Orbit Post Sitemap
The market is voting with its feet, awaiting the judgment of inflation data. On August 12, ahead of the CPI release, the three major U.S. stock indices all closed lower: the Nasdaq fell 0.6%, the Dow fell 0.34%, and the S&P 500 dropped 0.32%. The S&P maintained an extremely narrow intraday range for the fourth consecutive trading day, showing "undercurrents beneath the calm waters." AI Sector: Has the $500 Billion Positive Factor Turned into a "Negative Factor"? There is clear internal divergence within the AI sector. Semiconductors edged higher, but the gains fell far short of market expectations for major news. NVIDIA has joined forces with six Wall Street giants, including Blackstone and Goldman Sachs, to announce the establishment of a $500 billion financing platform for AI infrastructure construction. However, after the news broke, the market did not respond positively; instead, it triggered concerns about "circular financing"—that is, giants invest their funds in a circular manner among themselves, lacking a final supplier for demand. AI leaders plunged collectively: Google: down 3.84%, recording its largest single-day drop in nearly six months. Oracle: down 3.71%. Amazon: down 2.09%. Cloudflare: down 1.20%. The seven giants drag down the index overall, with the Nasdaq 100 closing just above the 50-day moving average. Storage Chips: Bucking the Trend Strength. Unlike the downturn in AI software, the memory chip sector performed impressively: SK Hynix: up over 4%. SanDisk, Seagate Technology: up over 2%. Micron: up 0.87%. The demand for AI infrastructure construction is being transmitted to storage hardware—large models need both computing power and storage. Memory chipsIn 2026, the biggest story in the global capital markets will still revolve around AI. Over the past two years, investors have focused on GPUs, chips, and servers. The rise of NVIDIA $NVDA tells the market: computing power is the most critical resource in the AI era. But after entering 2026, a new question is emerging: chips are in place, data centers are built, but what is truly limiting AI's continued expansion? The answer may not be chips, but electricity. The AI era is entering a new stage: from computing power competition to energy competition. 1. The biggest bottleneck in AI is shifting from chips to energy. In the past, the market believed that whoever had the strongest GPU would control the future of AI. So the funds are focused on chasing: GPU companies. Advanced packaging. High-speed storage. Server supply chain. However, as AI models continue to scale up, a real problem arises: large data centers consume massive amounts of electricity. Without stable power supply, even the strongest chips cannot operate. According to industry forecasts, global data center electricity demand will continue to grow rapidly in 2026, with AI-optimized servers becoming the main source of growth, and power supply capacity becoming a major limiting factor for AI infrastructure expansion. Simply put: future AI competition is not just about whose chip is stronger. It's also about who owns more: electricity resources. Data centers. Energy supply capacity. 2. Why is "electricity" becoming the new core of the AI industry chain? Many investors only see: NVIDIA is selling GPUs. But🔓Token unlocking does not equal a crash; the biggest bias in this market may be overturned by data. Tokenomist's study of 236 unlocking events reveals a counterintuitive pattern: the real price damage often concentrates in early-stage projects, characterized by extremely small circulating supplies and unlocking volumes that are large relative to total supply, which the market simply cannot absorb. In other words, unlocking itself is not scary; what is scary is a low circulating supply facing a large proportion of supply shock. 📊Zooming into the top 300 tokens by market cap, there are currently 10 projects that have not yet unlocked more than 30% of their total supply. This list itself is a potential volatility map: RaveDAO's $RAVE unlocking progress is 23.03%, Backpack's $BP at 25.00%, Raydium's $RAY at 26.00%, Lighter's $LIT at 26.10%, Zama's $ZAMA at 26.46%, Plasma's $XPL at 26.89%, SentientAGI's $SENT at 27.15%, chutes_ai's $SN64 at 27.94%, World Liberty Financial's $WLFI at 29.19%, and Falcon Finance's $FF is right at the critical 29.99% point. All are in the early supply release quadrant, meaning they still have a large amount of tokens waiting to enter circulation in the future. ⏳But the more critical variable lies in the time dimension BTC was mentioned 37 times in one hour, discussion speed still needs to be viewed over the entire day
OKX Onchain OS recorded 37 mentions of BTC in one hour at 14:00 on August 12, including 35 on X and 2 in news.
Compared to the 24-hour hourly average, this round's speed is 0.52 times, classified as "significantly slowed down"; the sentiment is 30% bullish and 16% bearish. There is no need to force these two lines into the same conclusion: popularity answers how many people are talking, sentiment answers which side the text leans toward, and neither can directly replace transaction volume and capital flow.
If the next round continues with speed, news sources, and actual market transactions together, confidence in judgment will increase; if it quickly returns to the average, this change is more like short-term noise. Real gold is soaring, while "digital gold" is playing dead — this major test of the safe-haven narrative is being reassessed.
On August 12, spot gold was around $4400/oz, climbing nearly 9% from $4048 on July 31 in just two weeks. COMEX August gold settled at $4383 on August 11, still far from the $5589 peak reached in January this year, but this rebound starting from the $4000 mark is the steepest of the year.
At the same time, BTC hovered at $63,594, down 0.53% in 24 hours and 0.89% over the week. The $64,000 level has become a battleground for bulls and bears, with resistance between $64,880 and $65,800. ETH looks even worse, fluctuating around $1890, having dropped nearly 2% on August 10 alone, struggling to hold above the psychological $1900 level. The Fear & Greed Index is at 26, clearly signaling "fear" in the market.
This comparison is painful. Over the past two years, the strongest narrative in crypto has been "BTC is digital gold," positioned as a hedge against fiat depreciation and sovereign credit risk. Now, with central banks buying gold, tariffs pushing up physical premiums, and inflation expectations fluctuating, safe-haven capital is voting with its feet — all flowing into that ancient asset with a 4,000-year history. Gold has risen 9% since August, while BTC remains stagnant and weak. This indicates that in the current macro environment, the market labels BTC not as a "safe-haven asset" but as a "high beta risk asset" — before CPI data is released, institutions' first reaction is to reduce BTC holdings and withdraw from ETFs, not to increase hedging positions. The July CPI, released at 8:30 PM ET on August 12, is expected at 3.4%. If the data is hotter, expectations for a September rate hike rise, and capital flows to yield-generating assets, with BTC taking the brunt. This is its nature, not its fault, but the "digital gold" certification is definitely under renewed scrutiny.
ETH's situation is another kind of awkward. Even the "digital silver" narrative is no longer mentioned — not disproven, just forgotten. Price-wise, $1850 to $1860 is recent support, with heavy resistance from $1930 to $1950. Since late July, ETH has been grinding within a narrow range of $1800 to $1950. Without an independent capital story or safe-haven attribute to leverage, its price moves entirely depend on BTC's mood, and since BTC itself is watching CPI closely, ETH's volatility is just a dampened echo.
Interestingly, SOL at $76.25 is up 0.39% in 24 hours and 3.66% over 7 days, showing relative strength amid a weak market. On August 10, SOL spot ETF saw a net inflow of $8.8 million, the best day since mid-May. With MoneyGram payment channels launching and on-chain weekly trading volume hitting records, capital is willing to assign SOL an independent logic. This, in turn, makes ETH look like an asset "with no support on either side": less of a safe haven than gold, and losing ecosystem capital heat to SOL.
So the core contradiction is clear: the demand for safe havens is real and strong, but the market is buying "sovereign credit hedges" — assets with zero volatility premium; meanwhile, the entire crypto curve remains a gauge of risk appetite. Gold tests credit, BTC tests liquidity, and ETH hasn't even entered the exam room. Until CPI is released, don't expect answers.过去十几年，比特币一直被市场赋予一个标签： 数字黄金。 很多投资者认为，比特币最大的价值在于稀缺性。 2100万枚固定供应，不受任何国家央行控制，因此它被看作是一种对抗货币贬值的资产。 但是进入2026年以后，一个更加明显的变化正在发生： 比特币$BTC 正在逐渐从一个另类投资品，转变为全球金融市场中的流动性资产。 它的价格走势，越来越受到美联储政策、美元流动性、科技股风险偏好以及机构资金配置的影响。 这意味着： 比特币正在进入一个新的阶段。 一、比特币正在告别过去的单一叙事 早期市场理解比特币： 更多来自于去中心化和货币替代。 那个阶段，投资者关注： 比特币是不是未来的钱？ 区块链能不能改变金融体系？ 但是随着机构资金进入，比特币的市场结构正在发生变化。 现在影响比特币价格的重要因素越来越多： 第一，美联储货币政策。 当市场预期降息，美元流动性增加，风险资产通常获得支撑。 第二，全球资金风险偏好。 当纳斯达克上涨，科技股情绪改善，比特币往往也会受到资金推动。 第三，机构资金流向。 现货ETF等金融产品让传统资金能够更加方便地配置比特币。 近年来，比特币市场流动性和机构参与度不断提高，$XSOXL is up around +5.04%, and this one caught my eye because semiconductor-related moves can accelerate quickly.
Still, I wouldn’t chase the current candle. I’d rather have a defined setup with a clear invalidation.
Trading Signal — LONG
Entry: 139–142
TP1: 146
TP2: 151
TP3: 158
SL: 135
If the entry area holds, I’d watch for continuation toward the targets. If 135 breaks, I’d leave the setup alone.Cloudflare $NET is currently caught between the reassessment of machine traffic and the lack of independent monetization of its developer platform, with a high market valuation premium awaiting the realization of actual customer unit prices.
