Orbit Post Sitemap
$DOGE
Looking back at the trend over the past six months, it has generally continued to oscillate downward with weak rebounds. Each time it surges on news related to Musk, it mostly fails to avoid falling back.
The fundamental hard issues cannot be ignored: there is no total supply cap, with a fixed annual increase of 5 billion coins, resulting in long-term continuous inflation and a lack of deflationary expectations to support it. Previously, the market bet on the X platform integrating DOGE payments, but it has yet to materialize, causing the biggest narrative to continuously fall short.
Another obvious point is that the pull effect from Musk's statements is weakening, and pure news is unlikely to trigger a large-scale rally.
The advantage lies in being an established meme coin with ample liquidity, maintaining elasticity when the overall market warms up. However, the current market highly depends on speculative sentiment, and without sustained positive catalysts, it is difficult to develop an independent trend.
Going forward, focus on two signals: progress related to X payments and changes in overall market risk appetite.
This is only a personal market record and does not constitute any investment advice. Hang Seng Tech fell again, JD dropped 10%, Meituan dropped 6%.
I really don't want to hold anymore; the holding experience is too bad.
Besides consumption and performance, there are quite a few problems to consider.
US tech stocks are crazily pouring capital expenditure into AI, $NVDA $META $GOOGL valuations are actually getting bigger, after all, this is buying future productivity.
Chinese tech companies are still making money and have decent cash flow, but they have only been optimizing existing businesses, cutting costs, and improving efficiency, without creating new profit pools.
Operating in mature markets, essentially all are intermediary businesses: Alibaba, Meituan, JD, Ctrip, Baidu...
Ultimately, the selling point is cash flow and dividends.
Hang Seng Tech will most likely continue to underperform global tech assets benefiting from AI,
China's intermediary business model is basically reaching its limit. Elon Musk Premium Zeroing In Progress: DOGE Is Losing Its "Godfather"
As of August 14, 2026, Musk hasn't mentioned DOGE on X for over two months. The last time was June 2, when someone dug up his 2022 old post "If McDonald's accepts Dogecoin, I'll livestream eating a Happy Meal," to which he replied "That's right"—the market's reaction was: no reaction. On that day, DOGE was worth only $0.0993; today, August 14, DOGE is hovering around $0.0696-$0.070, closing at $0.0697 on August 12, with a market cap of about $11.8 billion and a 24-hour trading volume under $400 million. The two resistance levels above at $0.0708 and $0.0729 are tightly holding it down, while $0.0682 is the recent last line of defense.
Looking at another set of data makes it clear that "premium zeroing" is not clickbait: at the start of the year, $DOGE was still at $0.118, down 40% over eight months; down 85% from the 2025 high of $0.4665; and down over 90% from the 2021 "Musk bull market" peak of $0.7316. Meanwhile, the Fear & Greed Index is at 27, indicating market sentiment is at freezing point.
What’s truly intriguing is the diminishing marginal effect curve. In 2020-2021, one word from Musk could pull DOGE up 50% or even double it; by the end of 2025 and early this year, a shout could still raise it 25%; on March 22 this year, when he revived the "DogeFather" persona for nostalgia, DOGE actually dropped 2.6% that day; and the June "That's right" didn’t even cause a ripple. The business of pumping coins by shoutouts has reached its limit.
The core contradiction behind this is that DOGE’s narrative is shifting from "Musk’s personal touch" to "X Money’s real functionality." On March 29, he confirmed on X Space that X Money would start public beta on April 9 and natively support DOGE payments. This is the only time in five years of pumping history that the narrative shifted from "meme" to "schedule." But the problem is, after payment integration lands, DOGE’s price has lost its speculative imagination—coins that become practical tools no longer have a reason for premium. The market now prices not "whether Musk will shout again," but "how many of the 60 million X users actually use DOGE to buy coffee."
For holders, it’s time to wake up: the DogeFather era is over. Going forward, DOGE’s price moves will depend on the overall market beta and real on-chain usage, not someone’s posting frequency. If $0.0682 doesn’t hold, the next support is $0.064; waiting for Musk to save you? The last time he spoke, your losses were probably less than now. SanDisk $SNDK: This is no longer an ordinary rebound.
The moving average structure has been fully restored.
EMA20≈1402, EMA50≈1332, EMA120≈1343, the stock price has risen above all; more importantly, the price is challenging the area above EMA200≈1413. The dense chip area formed by the long-term horizontal consolidation between 1200—1300 is now migrating upward.
This rise has the flavor of a "breakout + volume expansion."
Since starting near 1200, the trading volume has significantly increased, MACD DIF 75, DEA 51, the histogram remains positive, indicating short-term momentum is still bullish. However, RSI6 has reached around 80, RSI12 is close to 79, showing clear short-term overheating.
Therefore, the biggest taboo now is to FOMO chase the price after 1550 directly.
1598 is the first real resistance.
Your chart already shows resistance near 1598. If the price holds above 1598 with volume, the next stage has a chance to push towards 1650—1750; if after hitting 1598 there is a long upper shadow and a drop back below 1455, it is likely to form a "news realization + profit-taking sell-off."
1598 surge → pullback near 1455 → volume contraction → then volume expansion breakout.
If 1455 turns from resistance into support, the quality of this structure will be significantly higher than a direct rally.
Fundamentally, there is indeed strong support for the technicals now. On August 13, SanDisk announced new long-term targets at its investor day, expecting FY2028—2030 revenue to maintain mid-to-high single-digit growth, with an adjusted gross margin target of about 80%; the company is also expanding multi-year agreements and advancing HBF high-bandwidth flash memory for AI inference.
So the core logic of SNDK has gradually shifted from a "storage cycle stock" to a "revaluation of AI infrastructure + high-performance storage."
Will the funds spread to the crypto space?
This is where I find it more interesting.
If the AI infrastructure rally continues to strengthen, I will observe the capital chain of "AI computing power—data—storage—DePIN," rather than chasing a small coin that suddenly surges.
Worth putting into the watchlist:
$TAO — AI computing power/model network direction, a high Beta core target. The focus is not on how much it rises in a day, but whether volume and open interest expand synchronously during the rise.
$RENDER — Decentralized GPU computing power, directly mapped to the AI computing narrative. If the AI mainline regains risk capital, it is usually easier to become a capital recipient than pure concept coins.
$FIL — I value its "data storage infrastructure" attribute more. SNDK essentially trades storage demand; if the market starts mapping the AI data infrastructure industry chain, FIL is a veteran asset worth watching.
$AR — Permanent data storage narrative, logically connected to AI data growth. It is not the strongest in the short term, but if there is a "storage sector catch-up rally," its elasticity may be more obvious.
$AKT — Decentralized GPU/cloud computing direction, a higher-risk AI infrastructure Beta.
My ranking:
For research value:
$SNDK → $TAO → $RENDER → $FIL → $AR → $AKT
For potential catch-up rallies, I would focus more on:
$FIL $AR $AKT
Because SNDK has been strong for several consecutive trading days, the odds of funds continuing to chase the leader are decreasing; the real interest is to observe whether AI funds start rotating to second-tier infrastructure assets.
However, it cannot yet be said that these coins have seen "smart money inflows"—your chart only proves that SNDK itself has a clear volume-price breakout, but cannot prove that crypto funds have migrated synchronously. To confirm "hidden funds," the next step should be to look at spot CVD, contract OI, funding rates, exchange net inflows and outflows, and large address holdings changes.
This is my personal view and does not constitute any advice. Let's talk about my current observation logic, it's not complicated, three steps.
The market is a bit "selective" right now, I don't treat all altcoins the same.
Step one, look at $BTC first.
Bitcoin is currently hovering in the $63K–64K range, and the overall market is fluctuating within this relatively narrow range. If BTC is unstable, don't even talk about what comes next.
Step two, then look at $ETH.
ETH is now near an important area around 1.9K, how it reacts here will tell me if buyers are ready to push altcoins further.
Step three, if the first two conditions are met, then look at rotation.
