Author: Rita
Over the past four quarters, NVIDIA's revenue guidance has exceeded market consensus by an average of 4%, yet its stock price has fallen by an average of 3% and 5% in the subsequent 7 and 30 days, respectively. On August 24, JPMorgan released its preview report for NVIDIA's Q2 FY2027, reiterating an Overweight rating with a $280 price target. While surpassing earnings expectations is a foregone conclusion, for the stock price to rise, it depends on how management addresses four key questions.
GB300 rack shipments are expected to grow sequentially by approximately 15% to 17,000-18,000 units, with the Vera Rubin platform beginning to contribute revenue in the October quarter. Total revenue for the July quarter is projected to be in the range of $94-95 billion. Guidance for the October quarter could reach $107-108 billion, about 3% above the market consensus of $104.5 billion. JPMorgan believes that the numbers alone are no longer sufficient to drive the stock price; the market is truly waiting for positive responses from management on four issues: competition, the China business, HBM supply, and AI infrastructure investment.
Exceeding Expectations is the Norm, Market Pricing Logic Has Changed
Revenue for the July quarter is expected to be in the $94-95 billion range, with a midpoint of $94.5 billion, slightly above the market consensus of $92.1 billion. GB300 rack shipments are estimated to have grown sequentially by about 15% to 17,000-18,000 units, with GB200 simultaneously clearing out remaining orders. The growth rate of the Networking business has converged with that of the Compute business, both achieving mid-teens sequential growth. The Edge business grew by mid-single to high-single digits sequentially, with a year-over-year growth rate of approximately 20%, slowing from the 29% YoY growth in Q1.
The guidance for the October quarter is the main event. JPMorgan expects rack shipments to grow 13-14% sequentially to 19,000-20,000 units, including 1,000-2,000 units of initial Vera Rubin shipments. The blended Average Selling Price for Vera Rubin is expected to increase by 5-10%, with the Compute business growing at a mid-teens rate sequentially. Coupled with synchronized growth in the Networking business and the initiation of H200 shipments to China, the October quarter revenue guidance could land between $107-108 billion, about 3% above the consensus of $104.5 billion.
However, exceeding expectations is no longer a catalyst for the stock price. Over the past four quarters, revenue guidance has averaged 4% above consensus, yet the stock price has fallen by an average of 3% and 5% in the subsequent 7 and 30 days, respectively. JPMorgan believes the market has fully priced in NVIDIA's persistent outperformance, and the real focus has shifted to four more structural issues.
H200 Shipments to China Could Be the Biggest Upside Variable
Long-term approval for H200 shipments to China has been obtained but has been slow to commence, with recent signs of initial deliveries to customers. JPMorgan estimates that every 100,000 H200 units shipped could generate approximately $3 billion in incremental revenue. This is not only an upside variable for the October quarter but also creates upward pressure on revenue expectations for subsequent quarters if the shipment pace continues.
The China business has been largely stagnant over the past few quarters. The initiation of H200 shipments means this revenue stream is starting to contribute from scratch again. Given the scale of demand for high-performance computing chips in the Chinese market, this variable may be significantly underestimated by the market.
HBM Specification Reductions Are Supply Constraints
Recent supply chain checks indicate NVIDIA has adjusted the HBM configuration for the Rubin platform. Rubin has been modified from its original plan to two SKU versions, using either 288GB or 192GB of HBM4. Rubin Ultra has been reduced from 16-layer HBM4E to 8-layer or 12-layer HBM4E. The SOCAMM memory capacity for the Vera CPU has been halved from 1.5TB to 768GB.
JPMorgan judges these adjustments are almost entirely driven by supply shortages. To fulfill compute capacity commitments to customers, NVIDIA had to reduce memory configurations. The impact on the revenue side depends on the final SKU shipment mix, while the impact on gross margins depends on the relative change between memory costs and product pricing. How management addresses this issue on the earnings call will be a key focus for investors.
Vera Rubin Cost Reduction Magnitude is a Competitive Moat
Competition from AI ASICs and alternative compute platforms is a persistent narrative headwind, and JPMorgan believes it will be difficult for NVIDIA to completely eliminate market concerns. However, management can emphasize two facts. Platform flexibility is its key advantage in rapidly penetrating the enterprise and sovereign AI markets. Vera Rubin is expected to reduce platform cost per token by approximately 90% compared to Blackwell Ultra.
This magnitude of cost reduction means that even if ASICs or XPUs gain share, NVIDIA can maintain customer stickiness by continuously lowering unit costs. JPMorgan expects GPU, ASIC, and XPU shares of the total AI compute addressable market to trend toward parity over the next few years, but this does not mean NVIDIA's growth stalls; rather, the entire pie is rapidly expanding.
Valuation Has Fallen to Historical Lows
From a valuation perspective, NVIDIA's current stock price corresponds to price-to-earnings ratios of approximately 17x and 13x based on market consensus EPS for calendar years 2027 and 2028, respectively, well below historical averages. JPMorgan believes that the continuous upward revision of EPS expectations provides upward momentum for the stock price, a pattern that continues as profits expand further.
The $280 price target corresponds to a P/E of about 28x based on an exit annualized EPS of around $10 for calendar year 2026, aligning with the company's long-term EPS growth rate. On the downside, PC gaming demand could face pressure from macro headwinds, exposing about 53% of revenue. Reduced deep learning deployments by hyperscalers or intensified competition also pose risks.
JPMorgan's conclusion is clear. Simply exceeding expectations on numbers is no longer enough; the market wants positive answers on the four issues of competition, China, HBM, and AI investment. The valuation already allows ample room for uncertainty.
Disclaimer
This article is a compilation and interpretation of a third-party brokerage research report (JPMorgan, August 24, 2026) by Chaoxiang Research, combined with the collation of public market information. The ratings, price targets, earnings forecasts, and related judgments cited in the article represent the views of that brokerage's analyst and their institution only. They do not represent the views of Chaoxiang Research and do not constitute any investment advice.
The market involves risks, and decisions should be made independently. This article should not be used as a basis for buying or selling any securities.