Quarterly revenue rose to $696 million with a growth rate of 36%, coupled with DBNR rebounding to 120%, confirming the retention resilience of existing network layer calls.
Over 50% of traffic is already occupied by non-human requests, and the rapidly expanding base of new developers is prompting capital to reconsider the call frequency of AI Agents as the core growth engine.
Whether high-frequency automated traffic can be converted into higher API subscription tiers directly determines if the expansion of the developer base can offset the high bandwidth costs of the free tier.
If the expansion pace of large customers paying over $100,000 annually remains above 4,698, automated traffic will smoothly drive revenue tiers upward; if the growth rate of such customers slows to single digits per quarter, even with developers surpassing 7.4 million, the upward logic will fail.
Conversely, if 23% of proxy website entrances cannot be converted into independent business revenue, the overly inflated protection and bandwidth costs will erode profit margins; once quarterly revenue growth falls below 30% and DBNR declines again, valuation will face rapid compression.
The core of the current bull-bear divergence lies in whether the network infrastructure positioning can truly cross over into a high value-added commercial closed loop.
The single most important variable to watch in the next 7 days is the net increase trend of high customer unit price clients paying over $100,000 annually.
#Anthropic加快IPO进程，AI估值进入验证期 #贝莱德IBIT换购门槛降至100万美元 The "Dual-Currency Hedging" Game Before Tomorrow Night's CPI
On August 12, the crypto market collectively entered a "pre-data silent mode." Tomorrow (August 13), the US July CPI will be released. This is the only variable this week that can simultaneously influence US Treasury yields, the dollar, and risk appetite. The synchronized pullback of BTC and ETH today is the market's early response.
The difference in decline itself is a signal: under the same "pre-CPI deleveraging" logic, ETH's resilience is clearly weaker. This is not a coincidence—$BTC is supported by ETF funds, with a net inflow of $850 million into spot ETFs in the first week of August, 80% of which came from BlackRock's IBIT; meanwhile, ETH's institutional narrative has noticeably weakened in the second half of this year. When risk appetite weakens, the market sells beta first, and $ETH second.
On-chain and derivatives data tell the same story: BTC futures open interest shrank by about $2.3 billion over two days, with 90% of that from long liquidations, indicating typical active deleveraging rather than panic selling. The Fear and Greed Index is stuck at 28, in the fear zone, but the number of whale wallets holding over 10,000 coins has hit a six-month high—retail investors are exiting while large holders are accumulating. Such divergence often appears on the eve of a market turning point.
The core market contradiction is clear now: if CPI falls below 2.9%, expectations for a rate cut in September will be firmly locked in, and BTC is likely to retest the resistance zone between $64,900 and $65,000, with ETH following but with a weaker slope of recovery; if CPI exceeds 3.1%, against the backdrop of Brent crude oil already above $89, the "inflation rebound" narrative will crush rate cut pricing, and BTC will likely test supports at $63,000 or even $61,000, with ETH's downside risk even greater.
The answer will be revealed tomorrow night at 8:30. Until then, position sizing is more important than opinions. Tonight's CPI is most likely to kill not the shorts.
It's those bulls who think "good data means BTC will definitely rise."
Last week, nonfarm payrolls missed by 23,000, and May and June were revised down by a total of 103,000.
Employment data is already looking bad.
But BTC is still stuck below 64,000, and ETH hasn't truly reclaimed 1,900.
Because what the market is really struggling with now isn't whether employment is good or bad.
It's whether the Fed should trust employment data or inflation.
Cooling employment means continuing rate hikes will hurt the economy;
Oil prices are rebounding, and inflation remains above 2%, meaning not hiking risks runaway prices.
So whether to hike in September is again roughly a 50-50 call in the market.
Tonight, the overall CPI year-over-year expectation is 3.4%, and core year-over-year expectation is 2.5%.
The numbers look a bit lower than last month.
But what the market will really watch is whether core inflation month-over-month is picking up again.
If the overall data looks good and core is moderate, the market will believe again: weakening employment and falling inflation can coexist.
Then BTC reclaiming 64,000 and challenging today's 64,500 makes sense.
ETH also has a chance to truly reclaim 1,900.
But if the headline data looks good and core month-over-month is still hot, the most awkward situation arises.
The market will realize: employment is already deteriorating, but inflation hasn't fallen yet.
This is neither a rate cut trade nor a simple rate hike trade.
It's a repricing of "the economy is weakening, but rates can't come down."
In this environment, BTC support near 63,200 will be tested again; ETH's 1,850 area from yesterday might not hold either.
So tonight, I'm not in a rush to see if CPI is high or low.
I want to see if the market treats the data as a real positive after it comes out.
If it's positive but can't hold above 64,000, it means the market is weaker than the data.
If the data is average but it can break 64,500, it means the shorts haven't been able to push it down these days.
Don't use high leverage to guess the answer before the release.
Data is a multiple-choice question; market reaction is the real answer.
$BTC $ETH #今晚CPI公布，9月加息定价会改写吗？ #财报观察员：AI基建财报接力登场 #霍尔木兹通航谈判未果，美伊施压升级 The core of the current valuation debate for Cloudflare $NET lies in the revaluation of the high-barrier network traffic entry point towards AI Agent handoff points, which sharply offsets the high valuation digestion capacity caused by the developer platform not independently monetizing.
Market data confirms the infrastructure layer expansion logic, with quarterly revenue rising from $460 million to $696 million and growth accelerating to 36%. DBNR rebounded to 120%, validating the stickiness of existing application calls. The proportion of non-human traffic breaking 50% and 2 million new developers added in a single quarter have driven the market to reorder the driving factors as: Agent call frequency growth > developer base expansion > traditional reverse proxy share.
The prerequisite for the upside scenario is whether the high-frequency calls of automated traffic can effectively convert into high customer unit price conversion. If the number of large customers paying over $100,000 annually maintains an expansion speed above 4,698, and the growth of non-human traffic drives the API payment tier upward, it will support the valuation breaking through the upper boundary of a trillion market cap. This scenario requires continuous observation of the conversion rate improvement from free developers to paid API calls.
When the developer base breaks through 7.4 million but the growth rate of high customer unit price clients falls to single digits per quarter, it indicates an extremely high free tier but a blocked monetization chain, causing the upside logic to immediately fail.
The downside scenario logic is based on valuation premiums being eroded by unrealized business. If 23% of proxy website entries fail to generate independent revenue contributions on the developer platform, the market will quickly reprice the cost pressure of the free tier at high valuation levels.
Under this path, non-human traffic exceeding 50% may lead to disorderly increases in bandwidth and defense costs. If quarterly revenue growth falls below 30% and DBNR declines again, the price will face valuation compression and selling pressure.
In the next 7 days, key observations will focus on the net increase trend of high customer unit price clients and disclosures regarding the developer monetization mechanism.
#AI基建融资升温，英伟达英特尔路径分化 #特朗普媒体Q2加密亏损扩大，BTC持仓下降The focus tonight is to wait for the CPI to provide the interest rate path, rather than pressing positions on one-sided speculation before the data.
#CPI #Inflation #InterestRates #USD #Crypto #OKX
At 20:30, the US July CPI will be released: unadjusted annual rate previous value 3.50%, forecast 3.4%; seasonally adjusted monthly rate previous value -0.40%, forecast 0.1%. The judgment order is to first look at the actual value relative to expectations, then see if the USD and US Treasury yields confirm, and finally observe the secondary reaction in the crypto market and US stocks.
Crypto overall: If the cooler data leads to a decline in the USD and US Treasury yields, liquidity expectations improve, and BTC, ETH usually benefit first; but if it’s just a momentary sweep without USD and Treasury yields cooperating, the rebound’s sustainability will be discounted. Hotter data will suppress rate cut expectations, and the crypto market should first guard against a secondary drop in the high-leverage direction.
US stock risk assets: When close to expectations, the market may first fluctuate, then return to earnings and technical structure; below expectations usually benefits the Nasdaq and growth stock valuations; above expectations puts pressure on rate-sensitive sectors first. If the data is abnormally weak, it’s necessary to observe whether growth concerns outweigh the rate cut benefits.
For risk control, do not chase the first volatility at 20:30; first wait for 15-minute confirmation, then use the performance after the 21:30 US stock market open to verify the secondary reaction. The future AI computing power centers may no longer belong to the land.
Article | Alphabet AI
Following Elon Musk's plan to send AI computing power into space, another initiative is preparing to throw AI data centers into the sea.
Imagine, in a storm-ravaged offshore area, a huge spherical buoy is rising and falling with the waves.
Inside it is equipped with high-performance computing chips, continuously running AI models; externally connected to an energy system that can harness wave power; and the cold seawater serves as a natural cooling system.
This is not a scene from a sci-fi movie, but a future that a company named Panthalassa is trying to realize.
On August 7, it was reported that wave energy company Panthalassa is raising $225 million in new financing, targeting a valuation of about $2 billion, nearly doubling from three months ago.
The company's goal is to deploy GPU-equipped computing nodes offshore.
Traditionally, data centers have always been on land, close to cities and power grids, and reliant on large land areas.