BTC stabilizes + ETH momentum strengthens, that's when I start paying attention to the list:
$SOL
$HYPE
$BICO
$SNDK
I don't expect them all to rise together. What I focus on is rotation—where the strength appears first and whether it will spread to other targets.
For me, the layout is simple:
BTC stable → ETH strengthens → altcoins become interesting.
Let's see which token will catch the next wave.
$BTC $ETH $SNDK
#交易之声：你的经验值得被听到 SanDisk announced a new AI storage product plan called HBF. According to TrendForce, the first-generation product has completed tape-out, with sample delivery planned for 2027 and mass production in 2028.
It can directly reduce AI inference's dependence on HBM, will increase NAND's valuation, and may lead to a re-pricing of the existing storage sector.
HBM is high-speed memory next to the GPU, fast but expensive and limited in capacity; NAND has large capacity and low cost but is usually placed in more distant SSDs, with read speeds that can't keep up with the GPU.
HBF stacks multiple layers of NAND directly next to the GPU. Its response speed is not as fast as HBM, but it offers larger capacity and lower cost, which can alleviate memory pressure caused by long context and KV Cache, allowing HBM to handle data that requires higher speed.
For AI companies, the same number of GPUs may handle more requests, and the inference cost per token could potentially decrease.
For the storage industry, SanDisk is the most direct beneficiary. The market may no longer view it solely as a traditional NAND cycle stock but will increase valuation for AI memory.
SK Hynix is simultaneously developing both HBM and HBF, meeting the needs for high-speed computing and large capacity.
With HBF entering mass production, NAND stacking, advanced packaging, UCIe interconnect, and testing equipment may also receive new incremental orders.
$SNDK #闪迪投资者日后，长期目标成焦点 攻击者获取的是合法访问权限，链上无可辩别的指纹 量子攻击与传统盗窃最大的不同在于，它在链上看起来完全是"合法"的。攻击者通过 Shor 算法从暴露公钥反推出的私钥，是数学意义上的真实私钥；他们发起的每一笔交易，签名验证通过、脚本执行无误、节点照单全收。 链上观察者看到的只是一笔笔普通的转账，无法用任何技术手段把它和合法交易区分。这种完美的合法盗窃让所有基于【私钥 = 资产所有权】的传统假设同时失效。 当沉睡币开始流动时，没人知道是谁在动。 当中本聪那一百多万枚早期 P2PK 地址的币开始移动时，市场会瞬间会有三种解读： 有人相信"中本聪终于现身了"（创世者还在）; 有人怀疑"私钥泄露了"（币被偷了）; 有人意识到"是量子攻击"（底层密码学失效）。 这三种解读对应完全不同的市场含义，但链上看不出任何区别。第一个发出"权威解读"的人，会定义接下来几天的市场叙事，无论是分析师、政府官员还是自媒体猜测，都会被市场当成"准官方结论"，引发截然不同的买卖决策。 量子攻击会"伪装"成常见的安全事件。 Quantus Network CEO Christopher Smith 指出，早期的量子攻击很可The sequencing problem matters as much as the delay itself. CLARITY has cleared the Senate Banking Committee, but a full vote is expected only in September, while the SEC has postponed discussion of offerings, fundraising exemptions, safe harbors and tokenized securities.
That leaves both legislation and rulemaking moving more slowly than markets anticipated. My read: until one track provides a credible framework, token issuance and tokenized-securities pilots may remain cautious, with regulatory optionality valued over speed.
#CLARITYSECRulesDelayed$668 million unlocking crash vs. US SEC policy benefits, RWA leader ONDO is at the epicenter of a fierce bull-bear battle. On one side is the shadow of a "gradual decline" caused by massive token releases, on the other side is the hardcore support from ecosystem expansion and compliance advancement.
Facing a "cliff-like" unlocking in the next 6 months, early VCs' low-cost chips and foundation withdrawals create real selling pressure. Under a weak market, the rapid expansion of circulating supply easily triggers a sustained gradual decline.
Fundamentals remain solid, TVL steadily grows, strategic investment in Saturn expands credit assets, and with the SEC planning to launch 24/7 trading of tokenized stocks, the RWA sector's long-term narrative remains attractive.
Be cautious of selling pressure realization. Focus on on-chain treasury and VC wallet anomalies, spot trading volume changes, and whether the premium on Korean exchanges falls back.
ONDO is now at a "massive unlocking meeting policy benefits" moment. The amount sounds scary, but whether it crashes depends on whether VCs sell and if the market supports it. Don't blindly guess tops or bottoms in the short term; watch on-chain wallets closely. For long-term positioning, it's safer to wait until the unlocking bearish impact settles and the market stabilizes before entering. $ONDO #CLARITY表决待定，SEC规则未落地 In-depth Analysis of ACO Tokenomics: The Deflation and Distribution Logic Behind the Fixed Total Supply of 1 Billion 📊
To see if a public chain project can last long-term, the token model is crucial. Setting aside concepts, let's directly look at ACO's underlying issuance and distribution design:
💎 Fixed Total Supply and Distribution Mechanism
A fixed total supply of 1 billion ACO tokens with no risk of unlimited inflation.
55% allocated to network-wide ecological mining: the vast majority of tokens are linearly issued through community node construction and full-scenario interactions, ensuring decentralized token distribution.
🔥 Full-Scenario Burn and Deflation
On-chain DEX trading gas fees, instant swap fees, decentralized plaza feature unlocks, and live streaming rewards all include token burn and collection mechanisms.
As ecosystem applications (RWA + social + live streaming) become more active, the token deflation rate will dynamically accelerate, forming a sustainable underlying value support.
No storytelling, just logic. Do you think this issuance and burn mechanism can support a long-term value closed loop?
#Tokenomics #ACO Public Chain #DeFi #Blockchain For most of the past several years, Bitcoin and software stocks moved as if tethered together. Not anymore. The Split, in Numbers Since May 2026, the two have gone in opposite directions. The iShares Expanded Tech-Software Sector ETF (IGV) — a proxy for names like Microsoft, Oracle, and Salesforce — is down only about 1% year-to-date. Bitcoin, over the same window, has fallen roughly 29%. Their 20-day rolling correlation has now turned negative for the first time since May 2024, after years of t$ETH Ethereum abandons Poseidon after 8 years of research: it's not a mainnet issue, but the veteran SHA/BLAKE has made a comeback
Ethereum L1 roadmap abandons Poseidon hash, switching to SHA or BLAKE. The research plan, which took 8 years and involved an eight-figure dollar investment, is directly shelved. This is not due to vulnerabilities in the current mainnet's confidential algorithms, but specifically targets the future leanVM and quantum-resistant L1 plans, aiming for deployment around 2029.
Poseidon was designed specifically for ZK proofs and previously had very high efficiency in ZK. As proof systems evolve, the veteran SHA and BLAKE have caught up in performance within ZK, while the security risks of algebraic hashes like Poseidon have been re-evaluated. The official choice has shifted to mature, well-audited solutions.
Note, this abandonment only applies to the L1 roadmap; zkRollup and other ecosystem projects can still continue using Poseidon.
This development has no impact on ETH's short-term market performance and reflects Ethereum's willingness to abandon huge sunk costs to adjust its technical roadmap. #标普收盘再创新高，8000点预期升温 #CPI与PPI同步降温，加息分歧扩大
To raise rates or not, it's such a tough choice
The biggest challenge the Fed faces now is: how to continue fighting inflation without hurting employment
Hawks:
Core inflation remains above the 2% target, energy prices and the AI investment boom may continue to push costs higher
At the July meeting, three officials already advocated for a rate hike, which indirectly shows the hawks are serious this time, not just talking
Doves:
Employment has weakened, and businesses and consumers are bearing high financing costs
Recent CPI and PPI have cooled down, the Fed can definitely wait for a few more data sets
Hiking now might not precisely suppress inflation but rather add a brake when the economy is slowing
The split among Wall Street's big banks essentially reflects different bets on two paths:
One side believes inflation's stickiness will ultimately force the Fed to hike; the other thinks cooling employment will keep it on hold long-term
For the market, it's like immortals fighting and mortals getting hurt 🤕
Amid the tug of war, uncertainty is also soaring
When data heats up, US Treasury yields and the dollar may rise, growth stocks come under pressure; when data cools, rate cut expectations return
What the Fed really struggles to decide is not a single rate hike, but two kinds of mistakes: hiking too early and crushing employment; hiking too late and letting inflation rebound.Account Position Divergence Radar
Account direction reflects sentiment, while position weight reflects strength. This set specifically identifies where these two do not align.
