Huang Renxun has left Beijing, and NVIDIA has not disclosed whether Huang Renxun met with the DeepSeek founder.

Wall Street’s differing views on year-end U.S. stock market levels are narrowing, with 8,000 points for the S&P 500 emerging as a new psychological anchor. Fundstrat raised its year-end target for the index from 7,700 to 8,000; Goldman Sachs, Morgan Stanley, Deutsche Bank, and Societe Generale have also set targets near this level. Goldman Sachs previously lifted its 2026 S&P 500 target from 7,600 to 8,000, citing that earnings growth and AI investments continue to underpin the index, rather than relying solely on valuation expansion. Even the more conservative cohort is boosting targets: JPMorgan Chase raised its target from 7,600 to 7,800, while Barclays and Stifel also adjusted their year-end targets to 7,800. Barclays lifted its 2026 S&P 500 earnings per share forecast from $321 to $337, and set a 2027 target of 8,800 points. The shared rationale behind these moves includes upward revisions to corporate earnings, AI capital expenditure, improved visibility into tech sector profits, and easing geopolitical risks. However, this consensus does not equate to zero risk. JPMorgan Chase warned that momentum stocks, semiconductors, storage stocks, and second-tier AI concepts have become overcrowded in trading, and low-quality and speculative growth stocks may see sharp declines. It favors a barbell allocation strategy of "quality growth + low-volatility quality".

According to Bitget market data, South Korea's KOSPI index has extended its downward trend, plunging 7.00% intraday to 8303.60 points.

On-chain analyst Ai Yi (@ai_9684xtpa) has monitored that three addresses suspected to belong to the same whale or entity opened a total of 500 BTC long positions with 20x leverage last night, at an average entry price of $59,253, and are currently facing an unrealized loss of $258,000. The three addresses exhibit similar trading behaviors, and their miner fees were all bridged via Bridgers.

According to Bitget market data, spot gold has once again fallen below $4,000 per ounce, with an intraday decline of 0.68%.

According to Hyperinsight monitoring, ETH fell again early this morning, once approaching $1,500. The largest ETH long on Hyperliquid, the "BIT-linked whale", saw synchronized massive losses across four of its addresses’ long positions, with over $10.1 million lost in the past 24 hours. Its total unrealized loss has now expanded to $88.5 million. Calculated from its position opening principal of roughly $16.5 million, the loss exceeds 5.3 times the principal. To avoid liquidation amid ETH’s continued decline this morning, the whale’s addresses have added margin to several of their accounts over the past 24 hours, totaling around $8 million. After this top-up, the latest liquidation line has dropped to $1,143. The whale currently holds a total of 120,000 ETH long positions on Hyperliquid, valued at approximately $224 million, with an overall weighted leverage of around 18x. Based on the proportion of funds across addresses, its average entry price is roughly $2,261. It is understood that these addresses belong to BIT (formerly Matrixport), a digital financial services group. They have repeatedly received fund transfers from BIT, and previously held ETH long positions worth hundreds of millions, with profits exceeding $50 million. Addresses: 0xa5b0edf6b55128e0ddae8e51ac538c3188401d41 0x6c8512516ce5669d35113a11ca8b8de322fd84f6 0xa875890465dA20062bCF3b024Bf7d54E69C725a8 0xfd423284f6a9c73a2a3d53cab8921d6533533d97

Against the backdrop of ongoing fallout from Micron Technology’s (MU) earnings report, global tech stocks have faced broad sell-offs, with major tech giants including Apple (AAPL) also plummeting sharply, stoking market concerns over the profit outlook of traditional tech leaders. Analysts noted that while Apple has not been deeply involved in the AI infrastructure capital expenditure race, amid rising AI supply chain costs and pressure to adjust product prices, market fears that price hikes could dampen end-user demand, further weighing on revenue and profit performance in the coming quarters, prompting investors to lock in profits for now. Meanwhile, market focus has shifted from AI computing power demand to capital returns. Data compiled by Bloomberg shows that the massive free cash flow accumulated by the world’s five largest hyperscale cloud providers over the past two decades has shrunk significantly amid intense AI capital expenditure in the last two years. Analysts believe the current market is exhibiting the classic "sell shovels, buy shovels under pressure" trading logic: AI hardware and semiconductor suppliers are seeing their stock prices rise amid sustained orders, while cloud computing and internet giants bearing massive AI investments are facing pressure on cash flow and profitability. If AI hardware costs continue to climb in the future, and the AI infrastructure build-out cycle lasts roughly a decade as Jensen Huang has stated, who will ultimately bear the ongoing investments and how commercial returns will be realized will become the core test facing global tech stock valuations.