But as AI model sizes continue to grow, training and running these models require more and more power, cooling capacity, and deployment space, naturally raising a new question:
If land is no longer sufficient, where else can AI infrastructure go?
A $2 billion valuation: an ocean energy company entering the AI era
In a sense, Panthalassa is very similar to SpaceX; from their names alone, both companies reveal their goals. SpaceX points to space, while Panthalassa’s name comes from the Panthalassa Ocean that surrounded the supercontinent Pangaea hundreds of millions of years ago.
The sky and the ocean are humanity’s oldest directions for exploring the unknown. For thousands of years, humans have looked up at the sky trying to understand the universe, and repeatedly sailed the seas seeking new routes and resources.
Today, these two directions are connected again through AI infrastructure.
SpaceX attempts to send AI computing power into space, while Panthalassa is preparing to send AI data centers into the ocean.
On August 7, it was reported that Panthalassa is seeking about $225 million in new financing, aiming for a post-financing valuation of about $2 billion. Just three months ago, the company completed a $140 million financing round with a valuation of about $1 billion.
The doubling of valuation in just a few months reflects capital’s interest in new forms of AI infrastructure.
In Panthalassa’s vision, the ocean can also become a place for data center construction—it just happens to provide several conditions needed for AI data center expansion.
First is space: compared to land, the ocean offers a much broader deployment area without occupying the increasingly scarce land resources around cities.
Second is cooling: one of the biggest costs for data centers is how to dissipate the heat generated by servers, and the ocean naturally provides a large volume of cold water, offering a new cooling environment for high-density computing.
Finally, energy: as data centers scale up, power becomes a critical limiting factor, and offshore areas have abundant natural energy resources, providing another possibility for future off-grid computing.
Against the backdrop of rapidly growing AI computing power demand, any solution that can provide additional energy and computing space is attracting capital attention, making the ocean a new space for imagination.
For Panthalassa, the AI era has given this originally ocean energy research company a new application scenario.
Deploying data centers into the ocean
Deploying data centers into the ocean is not a completely new idea.
As early as 2015, Microsoft launched the Natick project to explore whether underwater data centers could become a form of future cloud computing infrastructure.
In 2015, Microsoft completed the first phase experiment by deploying a small data center prototype in the Pacific Ocean off the U.S. coast. Then in 2018, it started the second phase test by deploying a larger steel container near the Orkney Islands, Scotland. This device housed 864 servers and operated underwater for two years.
The experiment results proved that the underwater environment indeed has some advantages: the sealed environment reduces equipment wear caused by oxygen and humidity; seawater provides a stable low-temperature environment, reducing cooling pressure; and standardized modular design also allows for rapid deployment of data centers.
After recovery in 2020, Microsoft stated that the server failure rate inside the device was even lower than the on-land control group, only about one-eighth of the latter.
However, Natick ultimately remained at the experimental stage. In 2024, Microsoft Cloud Operations head Noelle Walsh confirmed that the company would no longer build underwater data centers and said the team would apply the experience gained from the project to other fields.
China has further attempted to push underwater data centers from experimental exploration to industrial application.
In 2026, the Shanghai Lingang Underwater Data Center project officially started operation. Located in the East China Sea, this project adopts an "offshore wind power direct connection + seawater natural cooling" model, building platforms offshore, deploying data pods underwater, directly powered by offshore wind, and cooled by seawater. The total planned scale is 24 MW, with a first-phase demonstration project of 2.3 MW.
Compared to traditional land data centers, this model attempts to solve several practical problems: on one hand, land resources in the eastern coastal areas are tight, making it difficult to build large-scale data centers; on the other hand, offshore wind power resources are abundant but have been limited by transmission and power demand matching issues, causing some power to be underutilized.
Deploying data centers offshore can bring green power closer to computing power demand.
According to Xinhua News Agency, the project’s designed PUE (Power Usage Effectiveness) target is below 1.15, green power supply rate exceeds 95%, and seawater natural cooling is used to reduce cooling demand.
What makes Panthalassa special is that while other organizations move data centers into the ocean, Panthalassa moves from the ocean into data centers.
This company, founded nearly ten years ago, has long studied how to generate energy from wave motion, accumulating extensive marine engineering capabilities: how to design equipment that can operate long-term in marine environments, how to capture energy from wave motion, and how to deploy and maintain systems far from land. $HYPE Contract Watch:
HYPE's recent performance truly lives up to its name: the story is full of HYPE, and the candlesticks know how to create HYPE.
Fundamentally, Hyperliquid has real trading revenue and a continuous buyback mechanism, so every deep pullback attracts capital to take over; however, from a market perspective, HYPE is still in a correction phase after its initial rise, with overhead supply not fully absorbed, so short-term bulls cannot rely solely on faith to add positions.
The contract strategy is very clear:
If the price holds the recent support zone and rebounds with volume to reclaim short-term resistance, one can wait for a pullback confirmation before going long, targeting the previous highs first; if it spikes up but quickly falls back into the range, it indicates heavy selling pressure above, and it might just be another "breakout experience card."
If the key support is effectively broken and the rebound fails to reclaim it, the bullish logic temporarily fails, and one can follow the structure to look for shorts instead of holding positions while analyzing the project's fundamentals.
Tonight, the US CPI data will be released. As a highly volatile asset, HYPE may react more intensely than BTC. If the data is cooler, risk appetite may rise, making HYPE a flexible player; if the data is hotter, it could quickly turn from "HYPE" to "HELP."
Conclusion: The project is promising for the long term, but contracts should be traded based on the current structure. Do not chase the first breakout, do not catch the first falling knife, wait for confirmation, use stop-losses, and reduce leverage. The coin won't run away, but the margin really will. Dan Bin liquidated Apple and Tesla, bought SanDisk, AMD, Broadcom, and Intel—what signal does this send?
Apple and Tesla are end-product companies. Due to supplier price increases, to maintain profits, product prices must rise. However, suppliers like SanDisk have profit margins that have multiplied, and with relatively lower market caps, their growth potential far exceeds that of high market cap companies like Apple and Tesla. Dan Bin's move also reflects that big investors are extremely optimistic about storage and chips this year. After a wave of declines, these hardware companies benefiting from AI development are expected to see another surge in stock prices!Tonight, the US stock market once again stands at a critical juncture. Over the past two years, artificial intelligence has undoubtedly been the strongest main theme in global capital markets. From Nvidia's GPUs, to data centers from Microsoft, Google, and Amazon, and even chip companies like AMD, the market has formed a massive industrial chain centered around AI. However, entering the second half of 2026, the market is undergoing a major shift: investors are no longer satisfied with the "AI story." The market began to ask: With so much investment in AI, when will it truly generate profit? Who are the real winners? Who is just being pushed up by market sentiment? Tonight's volatility in the US stock market may be a new round of selection. The AI market is moving from a nationwide celebration into a true elimination phase. 1. AI is not over, but the market is beginning to raise its demands. In the past, the market traded about: "Who owns the AI concept?" Right now, the market is trading "Who can make money from AI." These are two completely different stages. From 2023 to 2025, the capital market has given AI extremely high valuations. As long as a company is associated with AI, the market will always have higher expectations. But now, the investment logic has begun to change. Large tech companies continue to increase AI capital investment, while cloud computing giants like Microsoft, Amazon, and Google are still building data centers and AI infrastructure. Market attention has shifted from simply the scale of investment to whether these investments can deliver long-term returns. Simply put: the market used to ask, "Do you have AI?" Now the market asks: "Can your AI be able to do it?"BTC holding near $64K while ETH and SOL edge higher looks more like selective risk rotation than a broad return of conviction. The divergence matters: capital is still willing to move down the curve, but not aggressively enough to confirm a durable risk-on regime.
CPI expectations may reset Fed pricing, yet the haven bid in gold and rising Hormuz pressure argue against treating softer inflation as a clean liquidity signal. My bias is cautious: relative strength in majors can persist, but macro uncertainty still favors disciplined positioning over chasing momentum.
Just my read, not advice.Tonight, don't rush to take a position on direction; wait until 20:30 for the CPI release to see if the US dollar, US Treasury bonds, and risk assets form a directional confirmation.
#CPI #USInflation #USD #USTreasury #Crypto #OKX
US July unadjusted CPI year-over-year previous value 3.50%, forecast 3.4%; seasonally adjusted CPI month-over-month previous value -0.40%, forecast 0.1%. The focus this time is not just on a single number but on judging whether inflation cooling can continue and whether the market will reprice the rate cut path.
Crypto overall: If the year-over-year rate is below 3.4%, the month-over-month rate is no higher than 0.1%, and the US dollar and Treasury yields fall in sync, risk appetite usually recovers first. BTC and ETH can serve as proxies to observe the overall crypto market reaction; if the data is hot, the US dollar and Treasury yields rise, and the crypto market is more likely to deleverage first. If the data is close to expectations, don't rush to chase the first wave of spikes or drops; wait for the second reaction 15 to 30 minutes later.
US stock risk assets: Moderate cooling benefits Nasdaq, S&P, and growth stock valuations; hot data will re-elevate rate pressure, with long-duration growth stocks bearing the brunt first; if inflation weakens significantly, beware of the market shifting from rate cut trades to growth concerns, and don't take the first wave of gains as trend confirmation.