$LAB has more accounts leaning long, but the top position weights lean short, indicating that the apparent consensus has not yet translated into position scale. Both price and positions are falling together, releasing selling pressure. Which side is exiting cannot be confirmed by this data alone. The next step for the long side is not more accounts, but confirmation of the top position weights.
$DOGE account direction leans long, but top position direction leans short; the side with more people is temporarily not the side with heavier top positions. Price and positions are falling in sync, so this phase is treated as a sell-off. If price rises but top positions continue to lean short, position measurement conflicts are likely during pullbacks.
$XRP account count consistently leans long, but the top position ratio remains below 1, so the numerical advantage has not turned into a top position advantage. The decline has not led to position expansion; first, watch when risk exposure contraction slows. To resolve divergence, the top position ratio needs to rise, not just rely on increasing account numbers.#AMD completes the largest-ever USD bond issuance: raising $4.75 billion
AMD has issued bonds, raising $4.75 billion in one go, the largest bond issuance in the company's history.
What does this have to do with the crypto world?
First, the cost of computing power still needs to be borne. AMD is investing this $4.75 billion into AI infrastructure, which means chip production capacity expansion will not stop. Miners hoping for a short-term drop in computing power costs are unlikely to see it. AI chip players are still desperately expanding production to grab market share, so hardware prices will remain high.
Second, there is an indirect emotional benefit. AMD being able to issue bonds on such a large scale indicates market confidence in the AI sector. As long as leading players keep investing heavily, the risk appetite in the tech sector won't be too poor, and crypto, as a high-beta asset, will get some breathing room.
Third, competition is accelerating. Nvidia is setting up financing platforms, Intel is issuing more shares, and AMD is issuing bonds. Although the paths differ, the goal is to compete for AI chip production capacity and market share. The ammunition consumption in this battle is faster than most people imagine.
Regarding AMD as a coin.
In the short term, the news is positive. Being able to raise $4.75 billion through bonds shows market confidence in AMD's credit. But the stock price may not rise immediately, as financing also means increased debt pressure. Investors will watch whether this money can translate into revenue growth and whether AMD can really take market share from Nvidia.
In the medium term, AMD's AI chip revenue is still growing, with data center business accounting for more than half. As long as this line continues to grow, the fundamentals won't be too bad. But whether the valuation can hold depends on how much market share it can take from Nvidia.
In the long term, whether AMD can narrow the gap with Nvidia determines its valuation ceiling.
What do you think? $BTC $DOS Maybe my account will be banned one day, so I'll give a hardcore analysis of a project in advance. Projects incubated by Binance YZi Labs are rarely C-end; they are mostly B-end. The crypto projects are divided into ABC application ends. A-end refers to real-world enterprise-level applications, such as Ripple's cooperation and services with large banks, and VeChain's early provision of traceability code services for Mercedes-Benz, BMW, and BYD. These are A-end projects, which are currently hard to implement due to whether your technology and enterprises accept them. B-end refers to projects used by public chains or developers, such as the king of B-end, the oracle, and multi-chain projects. Other B-end projects are all about optimization and patching, which are hard to break out. The last type is C-end, directly serving crypto players, such as blockchain games. For example, the running shoes and running-to-earn projects back in the day were explosively popular, as you know. Although this technology is not particularly strong, it has the strongest explosive power. The more players, the greater the token consumption, and naturally, the token price rises sharply. Tokyo's most aggressive corporate Bitcoin buyer just published six months of results, and the picture depends entirely on which line you're reading. The Strong Half Sales for the period ending June 30 came in at ¥4.94 billion, up nearly 134% from a year earlier. Profit from operations climbed even faster — ¥3.33 billion, a gain of roughly 136% — putting the operating margin north of 67%. Most of that came from writing options and other derivative income tied to the firm's Bitcoin holdings, a busWall Street and AI Enter the Crypto Market Simultaneously: Will BTC and ETH Become More Stable?
Currently, the two most important forces in the crypto market come from Wall Street and AI. Actively managed multi-asset products allow traditional institutions to allocate $BTC, $ETH, and other assets just like stocks; intelligent agents begin to research, screen, and even execute operations on behalf of users. Many therefore judge that the market will become more mature and efficient, but maturity does not necessarily mean more stability; sometimes it means funds move faster.
The change brought by Wall Street is that asset selection shifts from belief to models. Fund managers adjust positions based on valuation, momentum, liquidity, and risk; BTC no longer only increases because it is the largest, and ETH cannot automatically receive high weighting just because of its rich ecosystem. Every macro data point, ETF flow, and on-chain indicator can be quickly incorporated into portfolio rules. Market narratives will be quantified, and rotation speed will increase.
The change brought by AI is in information processing and execution speed. Previously, a trader needed hours to read policies, financial reports, and on-chain data; intelligent agents can track numerous sources simultaneously and provide conclusions. Elon Musk pushing Grok toward complex workflows shows that agents are not just generating summaries but moving closer to real execution. As more funds use similar models, market reactions to news may shorten from minutes to seconds.
This is both beneficial and risky for BTC. BTC has the best liquidity and clearest rules, making it the easiest core crypto position in models; but it will also be the asset most easily sold when risks arise. If macro data disappoints, algorithms can reduce positions simultaneously; when risk appetite recovers, funds will quickly return. Institutionalization increases long-term participation but may amplify short-term correlations.
ETH faces a more complex situation. It is both an asset and on-chain financial infrastructure. AI agents may in the future use smart accounts, stablecoins, and contract settlements to create real demand for the ETH ecosystem; meanwhile, fund models will continuously evaluate network fees, staking yields, and competitive landscapes. If on-chain usage grows but does not translate back to ETH value, institutions will similarly reduce weighting. AI creates new markets for ETH but also poses stricter valuation questions.
On the positive side, institutional custody, product transparency, and automated risk control can reduce some original market problems. Funds no longer need to rely on opaque channels, portfolios can set clear limits, and AI can more quickly identify abnormal trades and risk concentrations. Crypto assets will move from isolated accounts into overall wealth management systems, helping to expand the base of long-term holders.
However, new systemic risks will also emerge. If many institutions adopt similar models, the market may act simultaneously on the same signals; if AI reads incorrect data or encounters malicious information, automatic execution will quickly amplify problems; if on-chain and ETF markets feedback into each other, trading session differences may cause price jumps. Efficiency eliminates some frictions but also removes time for calm reflection.
For ordinary traders, future advantages will not come from being faster than machines. More important is understanding what machines cannot easily price: whether policies can persist, whether network effects are real, why users stay, and how assets capture value. Short-term information gaps will shrink, but long-term judgment differences will become more valuable. AI can read all the news but may not know how much the market is willing to pay for trust ten years from now.
Therefore, I do not believe Wall Street and AI will simply make BTC and ETH more stable. They will deepen market foundations, increase capital sources, while making pricing more continuous and rotation faster. Extreme scams may decrease, but macro shocks may transmit faster. A mature market is not without volatility; rather, volatility becomes more structured and harder to explain with a single narrative.
$BTC will continue to compete for a place on the global balance sheet, $ETH will continue to compete for the settlement position in the machine economy and on-chain finance. Wall Street decides how much to allocate, AI decides when to act, and what truly determines long-term value is whether the network continues to be used after all models shut down. $INTC's over $20 billion private placement subscription has pushed risk appetite to a high level. The core conflict lies in the valuation re-evaluation caused by aggressive lock-up in the primary market versus the battle with the massive depreciation of advanced process technology suppressing position tolerance.
Market facts show that $INTC's issuance expanded to $20 billion, attracting over $100 billion in subscriptions, with management personally participating with $12 million. Huge funds are directly locked into AI foundry and advanced process capacity expansion, with primary market funds completing a concentrated lock-up of long-term chips in a short period.