Execution: first watch if the US dollar index, 2-year and 10-year Treasury yields, and US stock futures move in the same direction. Do not heavily position before data release; after the first wave of volatility ends, observe if the second reaction can continue.#今晚CPI公布，9月加息定价会改写吗？
The non-farm payrolls released unexpectedly weak data, showing negative employment growth in July, combined with significant downward revisions in the previous two months, clearly signaling a weakening labor force.
Strangely, the interest rate market has not directly ruled out the end of rate hikes; the probability of maintaining rates or hiking in September is almost evenly split.
The core logic behind this: The Federal Reserve is currently prioritizing inflation. Even if employment weakens, if inflation remains high, the option to raise rates is still on the table.
Tonight's July CPI is the most important short-term catalyst. Focus on core CPI:
✅ Inflation continues to decline: Weak employment combined with cooling inflation brings back easing expectations, benefiting overall risk assets.
⚠️ Inflation stronger than expected: Even if employment is weak, the market will reprice the risk of rate hikes, pushing up the dollar and U.S. Treasury yields, causing pressure and pullbacks on U.S. stocks, BTC, and ETH.
Everyone must avoid habitual thinking and not simply be bullish just because of weak non-farm data.
Weak employment does not immediately mean liquidity will ease;
Weak employment plus stubborn inflation is actually a nightmare for risk assets.
Before the CPI release, the market will most likely remain volatile and cautious; do not prematurely bet on a one-sided market.Gold is near $4,400.BTC is still around $64K.
Gold is benefiting from softer jobs data,weaker rate-hike expectations.
BTC is being treated differently. Despite periods of correlation with gold, its longer-term relationship remains unstable.Markets still view BTC more like a high-volatility risk asset .
So the key question is tonight’s CPI.
📉 Cool CPI→lower rate-hike odds→BTC could catch up with gold.
📈 Hot CPI→yields rise→BTC risks another downside test.
Gold has already chosen its direction.#比特币矿企Riot获Anthropic算力大单
The news that Riot secured a 10 billion AI computing power order has everyone shouting about a valuation restructuring for mining companies, but I think we shouldn't get too carried away.
Mining companies shifting to AI computing power essentially means mining profits are continuously shrinking, and they are looking for new ways to utilize their idle power plants and mining sites, which is a passive effort to develop a second growth curve.
The advantage is having ready power and site resources, so the startup cost is low; however, GPU operation and maintenance, client interfacing, and service stability are all more complex than mining. It's not as simple as swapping mining rigs for graphics cards to succeed.
During the previous bull market, I followed several mining coins, where people speculated on halving events and coin price elasticity, with sharp rises and even sharper falls. Now the narrative has shifted to AI, which sounds impressive, but most companies are just riding the hype and making empty promises. Very few can secure long-term stable large orders.
Moreover, even if the transformation is realized, mining companies will be competing with traditional IDC companies for market share, and their valuation logic must be completely recalculated. It's definitely not a simple "good news = price increase" scenario. I won't rush into mining coins because of this news; no matter how good the narrative sounds, if it doesn't translate into performance, it's all empty.
What do you think about this wave of mining companies turning to AI? Is it a real trend or just another round of hype?
$BTC $ETH Highly recommend everyone to pay attention to Cloudflare $NET. This company currently has a market value in the hundreds of billions online.
I only realized how promising this company could be after recently building my own website.
If you are like me now, rarely actively opening websites and mostly letting AI call them for you, then you will understand what I’m about to say.
Because Cloudflare has embedded one of its business scenarios into the AI-to-website intermediate link.
➠ In the past: I have a need → open website → website returns data
➠ Now it is becoming: I have a need → ask AI → AI accesses website/API → fetches and organizes data → returns to me
What it’s doing isn’t really creating a brand-new business. After breaking it down these days, I found that almost all its product launches revolve around this transmission path:
➠ Customer → Cloudflare → origin site/application ← Cloudflare ← external user or agent request
Including future business logic, it can be said that it is evolving from "hosting the internet" to "hosting the AI-empowered internet."
As of 2026 Q2, about 23% of websites worldwide already use Cloudflare as a reverse proxy (Figure 1).
In other words, Cloudflare already stands at the request entry point for a significant portion of websites.
Moreover, looking at data from the past seven quarters, it has preliminarily validated the growth capability of its original business segment (Figure 2).
➠ Quarterly revenue grew from $460 million to $696 million;
➠ Year-over-year growth rate increased from 27% to 36%;
➠ DBNR rebounded from 111% to 120%;
➠ Number of large customers with annual revenue over $100,000 increased from 3,497 to 4,698.
This original business isn’t really the sexy kind; it’s the old story.
The new story (AI requesting website access chain) is what makes me feel the space is huge.
Because one person cannot continuously open a dozen websites, repeatedly call dozens of interfaces, and perform tasks around the clock at the same time. But this behavior is exactly what AI can and will do.
According to Cloudflare’s own network observation, non-human traffic has exceeded 50% for the first time.
Although this includes not only AI agents but also traditional crawlers, automation tools, and malicious bots, the direction is clear: machines are becoming important visitors to the internet.
We also need to pay attention to the supply side that the platform itself is cutting:
➠ In 2026 Q2, Cloudflare added nearly 2 million new developers in a single quarter, while the entire year of 2025 added about 1.5 million developers. Currently, its active developers have exceeded 7.4 million (Figure 3).
This means more and more machines are "passing through Cloudflare," and more and more people are "building applications on Cloudflare."
More developers → more applications and agents → more computing and network requests → more calls to Cloudflare products → more usage and revenue.
This is where Cloudflare closes the loop of imagination:
In the past, it waited for websites to be built and then brought in traffic.
In the future, it hopes applications will be born on Cloudflare.
Of course, as I said, these are future matters and do not equal realized revenue yet.
Plus, Cloudflare has a large free tier, and the developer platform has not disclosed independent revenue, payment ratios, or average revenue.
But undoubtedly, this is a company worth tracking long-term.
The original post also included opinions on whether it’s worth entering now and how to enter, which I won’t repeat here, just picking some viewpoints as a supplement.#今晚CPI公布，9月加息定价会改写吗？
Tonight's CPI will be released, and it is highly likely that the interest rate will remain unchanged in September. There is no need to adjust positions prematurely based on this expectation.
The non-farm payroll data clearly shows the job market is cooling down, with significant downward revisions in the past two months. The Federal Reserve's policy always focuses on both employment and inflation. Since one side has already eased, even if core inflation remains somewhat sticky, it is impossible to raise rates solely based on one CPI report.
Currently, the market is split fifty-fifty on whether there will be a rate hike or not. Essentially, funds are gambling on expectations rather than there being a strong logic for a rate hike.
I've personally suffered losses betting on data several times. I used to think I could predict the market and would open heavy positions early to bet on the direction. Occasionally, I guessed right and made quick profits, but as soon as I was wrong once, the volatility would wipe out all previous gains.
After many mistakes, I realized that money made from data-driven trades is the hardest to earn due to its randomness.
My current approach is very steady: I hold my BTC base position without adding to bet on rate cuts or reducing to guard against rate hikes. I will wait for tonight's data to be released and for the market to show a clear direction before making any moves.
Earning a little less is fine; it's better than being trapped by a wrong bet.
$BTC Summary of today's US stock market
VIX remains low, the market hasn't turned bearish, but both SPY and QQQ are stuck in a high range, continuing short-term consolidation, waiting for CPI to break the balance.
VIX shows no risk signal
VIX closed at 15.29, which is relatively low this year, and continues to stay below the 10-day, 20-day, and 50-day moving averages.
There is still a large amount of Negative GEX near the 15 strike, and capital flow remains bearish, mainly because some are selling VIX Calls. Vol sellers still control the market.
For the past four to five months, VIX has mostly stayed below 20. This environment usually favors consolidation and buying on dips, rather than suddenly evolving into sustained panic selling.
So although the index is struggling to rise now, at least from the volatility structure, there is no sign of a major risk-off event yet.
SPY: 770 is the first line of defense, 760–767 is where I really want to buy
SPY closed around 772.8. After continuous gains earlier, it has clearly entered a high-level consolidation.
Short-term Gamma is mainly concentrated between 770–780, and the Put Wall has also moved up to 770, so recent oscillation around this area is not surprising.
The typical feature of a Positive Gamma environment is that there is resistance when prices rise and support when prices fall. So the first support level for SPY is 770.
If 770 does not hold, the next important support is 760–767.
In the longer term, the largest Gamma on the option chain is still concentrated at 800, so the overall trend is not bad.
There may be one more short-term pullback, but as long as 760–767 is not truly broken, it is still consolidation, not a trend reversal.
QQQ: 725 is a persistent resistance, 700 is the truly important support
QQQ is currently around 718. The 725 level above has repeatedly capped the price, becoming the most obvious short-term resistance. Around 700 below, there is the largest Gamma support, a gap, and multiple upward moving averages.
So QQQ is basically trapped in a clear range between 700–725. Recently, Call and Put flows have been switching back and forth without forming a clear one-sided direction. Also, volume has noticeably declined before CPI, and the market is reluctant to bet early.
Therefore, I think it is very normal to continue digesting within 700–725 over the next week or two.
Today's CPI is a catalyst to break the range-bound consolidation
JPMorgan currently expects:
Headline CPI MoM +0.12%, YoY 3.4%
Core CPI MoM +0.22%, YoY 2.5%
Their key threshold is Core CPI at 0.20%. If it is 0.20% or below, risk assets are more likely to see a Risk-On Rally.
If it is significantly above 0.20%, the market is more likely to trade down in the short term.
The options market currently prices about 0.9% volatility on CPI day. But an important background is that institutions' hedges are clearly biased to the downside. In other words, much of the downside risk is already protected in advance.
So if CPI is not particularly bad, hedge unwinding could push the index higher.
Another easily overlooked risk in CPI is gasoline prices
The average crude oil price in July actually dropped about 3% compared to June, but retail gasoline prices rose about 5%.
This means crude prices don't look strong, but consumers' energy costs have not decreased accordingly. If this part enters CPI, the data may be stickier than market expectations.
Combined with recent oil price rebounds, short-term inflation risk has not completely disappeared.
Another important variable in today's late-session decline: Treasury auction
After noon, bond auction demand was not particularly strong, and US Treasury yields rose again.
When yields rise, the Equity Risk Premium is compressed, macro books start reducing risk, and SPY and QQQ naturally get sold off in the late session.
Individual stocks
AMZN
AMZN has entered a clear consolidation after earnings.
Current price is about 272, short-term Gamma mainly concentrated between 270–280.
In the long-term option chain, 300 strike volume and positive Gamma are clearly increasing, so the market obviously still has players positioning for a move to 300.
But I still don't want to chase here.
A more comfortable position is the 4-hour EMA 21, around 266. If it can further return to the daily EMA 21, I think the odds will be better.
I am considering setting up out-of-the-money call spreads to play for a rally to 300.
NVDA
Similarly considering setting up out-of-the-money call spreads before earnings, aiming for a rebound to the previous high of 236.
SMCI
Revenue is strong, EPS significantly beats expectations, backlog and full-year guidance are also very impressive.
But the market still needs to verify whether this gross margin improvement can be sustained.
CRWV
CRWV surged directly after earnings, which is a very strong read-through for the entire Neocloud and AI Infra sectors.
Demand has not slowed, the real bottleneck remains power and computing capacity ramp-up speed. This is positive for similar companies like NBIS, IREN, etc.#黄金站上4400美元，避险需求升温
I am the mid-term intelligence analyst.
Gold breaking above 4400 superficially looks like a rise in safe-haven demand, but essentially it's a dual resonance of "interest rate expectations + credit repricing": July non-farm payrolls surprised to the downside, September rate hike expectations have receded, opening a window for real interest rates to decline, and a weaker dollar loosens zero-interest rates; although the Middle East has marginally eased, the US-Iran standoff and the Strait of Hormuz uncertainties remain, so the safe-haven premium hasn't truly withdrawn.
In the mid-term, I see a fluctuating upward shift in the central range, not a one-sided bull run. Central bank gold purchases (global +62% in Q2, China increasing for 21 consecutive months) and de-dollarization form the solid base, with 4000–4100 recognized by institutions as the bottom zone; however, a 7% weekly rise is already overbought, and if CPI rebounds or the Fed turns hawkish, a shakeout between 4360–4400 is inevitable.
Operationally, mid-term anchors expect a return to the upward channel within the year, with UBS forecasting 5000 in the first half of 2027. This wave is a pricing of monetary credit anxiety, not a pure safe-haven impulse, so buying on dips is more comfortable than chasing highs.
$XAU
#TradingVoice: Your experience deserves to be heardCracks in load-bearing walls are never marked on renovation plans, but when you see Sandisk's next quarter guidance—the median guidance, which acts as the main beam, suddenly half a notch below the market's expected load line—you should understand: this building, which just received an "overachievement" award, is putting up a "temporary support" sign for itself.
I come from an architectural design background and have spent my life dealing with foundations. Earnings surprises or misses are a delight or shock to retail investors, but to me, they are just concrete strength reports at different construction stages of a high-rise. Pouring the first three floors well doesn't mean you can build up to forty floors at the same pace—especially when you find that the diameter on the next floor's rebar purchase order has quietly been reduced.
Sandisk's problem this time isn't how solid the already built floors are, but that the construction permit for the next standard floor has been discounted. The guidance midpoint being below consensus is a negative deviation in the structural calculation report. Some peers might explain this deviation as a "seasonal wind speed adjustment," but a true structural engineer immediately picks up the geological report: Has the supply and demand of NAND flash's sand and gravel loosened? Is the planned tower in the AI storage roadmap actually growing from the foundation, or is it just a mass in a rendering?
Here lies a trap in construction organization. Last quarter's overperformance might have been achieved by compressing the curing cycle for a node sprint—if you've seen buildings where formwork was removed early to meet deadlines, you know that's not structural strength, but apparent strength. Market valuations are never based on the volume already poured but on the discounted future rental income of the entire building. Therefore, any load change in the guidance will cause the reinforcement ratio of that load-bearing wall in the model to be recalculated.
Regarding the $14 billion buyback: outsiders see it as positive—like a developer announcing an upgrade to the lobby stone. But my eyes focus directly on the capital allocation shear walls: Is this money pressed against the cross shear walls around the core tube, or diverted to decorative lobby curtain walls? If the buyback funds come from continuous external borrowing, it's like fixing a cantilever slab that shouldn't exist with high-strength bolts; when the wind blows, the joints will fatigue first.
As for the linkage with XAVGO, that's entirely part of the city's design logic. Two projects sharing the same site boundary and geological layer will have all crane swing data reflected on the central control room's screen. When Sandisk's settlement monitoring points issue warnings, XAVGO's external frame support plan must undergo the same review. The market isn't looking at today's profits but at the "geological survey report"'s rock shear zone—it doesn't just cut through one building; it slices a continuous path across the entire site.
So, the key isn't whether next quarter's guidance is cautious or pessimistic, but whether your blueprints have accounted for the "most adverse load combination." The iron law in construction is: the handbook strength of any material must be discounted. When Sandisk itself starts discounting, don't immediately say it's a "relaxed construction site"—you need to check if the crane base shows obvious uneven settlement. Foundation settlement never gives warnings; it only draws a diagonal crack on the load-bearing wall. #sandiskinvestorday The hourly chart's bullish divergence is starting to take effect, with the 633-635 dip yesterday and the 637-636 dip today. Both have some room to move. The market is fluctuating very slowly. Slowly enduring #今晚CPI公布，9月加息定价会改写吗？ #财报观察员：AI基建财报接力登场 #黄金站上4400美元，避险需求升温 $BTC 比特币持仓大户们，刚刚有了新动作。 Lookonchain监测显示，日本比特币储备公司MetaPlanet于1小时前转出1473枚BTC，价值约9382万美元。与此同时，比特币矿企Hut 8也在3小时前转出493枚BTC，价值约3136万美元。 两家公司合计转移1966枚BTC，总价值约1.25亿美元。 两家公司的背景 MetaPlanet是日本著名的“比特币财库”公司，其商业模式与Strategy类似——通过融资持续增持比特币作为核心储备资产。根据最新财报，该公司持有比特币账面价值约31亿美元，但2025财年因币价下跌录得约6.66亿美元的持仓估值损失。 Hut 8则是北美最大的比特币矿企之一，近年已从纯挖矿向能源基础设施平台转型，业务覆盖电力、数字基础设施和算力服务，管理约1020兆瓦能源容量。 这笔转移意味着什么？ 两家公司几乎同步转移大额BTC，可能的原因包括： 1. 托管钱包迁移或整理。 大型机构通常会将持仓分散在多个托管地址，定期进行归集或转移是标准操作。 2. 质押或借贷操作的前置步骤。 比特币抵押贷款市场正在快速增长，机构可能将BTC转移至特定托管方以获取流动性。 3.While BTC waits for the CPI at $64,000, the real turning point in the market is not direction but the quality of supply and demand. Ahead of the US July CPI announcement, BTC is consolidating around $64,000. This announcement is important not simply because of the inflation figure itself, but because it can reset risk asset preferences through its impact on Treasury yields and Fed expectations. The market has already priced in a significant portion of the Fed's September rate cut, so if the CPI comes out higher than expected, those expectations may retreat, putting downward pressure on both BTC and altcoins. Conversely, if the figure is lower than expected, it could strengthen rate cut expectations and act as a catalyst for risk appetite recovery. However, the key observation here is not predicting direction but the already confirmed change in the supply-demand structure. Last week, about $853 million net inflow occurred in US spot BTC ETFs. This means that institutional demand has not withdrawn while BTC consolidates; rather, accumulation is happening during the price adjustment phase. This contrasts with retail demand... If you feel that recent global macro data is getting harder to handle, don't doubt yourself, because the Federal Reserve has proactively smashed the market's streetlights.
After the new chairman Kevin Walsh took office, the Fed firmly held interest rates in the 3.5 to 3.75% range at the late July policy meeting.
This indicates that Walsh is quietly overturning the familiar forward guidance game played by his predecessors.
The mild Fed that used to politely negotiate with the market months in advance about whether to cut or raise rates is gone for good.
The exit of forward guidance has directly thrown global traders into an information black box.
In the Powell era, the Fed was afraid of excessive market volatility and would always give the market a heads-up through various leaks and official speeches before making decisions.
But Walsh’s approach is completely opposite; he prefers to maintain the mystery and absolute control of monetary policy, making all decisions fully dependent on the latest data.