In terms of event-driven ranking, the sharp boost in risk appetite is first, with over a hundred billion in subscriptions proving the market's strong willingness to revalue the computing power manufacturing cycle; next is the change in trading position structure, with chips concentrating in long-term capital easing short-term selling pressure; finally, the transmission factor is the hardware depreciation risk under heavy capital expenditure pressure.
The trigger condition for the upward scenario is the smooth landing of advanced process capacity and rapid confirmation of external orders. When the $20 billion capital expenditure converts into actual output, combined with long-term supply agreements locking orders upstream and downstream in the industry chain, institutional chip tolerance will continue to rise and steadily push up the valuation midpoint. Variables to watch at this time are the pace of foundry order fulfillment and yield ramp-up speed, with a failure signal being order confirmation progress falling short of expectations.
The trigger condition for the downward scenario is the cooling of the subscription frenzy and delayed yield ramp-up. If the $500 billion-level computing power financing triggers market caution about overheating bubbles, the massive depreciation caused by the $20 billion hardware investment will directly suppress profit margins, causing the institutional chips locked up to loosen. Variables to watch then are the risk transmission of the computing power financing cycle and signs of position withdrawal, with a failure signal being the market ignoring depreciation pressure and continuing to inject funds.
When external foundry order confirmation is interrupted, or advanced process yield ramp-up lags behind the capital expenditure accrual cycle, the premium logic established by aggressive lock-up in the primary market will completely fail.
The most important variables to observe in the next 7 days are Intel foundry order actual confirmation announcements and the advanced process production rhythm.
#霍尔木兹通航谈判未果，美伊施压升级 #CLARITY表决待定，SEC规则未落地#Harmony推进链上回滚，铸币漏洞修复已激活
Harmony's situation is quite severe. On August 12, over 3 trillion ONE tokens were directly minted out of thin air at the protocol level. This is no ordinary bug; it directly attacked ONE's supply verification mechanism, essentially a protocol-level inflation.
The attacker exploited the "empty block" vulnerability, completing massive minting through 6 abnormal blocks, with some tokens quickly transferred to exchanges during the price crash. Subsequently, ONE's price plummeted by over 30% at one point.
The official response was relatively swift: the cross-chain bridge was suspended the same day, and all validator nodes were urgently required to upgrade to patch version v2026.1.1. It has now been confirmed that the fix is activated, and an on-chain rollback plan is being advanced, with consensus reached among validators and exchanges. The team tracked 10,288 suspicious transactions involving 409 wallets and has notified exchanges to freeze the related deposits.
The ONE in your wallet may have been diluted by one-third out of thin air within minutes. Although the on-chain rollback can undo the attack, its execution is extremely challenging, and whether it will be successfully implemented remains uncertain. For ONE holders, these three days are likely to be more difficult than the past three years. Very optimistic about Chinese technology; it is only a matter of time before China's tech industry chain completely surpasses Europe and the US.
But the more optimistic you are about China, the more you should short China.
Just like China's photovoltaic industry, which accounts for over 90% of the world's capacity and patents. It can be said that in the photovoltaic tech sector, China is absolutely the leader, truly far ahead.
But what about the stock prices of photovoltaic companies? They keep falling.
Once China's tech industry becomes the world leader, the next steps are capacity expansion, price reduction, and ultimately lower profits, with everyone losing money and no one making a profit, leading to stock price declines.
The same applies to China's new energy vehicles. Domestic new energy cars have forced Tesla to keep lowering prices and have pressured BBA to the point of sales fractures or even withdrawal from the Chinese market, but they themselves are not profitable, with low or continuous losses, and stock prices keep falling.
Whenever there is a breakthrough in Chinese technology, one company becomes ten, ten becomes a hundred... then capacity increases and profits decline. So the more you believe in Chinese tech breakthroughs, the more you need to be bearish on Chinese tech stocks.
Believe in the US, invest regularly in the Nasdaq; believe in China, short China.$SNDK It all started when I saw users in the community trading with extremely high leverage and sharing their own disastrous returns. I have no objections to any trading method by any user; everyone should be responsible for their own account. However, try not to bring high-leverage trading strategies from cryptocurrencies into traditional finance. The market makers of altcoins have no conscience, but SanDisk is not the only one who has no conscience. Entire Crypto Market 24h Trading Volume Spot only: about $900-18 billion / 24h However, SanDisk SNDK (US stock stock) average daily turnover in ordinary trading days: ~$15-21 billion On trading volume days (8-13) with high volume and sharp rally: $32.7-33.1 billion What needs to be understood is future profits and changes in industry structure, not just damn candlesticks. The Nasdaq 100 has been rising for a long time, the S&P 500 has been rising for a long time, yet many retail investors still lose money. Why is that? Do retail investors not know that indexes can provide stable returns? Not sure about buying on dips? That's not the case. Everyone thinks they can achieve excess returns through timing and stock selection, and some even neglect position management because of this. This brings us to a very important aspect of traditional financial markets: implied volatility, or IV. If a trader has already started trading 50x SNDK perpetual but hasn't even looked at the current IV offered by the SNDK options market, then their research on the broken line is indeed insufficient. Because the options market has directly indicated what level of future volatility the market is willing to payOKB rallies against the trend, mainly due to the resonance between fundamental restructuring and scarcity narrative; meanwhile, BTC/ETH are suppressed by macro factors and capital outflows, leading to divergence due to different logics.
Why OKB can "rally violently"
- Supply-side revolution: A one-time burn of 65.26 million OKB, permanently locking the total supply at 21 million, aligning with Bitcoin's scarcity narrative and relieving inflation pressure
- Ecosystem value migration: OKB becomes the sole native Gas token of X Layer (zkEVM-based Layer 2), with ecosystem growth directly boosting demand
- Traffic entry binding: OKX Web3 wallet designates OKB as a "must-have asset," providing a foundational demand from tens of millions of users
- New valuation model: Value anchor shifts from exchange profits to on-chain ecosystem; holders can earn an annualized "shadow interest" of about 8.5% through new token launches and financial products
- Compliance and cooperation: Strategic partnership and investment from ICE, the parent company of the New York Stock Exchange, enhancing long-term development confidence
Why BTC/ETH "decline"
- Macro suppression: High interest rate environment weakens risk appetite; Bitcoin often negatively correlates with real interest rates
- Capital outflows: Continuous redemptions from the US spot Bitcoin ETF, withdrawing funds from the mainstream coin market
- Deleveraging: High leverage positions forcibly liquidated during volatility, creating selling pressure and negative feedback on prices
Market-level "seesaw"
- Capital rotation: When mainstream coins are weak, some funds shift to platform tokens with improving fundamentals, creating a phased risk-hedging or speculative "seesaw" effect. $OKB SanDisk投资者日释放关键信号：管理层正试图回答市场长期回避的问题——当前高利润水平是否已处于周期顶部。该公司给出的长期模型显示，FY2028至FY2030年间，营收有望维持中高个位数至两位数增长，Non-GAAP毛利率维持在80%左右，营业利润率约75%，调整后自由现金流比率约50%。这一指引的重要性远超高管口中“AI需求强劲”的表述。 上季度SanDisk实际毛利率达到84.6%，市场真正担忧的是，一旦NAND进入下行周期，这一利润率水平是否会重演历史性回落。管理层给出的答复是：此次周期可能不同，核心支撑在于NBM机制。目前SanDisk已与8家客户签署NBM长期合同，覆盖FY2027约50%的位元产出，FY2028覆盖比例接近三分之二。合同并非简单的购买意向，而是包含锁定量、合约框架、最低财务担保及结构化定价的约束性安排。 这一商业模式旨在改写NAND周期逻辑：传统链条为“涨价→扩产→供过于求→跌价→利润崩溃”；新框架则试图以长期合同锁定需求、控制产能扩张、提升营收能见度，从而降低周期波动。若该模型被市场验证，SNDK估值逻辑的最大变化或将不再是EPS增长幅度，而是市场可能不Last night, the brightest star in the US stock market was SanDisk SNDK.