For the crypto market, this is an extremely deadly volatility amplifier.
When the Fed no longer gives spoilers in advance, everyone can only react excessively like headless flies to each release of nonfarm payroll or CPI data.
Tonight’s CPI hasn’t been released yet, but the market options chain volatility has already been pushed up—this is a typical black box phobia.
This needs attention: the market’s pricing logic is undergoing a fundamental distortion.
In the past, everyone could comfortably look at the dot plot and plan asset allocation for the coming year.
But now, due to the lack of forward-looking signals, high-frequency algorithmic trading can only perform extreme tug-of-war within milliseconds of data release based on the absolute value of the data.
This means any slight inflation surprise can directly trigger a localized liquidity stampede.
Worse still, Wall Street is currently even trading defensive positions for further rate hikes in December, which was almost unthinkable before.
Personally, I think Walsh chose to smash the streetlights to reclaim the Fed’s absolute pricing power as a central bank.
When the market is fed too gently by forward guidance, the central bank gets hijacked by market expectations.
But for highly volatile crypto assets, without the Fed’s certainty milk, macro games driven purely by data will cause the market’s risk premium to soar sharply.
The Bitcoin in your hands has to endure more collateral damage and selling pressure released from traditional macro algorithms.
The flip side of data dependence is ubiquitous surprises and shocks.
In this spoiler-free black box era, any attempt to front-run the Fed may become cannon fodder feeding high-frequency algorithms.
Buckle up, because in the future macro sky, thunderstorms and clear skies may alternate in the blink of an eye.
#今晚CPI公布，9月加息定价会改写吗？ #黄金站上4400美元，避险需求升温
Gold has risen above $4400, reaching an intraday high of 4435, just one step away from 4500. COMEX was even stronger, surging to 4495 intraday. It has gained over 7% in a week, rebounding more than $400 from the late July low. This pace is quite fast for an asset of gold's scale.
Several forces are pushing simultaneously. The most direct is the nonfarm payrolls, which dropped by 23,000, turning negative. The probability of a September rate hike fell from nearly 60% to around 40%, weakening the dollar and causing U.S. Treasury yields to decline, thus lowering the opportunity cost of holding gold. On the other hand, the Hormuz Strait agreement stalled again, with Iran saying the strait will remain closed unless the U.S. accepts their conditions. Oil prices have returned above 88, fueling inflation expectations. Weak employment reduces rate hike pressure, while high oil prices support inflation, benefiting gold from both sides. Central banks are also buying; China's central bank has increased holdings for 21 consecutive months, adding 640,000 ounces in July alone. Global central banks net increased 289 tons in Q2, up 62% year-on-year. Funds are also flowing back, with domestic gold ETFs attracting over 10.5 billion since July.
But one detail is worth noting. After gold surged to 4435, it retreated to around 4368, closing below the 100-day moving average. The rally and pullback indicate significant profit-taking around the 4400 level. Today's CPI data is also due, with market expectations for core CPI monthly growth at 0.2%. If the data exceeds expectations, the probability of a September rate hike may bounce back.
The medium- to long-term logic for gold remains solid—central banks are buying, U.S. dollar credit is loosening, and geopolitical risks persist. Standard Chartered calls for 5000, RBC's high scenario sees 5300. But in the short term, I wouldn't chase above 4400. I'll wait for the CPI release and consider buying on a pullback to the 4250-4300 range. The direction is likely upward, but missing a day or two won't matter.Brothers
Tonight's US #今晚CPI公布，9月加息定价会改写吗？ Most likely to be flat or slightly soft, with a high probability of no rate hike in September.
Pre-market has already jumped ahead: gold taking off, Korean index surging, US stocks rallying pre-market. Risk assets are collectively betting on no rate hike.
If CPI meets or falls below expectations → weaker dollar, cooling rate hike expectations → high beta assets like BTC/ETH will directly follow the rally. Only if the core CPI unexpectedly heats up (month-on-month 0.3%+) might there be a short-term sell-off. Aggressive rate hikes before the election are difficult anyway; unless the data is explosive, crypto still has the tailwind.
Tonight is a window to confirm risk appetite, not a major reversal. What do you all think? Will it first spike then fall back, or take off directly? The global market is awaiting a key piece of data. Tonight, the US July CPI data will be released. Many investors believe this is just ordinary economic data. But in reality, for global risk assets in the second half of this year, CPI may become a key turning point in determining market direction. Because the market is now trading not just corporate earnings, but the Fed's monetary policy path in the coming months. US stocks, gold, Bitcoin, and even the entire crypto market are all waiting for this answer. 1. What the market is really waiting for is not the CPI, but the signal of rate cuts. Over the past two years, the Fed's biggest task has been to control inflation. After entering a cooling cycle from the era of high inflation, the market has been eagerly awaiting the question: When will interest rate cuts begin? How big is the rate cut? When will funds flow back into risk assets? Therefore, every CPI release becomes a moment for the market to reprice. Currently, the market expects inflation to continue to slowly decline in July, and investors hope to see clearer signals of inflation easing, supporting a more accommodative monetary policy in the future. Simply put: CPI is below expectations, and the market will believe inflationary pressures continue to decline. The Fed has more room to cut rates. The US dollar may weaken. Funds are flowing back into stocks and crypto assets. And if CPI exceeds expectations: the market may re-worry about fluctuating inflation. Interest rate cuts have been delayed. U.S. Treasury yields rose. Tech stocks and the crypto sector are under pressure. So tonight's data, on the surface, looks like inflation data, but in reality, it's a vote on liquidity. 2. Beauty【 Claude Starts Competing for Electricity with $BTC Mining 】
Electricity and data centers are occupied -> BTC mining competition decreases -> BTC mining difficulty lowers -> BTC costs decrease
Anthropic and Riot sign a new long-term AI data center contract
Locked in 191 MW capacity at Riot's Rockdale, Texas campus for 20 years
Riot Platforms is a US Bitcoin mining company and data center operator, Nasdaq ticker RIOT
This long-term contract is enough to power about 143,000 households
Riot expects contract revenue of $16.1B
The first 96 MW is expected to be delivered by December 2027
Fully operational by June 2028
The contract runs until June 2048
Mining machines themselves cannot run Claude
Mining companies have land, power access, cooling, and construction capabilities#CPIToResetFedBets #AIInfraEarningsWatch #Gold4400HavenBid BTC Today's Trend and Market Outlook
Today, BTC maintains a high-level pullback pattern, repeatedly testing the $65,000 resistance without success. Bullish momentum continues to weaken, and the market has entered a phase of consolidation and digestion. The daily chart keeps showing upper shadows, the 4-hour indicators are weakening, and the market lacks incremental buying pressure to push prices higher. ETF funds have seen a phase of outflows, and short-term selling pressure is gradually emerging.
The current market is highly tied to U.S. inflation data. Hawkish remarks from the Federal Reserve repeatedly disturb the market, delaying rate cut expectations. The overall market struggles to develop an independent trend, with small-cap and platform tokens strongly linked to BTC's movements. The first short-term support is at $62,400–$62,600; if this holds, the consolidation range will continue. If there is a volume-driven break below, it will trigger contract liquidations, with strong support expected at $61,000. On the upside, resistance is at $65,000; only a volume-backed close above this level will restart the bullish trend.
From an operational perspective, avoid chasing highs at present. Short-term traders should consider light positions only after support shows signs of stabilization, with strict stop-loss settings. Those already holding positions should closely watch the defense strength around $62,400. Long-term investors need not be disturbed by short-term volatility and should wait for pullbacks to build positions gradually. The overall market is at a directional choice window with increased uncertainty; it is essential to control position sizes and avoid heavy leveraged bets.Ethereum $ETH and Bitcoin $BTC have gained over 4% so far after accumulating at low levels yesterday. Entering with small leverage and divided positions, Ethereum $ETH is expected to have a slight rise in the next few days, then continue to oscillate for a while to shake off the undecided friends. My expectation is that it will still reach above 2300 later, so brothers, keep your mindset steady and stay firm #今晚CPI公布，9月加息定价会改写吗？ #机构逆势加仓：SharpLink增持近4万枚ETH #交易之声：你的经验值得被听到 🔔The evening CPI data is coming in strong, will the consolidation pattern be broken?🔥🔥🔥
CPI is the Federal Reserve's monetary policy indicator: the higher the inflation, the weaker the rate cut expectations, the stronger the dollar, and the more pressure on cryptocurrencies; when inflation cools, rate cut expectations rise, the dollar weakens, benefiting the crypto market.
This time, focus on two key indicator groups:
1. Overall CPI year-over-year: previous 3.5%, expected 3.4%
2. Core CPI year-over-year (more closely watched by the Fed): previous 2.6%, expected 2.5%
Three data scenarios correspond to market trends
Inflation across the board exceeds expectations (bearish, decline)
Condition: CPI year-over-year > 3.4% and Core CPI year-over-year > 2.5%
Interpretation: Inflation decline is less than market expectations, the Fed's September rate cut expectations cool significantly, high interest rates remain longer.
Market performance:
1. Dollar index surges, Nasdaq plunges;
2. Market quickly sells off, short-term spike down to support levels;
Trading approach: wait and see, do not bottom fish, if rebound faces pressure, follow the trend to short, strictly use stop loss.
Inflation across the board below expectations (bullish, rebound rise)
Condition: CPI year-over-year < 3.4% and Core CPI year-over-year < 2.5%
Interpretation: Inflation continues to cool, market prices in soaring probability of September rate cut, easing liquidity expectations improve.
Market performance:
1. Dollar weakens rapidly, risk assets collectively recover;
2. Crypto market directly rallies, breaks short-term resistance line;
Data divergence (bull-bear tug of war, wide volatility)
Typical divergence: overall CPI meets decline target, but core CPI exceeds expectations; or vice versa.
Interpretation: Market shows divergence in Fed policy judgment, funds quickly switch between long and short.
Market performance: first spike down then pull back, or surge then quickly fall back, intense back-and-forth within 15 minutes causing stop losses.
Trading approach: do not enter immediately after data release, wait 30 minutes for the market to digest and form a clear range before trading, reduce position size to lower risk.
$BTC $ETH Gold long-term layout continues with low-buy operations $BTC
Gold H1｜Moving averages show a strong bullish trend, but the MACD momentum bars continue to shrink, with price making new highs but momentum lagging, signaling a bearish divergence warning.
At this stage, do not chase highs; wait for a pullback to support and stabilization before following the trend; do not short solely based on exhaustion signals at the top, wait for a confirmed breakout.
#XAUUSD #PreciousMetalsTechnicalAnalysis
⚠️ For review and communication only, not trading advice Complete list of today's gainers and losers
🟢 Gainers
$APR MEV Infrastructure Track | +77.76%, trading volume 859 million, short-term funds concentrated attack
$CRWV AI Computing Power Track | +21.57%, trading volume 127 million, sector funds significantly returning
$BEA AI Blockchain Gaming Track | +18.67%, trading volume 4.484 billion, sector funds rotating upward
$SKUU Semiconductor Leveraged ETF Track | +14.44%, trading volume 14.8934 million, thematic rotating funds entering
$AEHR Semiconductor Equipment Track | +13.62%, trading volume 5.2984 million, sector heat continues to ferment
$SNXX Storage Chip Leveraged ETF Track | +13.28%, trading volume 389 million, funds continuously positioning
$NBIS AI Cloud Computing Track | +13.22%, trading volume 107 million, sector heat warming up
$KORU Stock Index Leveraged ETF Track | +13.09%, trading volume 398 million, incremental funds continuously flowing in
$AEON Emerging Public Chain Track | +12.37%, trading volume 98.5905 million, new coin heat rising
$CAP Crypto Payment Track | +11.76%, trading volume 1.308 billion, sector funds significantly returning
$RAM Technology Stock Leveraged ETF Track | +11.58%, trading volume 3.1585 million, short-term sentiment heating up
$HUS AI Social Track | +11.38%, trading volume 44.8409 million, sector rotation active
$RKLB Commercial Aerospace Track | +11.12%, trading volume 58.5643 million, thematic funds entering
$BOT Quantitative Strategy Track | +10.92%, trading volume 11.7438 million, sector heat continues to ferment
$NOK Communication Hardware Track | +10.87%, trading volume 6.4867 million, tech concept strengthening
$MVLL Optical Module Leveraged ETF Track | +10.79%, trading volume 5.8508 million, short-term funds positioning
🔴 Losers
$ONE Cross-chain Public Chain Track | -38.32%, trading volume 119 million, large capital outflow
$DOS Web3 Infrastructure Track | -30.73%, trading volume 935 million, profit-taking concentrated
$KAITO AI On-chain Data Track | -15.84%, trading volume 438 million, short-term sector sentiment fading
$BICO Gas Relay Infrastructure Track | -14.16%, trading volume 364 million, selling pressure continuously released
$SKDD Semiconductor Inverse Leveraged ETF Track | -13.63%, trading volume 8.4194 million, inverse targets weakening
$MMT On-chain Strategy Protocol Track | -12.15%, trading volume 176 million, short-term funds withdrawing
$BABY BTC Staking Track | -11.63%, trading volume 10.7125 million, sector heat declining
$UNI Decentralized Exchange Track | -10.71%, trading volume 321 million, DeFi sector under pressure
$OFC GameFi Gaming Track | -10.69%, trading volume 8.8621 million, thematic speculation ended
$LA Zero-Knowledge Proof Infrastructure Track | -10.67%, trading volume 15.4315 million, funds continuously exiting
$WAL Decentralized Storage Track | -9.48%, trading volume 1.8904 million, market attention declining
$RVN POW Asset Issuance Public Chain Track | -8.95%, trading volume 30.3289 million, selling pressure gradually increasing
$UB On-chain Data Platform Track | -8.62%, trading volume 131 million, short-term sentiment weak
$MEGA Ethereum Layer 2 Track | -8.40%, trading volume 16.1761 million, sector entering adjustment phase
$BILL RWA Real-World Asset Track | -7.27%, trading volume 21.2433 million, bullish confidence insufficient
$ZBT Modular Public Chain Track | -7.04%, trading volume 26.9245 million, price oscillating downward
(Data as of the time of publication)
⚠️ Reminder
1. The big gainers are all short-term speculative funds colluding with no long-term positive support; heat fading will cause a cliff dive, chasing highs will lead to deep losses;
2. The declining coins face continuous selling pressure; bottom fishing can easily lead to more losses, do not rush to bottom fish;
3. Contract volatility is extreme; a 10-20 point rise or fall can directly liquidate and wipe out principal; heavy positions and borrowing to trade are prohibited;
4. The current market is severely fragmented with huge divergence in gains and losses; do not follow trends based on feelings, mainly observe.
⚠️ Risk Warning: Virtual currency trading is highly risky, and cryptocurrency contract volatility is extreme! This article only organizes market data and does not constitute investment advice, does not encourage any trading operations, please do not blindly follow trades. Chip concentration has risen to 14.8%!
Half a foot has stepped into the "high-risk zone." Attention! The risk here does not refer to rising or falling, but to volatility.
Chip concentration cannot predict direction, but based on historical data, my friends and I have found that there seems to be a pattern:
When the curve starts to turn, if BTC's price was rising before this, then the probability of continuing to fluctuate upward is greater; conversely, the probability of continuing to fluctuate downward is greater (as shown in the chart);
However, at this moment, the curve is still rising continuously. So, we cannot yet predict which direction has a higher probability next.
But what is certain is that risk is accumulating, and volatility is brewing......🚨 BTC & ETH JUST GOT HIT — AND THE REAL DRIVER ISN’T CRYPTO.
Something felt off as the night session opened.
BTC and ETH sold off sharply after fresh Strait of Hormuz tensions, while oil $CL held above $82. That’s a sign markets are starting to price geopolitical risk back in.
Here’s the chain traders are watching:
🛢️ Hormuz risk → oil higher
📈 Oil higher → inflation expectations rise
🏦 Higher inflation → fewer Fed cuts
⚠️ Less easing → pressure on risk assets
The US-Iran talks still look far from a real breakthrough, with the biggest issue being how any agreement would actually be implemented.
Now comes the key test: CPI.
If inflation continues cooling, some of this pressure could ease.
But if CPI comes in hot, crypto could face a nasty macro + geopolitical double squeeze.
$ZBT
$ETH
#Gold4400HavenBid #CPIToResetFedBets #AIInfraEarningsWatch #Gold4400HavenBid Even with a market correction, Solana's ecosystem value locked continues to soar, and on-chain active addresses are approaching historic highs. The market is betting on its "Ethereum killer 2.0" narrative, but $SOL's valuation is no longer cheap, and you need a clear framework to handle volatility. Outline of this article - 🏢 What exactly is it - Why 🔥 is it being hyped now - 📊 Fundamental highlights - ⚖️ Long-Bear games - 🎯 How to view and participate 1. What 🏢 exactly is Solana is a high-performance public blockchain, with its core selling point being "fast and cheap"—theoretically processing thousands of transactions per second, each with a fee of less than 1 cent. Unlike Ethereum, which relies on Layer 2 scaling, it directly combines PoH (Proof of History) and PoS consensus at Layer 1, enabling smooth operation of on-chain DEXs, lending, and NFT markets. The team divides the ecosystem into clear modules: DeFi includes Jupiter (aggregate trading), Raydium (AMM), and Marinade (liquid staking); NFTs include Mad Lads and Tensor; Payment is Solana Pay; Even in the DePIN (decentralized physical network) sector, Helium and Hivemapper have migrated their devices to Solana. Its revenue source is straightforward: fees + MEV (Maximum Withdrawable Value). The more prosperous the ecosystem, the $SOL#CPIToResetFedBets The July U.S. CPI report is the market’s most important immediate catalyst. July payrolls reportedly fell by 23,000, while May and June were revised down by a combined 103,000. That initially reduced expectations for another Federal Reserve rate increase, but CME probabilities have returned to nearly a coin toss. Economists expect headline CPI to rise approximately 0.1% month-on-month and slow from 3.5% to 3.4% annually. Core inflation is forecast near 0.2% monthly and 2.5% annually.
A cooler report would reinforce the weak-employment argument and could support bonds, equities, BTC and ETH by reducing pressure on interest rates. A hotter core reading—particularly in housing or services—could strengthen the dollar and lift Treasury yields. My view is that the composition will matter more than the headline alone. Energy prices may create additional inflation pressure in future reports, so even a favorable July number would not completely settle the September debate. The divergence between Nvidia and Intel in AI infrastructure financing increasingly resembles a divide between two eras.