During an investor day, the stock price surged nearly 20% intraday, also lifting the entire storage sector: Western Digital WDC rose nearly 10%, Micron MU rose nearly 7%, and Seagate STX rose nearly 6%.
The market's excitement is simple: SanDisk presented an almost "unbelievable" long-term goal—FY2028 to FY2030 revenue to maintain mid-to-high double-digit growth, non-GAAP gross margin around 80%, operating margin around 75%, free cash flow margin around 50%, and remaining cash will be 100% returned to shareholders.
More importantly, this time it’s not just empty promises. SanDisk has already signed long-term agreements with 8 major customers, covering about 50% of shipments in FY2027 and about two-thirds in FY2028. The most fatal problem of traditional NAND is price volatility, but long-term volume locking and structured pricing are essentially attempts to transform a cyclical business into a "long-term contract business."
The biggest bet behind this is AI inference: the more tokens, the larger the KV Cache, and data centers will need high-capacity, low-power flash memory. SanDisk expects the enterprise flash market to reach 1.2ZB by 2030 and is betting on a new AI storage tier with HBF and high-density QLC.
Of course, an 80% gross margin sounds less like traditional NAND and more like Nvidia. Tonight, the market chooses to believe first; next, it depends on whether SanDisk can turn this "epic PPT" into cash flow. $SNDK Why do people think Bitcoin will still rise this year?
It's nothing more than a modern retelling of the fable of carving a mark on a boat to find a sword.
Some people look at historical data and find: Bitcoin halves roughly every four years, and coincidentally, there has been a spectacular rally after each halving.
Oh? So based on the current situation, the next halving will be around 2028, and Bitcoin should take off around 2029.
Thus, the 2029 big rally theory was born.
How to put it, it has some theoretical basis, but not much.
It's mostly superstition and shouldn't be taken as investment advice—just for fun.
When someone feels utterly powerless about the current situation, they are easily "forced" to find some patterns.
My cousin experienced three major breakups in 2016, 2018, and 2022. She never examined her taste in partners nor seriously reflected on her own issues.
Her final conclusion was: even years are bad for dating. The world's clown watches Gotham; China's clown watches my cousin.
It's the same in financial markets. As long as a chart looks neat enough, with several peaks connected by a line, many people get the illusion that they've uncovered a secret.
But the question is, were past Bitcoin rallies really just because of halving? Not necessarily.
Bitcoin in 2017 coincided with exchange expansion and the digital currency concept going mainstream.
The bull market after 2020 was driven by global liquidity injections, low interest rates, and extremely loose monetary policy.
Later, institutional money entering and spot ETFs opening channels brought another set of dynamics. Halving happens every time, but what actually pushes the price up each time is never exactly the same.
Halving only means fewer coins are issued; it doesn't magically create money to buy coins.
Miners selling less daily reduces selling pressure, of course, but for Bitcoin to rise from one price to another, someone still has to throw real money in.
If after 2028 the world re-enters a loose monetary cycle, with ample USD liquidity and continuous institutional inflows, then a big rally in 2029 is certainly not surprising.
But if interest rates remain high, capital tightens, regulations tighten, and market risk appetite is low, then halving alone won't save the day.
Many people confuse cause and effect.
They think halving creates the bull market.
In reality, halving provides an easy-to-tell, easy-to-remember, and easy-to-spread story, while what truly ignites the rally are liquidity, sentiment, and new capital.
Moreover, today's Bitcoin is no longer the speculative asset in a small pond from ten years ago. Back then, Bitcoin's market cap was small; a few tens or hundreds of billions of dollars moving could flip the price upside down.
Now the market cap is much larger. To replicate early gains of tens or hundreds of times, the required capital isn't just a bit more, but a lot more.
So Bitcoin's cycles may still exist, but their power will likely weaken.
It may still rise and hit new highs, but it might not suddenly lift everyone to the sky in a single year like before.
There's an even more troublesome issue.
Once a pattern is known by everyone, it often no longer happens the same way.
If everyone agrees 2029 will be a bull market, truly smart money won't wait until 2029 to enter; they might start positioning in 2027 and begin trading the halving expectation in 2028.
By the time ordinary people finally reach 2029, full of confidence to witness history repeat, the early movers might already be selling.
Financial markets are never like a bus.
Just because the schedule says the bus leaves at 9 o'clock doesn't mean it will definitely pick you up exactly at 9.
So, saying Bitcoin has a chance to rise in 2029 is fine.
But saying it will definitely surge in 2029, or treating it as a certainty, is somewhat like carving a mark on a boat to find a sword.
Halving is an objective fact; it does reduce new supply. But supply reduction is only one condition for a rally, not the whole story.
What truly determines whether Bitcoin can rise is whether there is enough money in the market, willingness to take risks, and when those who positioned early decide to hand over their chips to later buyers.
Bitcoin's greatest strength has never been just its limited quantity, but that it always finds a reason in different eras to convince people it will keep rising.
In the past, it was halving; later, inflation; then institutional entry. By 2029, there will definitely be a new narrative.
Whether the price will rise as people imagine is another matter. The only variable here is that even a consensus among fools is still a consensus.
When you bind enough fools to reach agreement, it really can create an unstoppable momentum. $BTC $SNDK #加密估值转向收入，BTC如何定价？ #Strategy再卖1690枚BTC，企业财库出现分化 Goldman Sachs bought NEOS, and BTCI also entered the big institutional shelf together. All eyes on me: the acquired party is the fund manager, not the Bitcoin network, so don't directly factor the acquisition price into the coin price.
BTCI uses Bitcoin ETP plus selling call options to exchange for monthly distributions. When the market surges rapidly, options may cut off some upside; when it falls, the premium is not a bulletproof vest either.
I will compare BTCI's net asset value total return with the spot Bitcoin index, then look at the premium or discount. If the gap widens, it indicates that the yield packaging has a cost.
I react a bit slowly, but the advantage is that I don't applaud the first candlestick. First, let's see if the post-acquisition fees and option coverage ratio will change.
This article is for informational and educational purposes only and does not constitute any investment advice. Digital asset prices fluctuate greatly; please make independent judgments and pay attention to risks. #$BTC On August 12th, the daily transaction volume of Ethereum Layer 2 reached 12.42 million transactions, setting a new historical high.
Base alone accounted for 3.6 million transactions, surpassing Arbitrum since early July to become the largest L2 by transaction volume.
But what about the mainnet? It only processes about 20 user transactions per second, while L2 handles 1,270. On-chain analytics firm Tanaka made an estimate—Ethereum's application layer generated $1.79 billion in fees in Q2, with the mainnet capturing only $88.4 million, less than 5%. The busier the L2, the less fee ETH collects.
Of the $1.79 billion ecosystem revenue, ETH only gets less than 5%, with the rest all taken by L2.
ETH has dropped about 32% this year, while BTC only fell 11%. That's where the price difference comes from—BTC captures its own value, while ETH's value is diverted by L2.
$ETH At the beginning of August, BlackRock did something on Ethereum by tokenizing a €31.1 billion European money market fund.
This is not a trial; they moved the entire fund onto the Chain.
BlackRock's BUIDL fund has been running on Ethereum for 7 months, reaching a scale of $700 million.
Franklin Templeton is doing the same.
The entire RWA market reached $38 billion at the beginning of August, with Ethereum holding $17.3 billion, a 45% to 46% share. Year-over-year growth is 315%, exceeding the combined total of BNB Chain and Solana.
Institutions choose Ethereum not because it is cheap—on the contrary, mainnet gas fees are much higher than L2. They choose the settlement layer's trustworthiness. BlackRock would not place a $30 billion fund on a Chain that might fork or have governance chaos. After institutions enter, these assets won't be frequently traded; they will settle on the mainnet to form long-term liquidity. L2s compete for retail trading volume, Ethereum competes for institutional balance sheets. Which path is more valuable is still uncertain. $ETH At the beginning of August, there was a highly controversial proposal—EIP-8363.
The core logic is: when the ETH staking rate reaches 50%, all new issuance rewards for validators will be burned, and staking yields will drop to zero.
SharpLink's CEO Joseph Chalom directly opposed it, saying this would undermine ETH's core advantage over BTC—native yield. Aave founder Stani Kulechov also spoke out against it. At the core developers' meeting on August 6, the proposal was shelved.
The proposal hasn't moved forward for now, but this issue will have to be faced sooner or later. The ETH staking rate is already at 34%, increasing at about 1% per month. If EIP-8363 is eventually passed, the narrative of ETH as an "interest-bearing asset" will be completely rewritten—the staking yield will drop to zero, removing one reason for institutions to allocate ETH. If it doesn't pass, the staking rate will continue to rise, validators will increase, security costs will go up, and ETH's supply inflation pressure will persist. Both sides have problems, and both need solutions. $ETH Recently, as inflation data softened, the market immediately started going long on growth.
The logic is very standard: money becomes cheaper, discount rates drop, and future cash flows become more valuable.
But I would ask one more question.
Money has become cheaper, but have electricity, advanced packaging, high-bandwidth memory, and top engineers increased?
No.
These physical bottlenecks typically take 2 to 5 years to ease.
Interest rates can turn in a few months, but fabs cannot.
//
So where does the easy money flow?
It doesn’t flow to increasing supply, but to competing for existing scarce resources.
The result is: asset prices rise first, but real capacity doesn’t keep up.
Everyone thinks they are solving the bottleneck, but in reality, they are collectively making the bottleneck more expensive.
//
I no longer take "interest rate decline → long growth" as a default action.
I first ask myself: what is the tightest physical bottleneck right now? How long will it take to ease?
Is the easy money shortening this time, or extending the competition for it?
Interest rates can change quickly, scarcity cannot.
When these two are mismatched, the first thing exposed is not the growth slope, but the gap between asset prices and real supply. #USStockMarket Here's some data I just saw today: the total number of Ethereum developers has exceeded 1 million.
The volume of smart contract deployments from August 5th to 7th was 50% higher than the three-month average.
The ETH balance on exchanges has dropped to its lowest level since 2016, with about $25.6 million flowing out weekly.
Developers are building, exchanges are becoming emptier, and staking rates are rising. All three trends point to the same conclusion—ETH is transitioning from a "trading asset" to a "productive asset." Gas fees have dropped below 10, cooling on-chain speculative demand, but developers are deploying contracts, institutions are moving RWA, and stakers are locking up tokens. When ETH reached 4800 in 2021, none of these three things were happening. The fundamental logic of ETH has changed a lot; the price just hasn't caught up with this change yet.
$ETH A major coin price drop is not a survival issue but a capital allocation issue. Isn't there a time lag between the market's fearful reaction and the actual capital outflow? Recently, many views express fear over the price correction of large coins, but in reality, this decline is just profit-taking supply that pressed down the price ceiling to realize previous gains. As mentioned in the original text, a price drop does not necessarily mean a value drop, and the actual selling pressure on-chain is concentrated among short-term holders and leverage liquidation volumes. Looking at the market structure, large coins are still in a phase where institutional funds and passive allocation demand support the downside, a pattern repeated since 2020. However, this cycle differs from the past. The phase where simply holding the price, like in 2017 or after March 12, 2020, generated profits is over. The current market has shifted to a structure where real demand capital enters after verifying the project's actual usability and liquidity depth, and simply holding coins no longer guarantees excess returns On August 13, the net inflow was $6.7169 million, with the Grayscale Ethereum Mini Trust ETF accounting for $6.4748 million.
Although the volume is not large, the direction is shifting.
At the same time, the Bitcoin spot ETF saw a net outflow of $131 million.
Money is moving from Bitcoin to Ethereum, and this trend has been ongoing for some time.
In August, Bitcoin ETFs received over a billion dollars, but the price barely moved because miners and retail investors were selling simultaneously. The situation with Ethereum is the opposite—ETFs are continuously making small regular investments, miners are not exerting large selling pressure, and the staking rate has reached over 34%, meaning more than one-third of the supply is locked in staking contracts and cannot be withdrawn. Whether Ethereum at 1,888 is cheap or expensive depends on the time frame you consider.
After bottoming near 1,863, Ethereum is recovering, with moving averages trending upward to provide support, indicating a consolidation phase after a pullback. Short-term support is between 1,873-1,875, with strong support at 1,862. Resistance lies at 1,900 and 1,970. If volume increases and it breaks above 1,900, bears might be forced into a round of losses. According to Coinglass data, if $ETH breaks above 1,975, the cumulative short liquidation intensity on major platforms will reach $788 million. Conversely, if it falls below 1,791, the long liquidation intensity will reach $697 million. The ETH supply ratio on exchanges has dropped to 0.129, the lowest point since 2016.
More and more coins are being locked up, and fewer are available for sale.
But the price is still hovering around 1,888, demand hasn't caught up with the shrinking supply yet.
Someone is doing one thing—withdrawal, staking, locking. A whale moved 37,000 ETH out of Gemini, then deposited it in batches into the beacon chain for staking. No selling, no swapping, no operations, just locking the position. Retail investors are watching, big players are locking up. Same price, two completely different operational logics.
Vitalik's version of the Lean Ethereum roadmap is still progressing, with quantum resistance, privacy, and STARK verification all scheduled. But on the macro level, PPI year-over-year at 4.7% is slightly below expectations, inflation is easing. The urgency for Fed rate hikes is decreasing, which is good for risk assets, but Bitcoin and the US stock market diverged yesterday; the stock market hit new highs, while Bitcoin is still stuck around 63,000. Ethereum also remains stuck. Macro logic and capital flows are conflicting, the direction is not yet clear.
The 2026 market is a completely different world from the 2021 DeFi bull market. Back then, coin prices rose, gas fees rose, and market sentiment rose—all three lines resonated in the same direction. Now, coin prices are sideways, gas fees have dropped below 10, staking rates are rising, and the three lines are moving independently. The structure is changing, but the direction of change has not yet been reflected in the price. $ETH This morning, the entire network liquidated 238 million, with long positions liquidated at 131 million and short positions at 108 million.
Bitcoin long positions liquidated 24.99 million, short positions liquidated 8.65 million. Bulls took three times more hits than bears.
Those who were bullish yesterday probably won’t feel too comfortable today.
At the 1,900 level, short sellers didn’t gain much advantage either; the price is still hovering around 1,888.
Spot ETF inflows have plummeted by over 80% since mid-July. The continuous net inflow momentum seen in July has clearly slowed down in August. Demand is cooling off, and prices are following suit. The US stock market is hitting new highs, but Bitcoin isn’t following. The macro environment is improving, yet Bitcoin isn’t responding. The market is waiting for something that can truly ignite sentiment—possibly progress on the CLARITY Act or the SEC’s Reg Crypto rules coming into effect.
Before that happens, the 1,800-2,000 range will most likely continue to consolidate. Every time it seems like it’s about to break out, it pulls back. Every time it looks like it’s about to crash, it stabilizes. Long or short at this level is uncomfortable, but often the trend only truly chooses a direction when most people feel uneasy $BTC $ETH #FinancialReportObserver: AI Infrastructure Earnings Reports Take the Stage Everyone, the AI infrastructure earnings reports so far show impressive numbers, but the market's attitude has clearly shifted.
Lumentum's revenue grew 109% year-over-year, with next quarter guidance between 1.225 billion and 1.275 billion. Coherent's revenue increased 34% to 2.05 billion, with guidance exceeding expectations. Cisco's Q4 revenue was 17.3 billion, up 18%, with full-year AI infrastructure orders reaching 9.3 billion. Applied Materials' Q3 revenue was 9.12 billion, up 25%, EPS 3.50, and guidance also above consensus.
The numbers are solid, but after earnings from Coherent, Cisco, and Applied Materials, their stock prices have come under pressure. The market is no longer satisfied with just "growth"; it is focusing on three more detailed factors.
First, can profit margins be maintained? Revenue is growing fast, but if costs rise faster, squeezing profit margins, the market will reprice accordingly.
Second, capital expenditure efficiency. Investment is expanding, but how much revenue is generated per dollar invested is a metric that influences valuation more than revenue growth.