Nvidia is packaging GPUs, data centers, and computing power cash flows into a new asset class, pulling Wall Street along for financing. It’s not just selling chips but helping customers solve the problem of "not being able to afford chips." Computing power becomes collateral, AI factories become infrastructure, and financial institutions come in to capture long-term returns.
Intel’s path is more traditional: relying on CPUs, packaging, foundry services, heterogeneous architectures, and partnerships to reinsert itself into the AI infrastructure chain.
This isn’t about who’s right or wrong, but about different positions. Nvidia’s challenge is the fear of being accused of circular financing and asset bubbles; Intel’s challenge is the fear of falling behind in technology and ecosystem, with capital unwilling to wait long enough.
I think this divergence is very important. The second half of AI is not just a chip performance battle but also a financing capability battle. Whoever can convince capital that "computing power will continue to generate cash flow" will be able to keep expanding.
#AI基建融资升温，英伟达英特尔路径分化 Launch of Staking Ethereum ETF: How Traditional Finance Redefines ETH Asset Attributes
With the official listing of BlackRock's staking Ethereum ETF, the spot Ethereum sector is undergoing a critical transformation. Early U.S. spot Ethereum ETFs only provided pure price exposure and could not participate in on-chain staking rewards, which was a core reason many native crypto participants remained cautious about such products. After regulatory easing, the staking rewards mechanism has been incorporated into the compliant ETF framework, opening channels for traditional capital to access Ethereum's on-chain yields.
From a product structure perspective, fund managers will entrust professional node operators to stake most of the ETH holdings on the network, and after deducting management fees, distribute the staking rewards to holders. This change means Ethereum is no longer just a pure price speculation asset but begins to exhibit characteristics of an "income-generating infrastructure asset" within traditional asset allocation models.
However, there are clear market divergences. Some asset management institutions proactively remove staking-related clauses in their filing documents, opting to simplify product structures and prioritize regulatory approval certainty. The coexistence of these two paths means the Ethereum ETF sector will continue to diverge internally.
On the capital side, the overall scale of Ethereum ETFs remains significantly smaller than Bitcoin ETFs, and institutional capital deployment is more cautious. It is also important to recognize that ETF capital flows are influenced by multiple factors such as macro liquidity, regulatory news, and market sentiment; net inflows do not guarantee a perpetual trend.
From an industry perspective, the greatest significance of staking ETFs lies in lowering the technical barriers for traditional institutions to participate in the Ethereum ecosystem. However, regulatory uncertainty remains long-term, and countries have yet to unify rules regarding crypto asset yield products. All market participants should rationally distinguish narrative expectations from fundamental realities and fully acknowledge the inherently high volatility risk of digital assets
#今晚CPI公布，9月加息定价会改写吗？ #财报观察员：AI基建财报接力登场 #黄金站上4400美元，避险需求升温 $BTC $ETH $BEAT #今晚CPI公布，9月加息定价会改写吗？
$BTC AI hardware collectively surges, is the pre-CPI rally a rush of funds or a leak of information?
Before tonight's CPI release, there is a signal on the market worth pondering: risk assets suddenly rise collectively, sweeping away the weakness of the past few days. BTC rebounded from the intraday low of $63,204 to around $63,794, temporarily halting the short-term downtrend; pre-market US stocks in the AI hardware sector are all in the green, $SNDK up over 2.6%, $MRVL up nearly 1.8%, LITE up 0.8%, Nasdaq futures also rose about 0.4%, the whole market strongly showing a Risk-On sentiment.
This rally is not baseless speculation; there are two solid logical supports behind it.
On one hand, US Treasury yields have temporarily eased, with the 10-year yield falling from yesterday's high of 4.735% to around 4.68%. As the risk-free rate declines, valuation pressure on growth stocks and crypto assets naturally eases, providing a window for funds to bottom-fish. On the other hand, AI hardware stocks have their own fundamental support: SNDK's data center revenue surged nearly 400% year-over-year, LITE's quarterly revenue exceeded $1 billion, more than doubling year-over-year, and MRVL continues to benefit from AI custom chips and high-speed interconnects. With these solid performance fundamentals, any sentiment recovery easily triggers a leading rebound.
As for the market talk about "CPI data leaks and funds positioning early," I think the probability is low.
If there were insider funds acting early, the gains would not be so restrained, nor would the market move so sluggishly. This looks more like a routine large fund positioning on the eve of CPI—betting on moderate inflation decline, adding some risk asset exposure first, a typical expectation trade, not really a "leak or early run."
Of course, all predictions must wait for the 20:30 data release to verify.
If CPI is truly below expectations, US Treasuries continue to fall, and Nasdaq and BTC can rally with volume and hold gains, then this rebound can be considered a confirmed trend; if the data seems positive but the market collectively spikes then falls back with funds selling into the rally, then today's rise is just an early anticipation, and caution is needed for a possible reversal pullback.
In the end, the pre-market rise is all about sentiment and expectations; the strength of the follow-through after the data release is the real answer tonight. #财报观察员：AI基建财报接力登场 #黄金站上4400美元，避险需求升温 88 hours ago this thing was squeezed 40% up on a $2.5B FDV nobody could justify. Two days later shorts piled in on a -43% crash while longs still outnumbered them 62/38. Now it's sitting at -27.7% from that first flag.
No squeeze, no rescue, just a slow bleed while longs kept averaging down the whole way. The mechanic that usually pays out here is a crowded short getting run over, this time the crowd was long and the market just walked away from them instead.
NFA.The Federal Reserve injected $51.79 billion in liquidity in a single day to stabilize the market
Short-term funds loosened, BTC and the US stock market saw an immediate rebound, giving bulls a temporary breather
But the key point is, this is just an overnight short-term repo, with all funds withdrawn the next day, so it’s not a full QE
The real direction will be set by tonight’s CPI data; if inflation exceeds expectations, this rebound is just a bull trap, leading to a sharp drop. Only when inflation cools down can a sustained rally occur! Before tonight's CPI release, the biggest risk in the market is not the data itself, but the overly extreme positioning of funds.
CTA trend funds have pushed bond short positions to historic highs, effectively turning the market into a pressure cooker.
1. Risk is extremely asymmetric
The space for shorts to push lower is very limited, but if the CPI is even slightly lower or core inflation signals some cooling, it will immediately trigger a short squeeze. Shorts clustered at high levels will have to frantically buy back government bonds to cut losses, causing bond prices to surge and yields to plunge instantly.
2. Why are funds daring to bet?
Recently, the oil price rebound combined with large bond issuance has instilled a stubborn inflation expectation in the market. Traders bet that the Fed will still raise rates in September and have been adding shorts along the trend, pushing positions to the limit.
3. Market outlook
Even if tonight's data only meets expectations, the probability of a bond market rebound is very high, with yields adjusting downward. The positioning structure determines short-term elasticity, and the safety margin clearly favors the longs now. Only a CPI surge far above expectations would allow shorts to exit unscathed, but that probability is low.
In such an overcrowded trade, blindly chasing bond shorts has very poor risk-reward. Betting on a yield pullback is actually more stable.
Do you think tonight's data will force these shorts to cover?
Not investment advice, DYOR
#今晚CPI公布，9月加息定价会改写吗？ SK Hynix is making big moves in Dalian.
The NAND Phase II factory, which has been halted for four years, is now officially restarting. Phase I has a monthly capacity of 100,000 wafers, and Phase II will add about 50,000 more, increasing overall capacity by 50%. Why was it stopped for four years? The memory market was too bleak. Why restart now? AI has directly exploded enterprise SSD demand, making expansion unavoidable.
Counterpoint data shows that enterprise SSDs now account for 48% of global NAND shipments, up from 26% this time last year. Kioxia has also stated that flash memory orders are already booked through 2027. SK Hynix is not expanding ahead of demand; it is being forced to expand by demand.
But there is a divergence worth considering. On one hand, demand is indeed fierce, and AI data centers are still being built frantically. On the other hand, capacity is being released in concentration, with institutions predicting that NAND supply may become loose in the second half of 2027, putting downward pressure on prices. Short term, there is still a shortage; mid-term supply is accelerating; long term, prices may ease.
What does this have to do with the crypto world?
No direct short-term impact; NAND is not the core chip for mining machines, so it has limited effect on miner costs. But there is a more important signal behind this—the entire semiconductor supply chain's expansion cycle is accelerating. DRAM, HBM, NAND are all expanding, and upstream wafer foundries, equipment, and materials are all running at full capacity. The cost of the entire computing infrastructure will not decrease in the short term; miners and AI computing projects will have to continue bearing high hardware costs.
Here is my view. SK Hynix restarting Dalian Phase II is driven primarily by AI-driven enterprise storage demand. As long as AI computing power keeps expanding, storage demand will not stop. But capacity ramp-up takes time; from equipment installation to mass production still requires over half a year, during which the supply gap remains.
Storage is always in a cycle of shortage, expansion, oversupply, cutbacks, and shortage again. SK Hynix is betting that this AI-driven cycle will be longer than before. Whether this bet pays off will be clear by 2027.
$BTC $ETH $SKHY