Third, order visibility. Previously, the market accepted order growth at face value; now it demands to see whether orders can sustainably convert into profits.
The growth story of AI infrastructure continues, but the market has shifted from "pricing based on expectations" to "pricing based on efficiency." The recent earnings-related stock price pressure is not because the industry is failing, but because the pricing logic has changed.
Folks, the AI infrastructure sector is solid long-term, but the short-term valuation re-rating is not over yet. At this point, let's wait for the market to fully digest this round of earnings. What do you think about the future direction of AI infrastructure? Let's discuss in the comments. Have a great weekend. $SNDK $ETH $BTC [Market Analysis]
Converging consolidation patterns at both large and small levels, whether upward or downward, offer good potential space.
I have been bullish recently, or rather, bullish in the mid-term.
Since it is still a bear market, I am only looking for a rebound. Although the bear market only has the last three months left, the potential decline in these three months should not be underestimated.
Back to the current situation:
Bullish.
The reason is that CPI, PPI, and employment conditions do not allow for rate hikes (which are unlikely anyway, as US Treasury yields have damaged the global economy). Therefore, the inflation expectations driven by rising oil prices in recent months are cooling down, leading to a rebound in the US stock market.
How far can the bullish trend go? Approximately to the resistance level of 67.5k-68k.
This is firstly the starting point of the last round of one-sided decline, with huge trapped positions exerting selling pressure.
Additionally, although a mid-term bullish view is possible in the short term, there is also strong potential liquidity risk. The most representative "gray rhino" is the future fact of the Bank of Japan's explicit rate hike. Japan's rate hike is almost certain; the interest rate differential between the US and Japan will cause the yen to continue depreciating, and not hiking rates only treats the symptoms, not the root cause.
This transmission chain will hinder the US's strong dollar national policy.
In summary, yen carry trades will be forced to reduce positions under the expectation of a yen rate hike, which will exacerbate liquidity squeezes.
What we are waiting for now is a reliable high point to short, and this high point must be brought by "rate cut expectation trades," which are caused by "cooling US inflation and a complete drop in energy prices."
With rate cut expectations, all risk assets including cryptocurrencies will come under downward pressure, completing the final round of leverage clearing.
This is roughly around November to February next year, since there will be no rate changes in September.
Therefore, the rate cut expectation trades will most likely occur between September and November, and we will probably see a suitable high point to short.
So, in August, I am more inclined to accumulate long positions, even if we might drop to the 61.8-60.6k range.
#CPI与PPI同步降温，加息分歧扩大 DeepSeek announced two things today: the official launch of DeepSeek-V4-Pro and the open-source developer preview of DeepSeek Harness v0.1.
The core logic of Harness is a formula: "Model + Harness = Agent." This is not a new model, but a programming intelligent agent framework that enables models to autonomously complete multi-step tasks—directly competing with Anthropic's Claude Code. Licensed under MIT, the code is open source on GitHub.
From the team's founding to the developer preview, DeepSeek took about five months, which is extremely fast even within China's tech circle. The team is led by Cui Tianyi, who joined DeepSeek this March from the renowned quantitative firm Jane Street.
Why this is an important structural change for the AI industry: models are becoming easier to replace as interfaces standardize; however, the "harness layer" that controls how agents reason, call tools, edit code, and run continuously across tasks is hard to replace. DeepSeek is competing for this layer.
This is the second time this year that a Chinese AI product has directly impacted the AI programming tools market—DeepSeek R1's release in January caused chip stocks to lose $600 billion in one day, and in July, Kimi K3's open-source weights triggered another sell-off in chip stocks. The logic behind Harness is different this time: it's not about "my model is cheaper," but "I control your workflow."
Regarding BTC: intensified competition in the AI tools layer → chip stocks' short-term sentiment may be further suppressed → risk assets under pressure.
$BTC $AAPL Recently, AAPL has been continuously falling from the high of $344.57 at the end of July, with the lowest point retesting around 300, a maximum drawdown close to 13%. As of the close on August 13, the stock price returned to $305.26. Although it has climbed back above the 5-day, 10-day, and 20-day moving averages, it is still below the 50-day, 100-day, and 200-day moving averages, so it currently looks more like a correction after a decline, and it cannot yet be confirmed that a new upward trend has started.
I opened a long position at 305, currently basically near the cost. This position is not bad because it is close to the 300–302 support area, but discipline is necessary.
✔ If 300–302 holds, I will continue to expect a rebound
✔ 307–310 is the first resistance; if volume supports a stable break above 310, then look toward 315–319
✔ If it repeatedly fails to break through 307–310, I will consider reducing my position a bit first
✔ If the daily chart effectively breaks below 300, it means the rebound logic fails, and I will not blindly add to my long position
Apple’s fundamentals are not bad; the latest quarterly revenue was $109.4 billion, a 16% year-over-year increase, and earnings per share grew 29% year-over-year. The real issue now is not whether the company can make money, but that the valuation is still relatively high, and the market’s expectations for AI, new products, and subsequent growth are also increasing.
My judgment is that 305 can be used as a defensive long position level, but it is not a blind buy. If 300 holds, there is room for a rebound; if 300 does not hold, downside risk will reopen. $SNDK is currently not a good position to open a new short; it belongs to the "right logic, wrong timing" early left-side zone. The short logic is not broken, but the mid-August position is an entangled zone of "already dropped once + cycle not falsified + buyback support," so naked shorting has a low success rate.
The current trend's bullish pattern has not been broken; as long as there is a pullback, it is a buying opportunity, with an upside target of 1630. You don't necessarily need to catch the exact bottom. You need a strategy you can stick with. Imagine allocating $100 every month since 2022. Historical performance: 🥇 $TRX +195% 🥈 $BTC +54.6% 🥉 $XRP +51.2% $SOL +43.3% $ETH -12.5% $ADA -53.3% The difference between winners and laggards is impossible to ignore. DCA helps with consistency. Research helps with selection. Risk management keeps the strategy sustainable. Past performance is not a guarantee of what's next. What's your long-term pick$2Z real demand is far from proven yet
Low latency is indeed valuable for high-frequency trading, MEV, and some validators, but most ordinary validators and users have very limited willingness to pay for "a few milliseconds faster."
If network usage (real paid traffic) does not keep up, token demand will remain weak for a long time.
2. Token design determines selling pressure
Contributors receive 2Z rewards by providing bandwidth and hardware.
This means that as long as the network is running, new tokens will continuously be released into the market.
In the early stages, when user base and payment scale are not large enough, this design naturally leans toward supply-side pressure.
3. Current price level is not cheap
It has dropped significantly from 0.09 to around 0.05, but compared to the actual network revenue and locked value it generates, the valuation is still high.
The rebound is more of a technical correction after overselling rather than a fundamental-driven reversal.
4. Common issues in the DePIN sector
Many DePIN projects initially push prices up through narratives and airdrop hype, then enter a long "prove yourself" phase.
Only a few survive; most will slowly decline for a long time after the hype fades.
2Z has not yet emerged from this phase.
Charge forward fiercely and win a chicken dinner In-depth Analysis of ACO Tokenomics: The Deflation and Distribution Logic Behind the Fixed Total Supply of 1 Billion 📊
To see if a public chain project can last long-term, the token model is crucial. Setting aside concepts, let's directly look at ACO's underlying issuance and distribution design:
💎 Fixed Total Supply and Distribution Mechanism
A permanent fixed total supply of 1 billion ACO tokens, with no risk of unlimited inflation.
55% allocated to network-wide ecological mining: the vast majority of tokens are linearly issued through community node construction and full-scenario interactions, ensuring decentralized token distribution.
🔥 Full-Scenario Burn and Deflation
On-chain DEX transaction gas, instant swap fees, decentralized plaza feature unlocks, and live streaming rewards all include token burn and collection mechanisms.
As ecosystem applications (RWA + social + live streaming) become more active, the token deflation rate will dynamically accelerate, forming a sustainable underlying value support.
No storytelling, just logic. Do you think this issuance and burn mechanism can support a long-term value closed loop?
#Tokenomics #ACO Public Chain #DeFi #Blockchain First, technically, $SPCX surged too much yesterday, so today's pullback is part of a cleanup. Secondly, the previously supportive bullish news has also been diluted.
For example, Musk initially said Grok4.6 was very strong, but shortly after, DeepseekV4PRO was released, kicking it off before it even got warm. This directly impacts the expected revenue from computing power rental.
For example, the internal speech released on August 11 claiming AI accounts for 99% of SpaceX's value has been highly controversial, with some considering it a distraction. This directly hits investors attracted by Starship, Starlink, and defense contracts.
Also, the so-called Wall Street report claiming 300 billion in revenue is full of holes meant to hype the stock price. First, completing 10GW computing power by year-end is basically impossible. Second, it linearly extrapolates the premium during the current computing power shortage.
Moreover, cross-industry hoarding of computing power actually proves Grok itself cannot absorb the computing power. The so-called rapid construction of computing centers using second-hand Chinese power equipment will inevitably face many difficulties in implementation.
Market outlook:
Holding 139–140: can still be defined as strong turnover after a breakout, as this is the original resistance zone from August 10–11.
Recovering 143–145: indicates today's main action is profit-taking cleanup.
Closing below 138.7–139: the gains from yesterday's AI PR combo are basically given back, and the next likely test is the 135 issue price.
If 135 also fails: return to the 130–133 range to find support again.
Only by retaking 146.15: can it be considered qualified to attack 150 again. An address is gradually transferring old BTC out on-chain in batches. These addresses did not sell when BTC was at $120,000 in December 2021, and now they are starting to move.
Early-stage old money at the bottom is beginning to loosen up, which is a signal of chip turnover.
However, if the volume of transfers out continues to increase over the next few weeks, a reassessment will be necessary.
The 63,000 level has been repeatedly tested from the end of July until now, fluctuating within just a few thousand dollars.
The inflow of BTC to exchanges has dropped to the lowest in nearly three months; no one is depositing coins to sell. Sellers are contracting, which is more noteworthy than buyers hesitating.
Price consolidation usually means both buyers and sellers are inactive. But if sellers stop selling first and buyers make a slight move, the price could bounce up.
FTX compensation is still ongoing; on August 12, 2,300 BTC worth $146 million was transferred to exchanges. Weekly volumes of 2,000-3,000 BTC are not large and the pace is very steady. Spot trading volume has shrunk to the lowest level since 2019. Back then, after volume shrank to this level, there was a 10x rally. History doesn’t simply repeat, but extreme contraction itself is a signal.
On the miner side, hash rate has dropped from a peak of 1,150 to 886, a 23% decrease. Core Scientific has already started converting mining farms into AI data centers; miners are transitioning. Selling pressure is continuously consumed but not exhausted. Long-term holders are starting to lose money. In 2015, 2018, and 2022, every time long-term holders lost money, the market was near bottom regions.
If 63,000 breaks, 62,000 and 61,000 are supports.
My strategy?
Hold spot, no leverage, no new positions. Wait for the CLARITY Act or SEC’s Reg Crypto results before making moves. Watch more, act less until direction emerges.
Bottom signals are triggering one by one, but triggering doesn’t mean an immediate reversal.
$BTC Before the long-term financial model was released, SNDK hovered between 1300-1380 for several days.
After the long-term model was published, a single bullish candle pushed it directly above 1500.
That day's breakout wasn't an instant spike; it was gradually pushed up over nearly an hour.
Some investors slowly bought after confirming the fundamentals, not chasing the rally. SanDisk's revenue is expected to maintain "mid-to-high double-digit" growth from 2028 to 2030, with a gross margin around 80% and an operating profit margin of 75%. Institutions focus on these fundamentals, while retail investors look at the candlestick charts.
Out of 23 analysts, 21 gave buy ratings, with a target price around 1,800, about 15% higher than the current price. The market is re-pricing SanDisk from a NAND cycle stock to an AI infrastructure company with long-term visible revenue.
The NBM protocol has signed 8 clients, with a minimum contract revenue of 93.9 billion, and over half of the capacity for fiscal year 2027 is already locked in. Tokenized stocks trade 24/7, effectively bridging US stock trading hours with the crypto market.
SNDK's liquidity is much worse than the underlying stock; the stock's daily average volume is tens of billions, while SNDK trades 50 million a day. When trading this asset, position control is more important than directional judgment. Don't use the stock's liquidity to trap yourself in SNDK.
I'm waiting for a pullback; 1450-1480 is the first support, and if broken, look at 1400-1420. Resistance above is 1600-1650. As long as the underlying stock's long-term logic remains unchanged, SNDK's trend direction won't independently reverse. $BTC CryptoQuant data shows that BTC inflows to exchanges have dropped to the lowest level in nearly three months. No one is depositing coins to sell; sellers are actively contracting. There are few sellers left in the market, but buyers haven't entered either, so the price remains sideways.
The batch of GBTC addresses from Grayscale has started transferring out old BTC in batches. These addresses didn't sell even at the previous bull peak of $120,000, but now they're moving. Early-stage old money at the bottom is starting to loosen up, which is usually a signal of chip turnover, but if the outflow volume continues to increase over the next few weeks, a reassessment will be necessary.
FTX compensation is ongoing, with 2,000-3,000 BTC flowing into the market weekly at a steady pace. This selling pressure isn't large but happens daily. Spot trading volume has shrunk to the lowest level since 2019. After shrinking to this level in 2019, BTC had a rally; history won't simply repeat, but extreme contraction itself is a signal.
On the miner side, Core Scientific and TeraWulf have started converting mining farms into AI data centers. Hashrate dropped from a peak of 1,150 to 886, a 23% decrease. Miners are transitioning to AI, selling pressure is continuously consumed but not yet finished.
Long-term holders have started losing money. In 2015, 2018, and 2022, every time long-term holders lost money, the market was near the bottom area. Bottom signals are triggering one by one, but triggering doesn't mean an immediate reversal.
63,000 has broken; below are supports at 62,000 and 61,000. At this level, neither bullish nor bearish bias is held; I only know one thing—the sellers are much weaker than a month ago. Hold spot, no leverage, no new positions. Before the direction emerges, watch more and move less. The longer the grind, the stronger the breakout will be. $BTC $INTC completed a $20 billion private placement expansion and attracted over $100 billion in subscriptions, with massive funds flowing directly into AI foundry manufacturing. The aggressive risk appetite in the primary market quickly locked in long-term chips, and institutions are focusing on re-evaluating the capital expenditure cycle of advanced process technology. If subsequent capacity ramp-up drives external order volume, the valuation midpoint is expected to steadily rise, but massive depreciation may suppress position tolerance. When subscription enthusiasm fades and yield ramp-up lags, the logic will fail; the next step is to observe foundry order confirmations and actual production schedules.
#加密估值转向收入，BTC如何定价？ #标普收盘再创新高，8000点预期升温 #CLARITY表决待定，SEC规则未落地Tokenized shares must go through three stages: issuance, distribution, and clearing before reaching users, but there is no fixed rule for who does which step. The real competition happens between these three stages. According to RWA.xyz data, as of August 12, the total size of the tokenized stock market was about $2.5 billion. Ondo leads with $866 million (34.6% market share), bStocks surpasses to second place with $614 million (24.53%) less than two months after launch, and xStocks slips to third place with $560 million (22.37%). The top three together account for about 80%. The dilemma of independent distributors is that they lack their own users and liquidity, and must rely on channel distribution channels. Ondo pursues a multi-chain infrastructure route, xStocks expands horizontally and supplements governance voting rights vertically, while Securitize focuses on institutional efforts to avoid direct competition. But once channel owners step in to build their own distribution, distributors go from partners to replaceable suppliers. bStocks, Bitget's Reality platform, and Gate's gStocks are all like this. The brokerage direct connection route directly targets users' most basic needs: buying US stocks that are truly backed by underlying assets. In the first nine days after Mou'an's real stock spot was listed, the average daily turnover was $143 million, three to four times that of tokenized spot during the same period. Gate, Coinbase, and Crypto.com have also launched real equity products through licensed clearing channels like Alpaca. The value of tokenization remains—self-custody, 24×7 trading, composability