Nvidia Stock Slips 1% Ahead of Earnings as AI Server Prices Rise 15%

Key Takeaways

• Nvidia stock closed 0.98% lower at $214.72 as investors positioned cautiously before Wednesday’s fiscal second-quarter earnings.
• Wall Street expects revenue of approximately $92.16 billion and adjusted EPS of $2.09, with both figures nearly doubling year on year.
• Reported AI server price increases above 15% have placed memory costs, gross margins and customer spending plans under closer scrutiny.

Nvidia stock moved lower ahead of one of the most closely watched earnings reports of the quarter, with investors balancing expectations for another period of exceptional growth against concerns about elevated valuations and rising AI infrastructure costs.

Shares finished Friday at $214.72, down 0.98%, extending a pullback that has left the stock roughly 9% below its May record above $236. Nvidia remains up approximately 15% in 2026, but the recent decline suggests traders are becoming more selective as the company approaches an earnings release carrying unusually high expectations.

Nvidia Stock Retreats Before Wednesday’s Earnings

Nvidia will publish its fiscal second-quarter 2027 results after the US market closes on Wednesday, 26 August. The report could influence sentiment across semiconductor stocks, AI infrastructure companies and the broader Nasdaq 100 because Nvidia now carries a market value of more than $5 trillion.

The stock’s recent weakness reflects more than simple profit-taking. Investors are questioning whether even rapid growth will be sufficient to support another rally after Nvidia shares declined following each of the company’s previous four quarterly reports.

Options markets imply a move of approximately 6% in either direction by the end of earnings week. Such a swing could lift the stock back towards its May record or push it below $205. The implied volatility underlines how sensitive Nvidia’s valuation has become to relatively small differences between reported results, guidance and Wall Street expectations.

Nvidia Earnings Forecast Points to Near-Doubling Revenue

Analysts expect Nvidia to report record quarterly revenue of approximately $92.16 billion, representing growth of nearly 97% from the $46.7 billion recorded a year earlier. Adjusted earnings are forecast at $2.09 per share, almost double the previous year’s $1.05.

Data centre revenue is expected to reach approximately $85.67 billion, up 108% year on year. That would mean the division contributes more than 90% of total revenue, highlighting Nvidia’s dependence on spending by hyperscale cloud companies and other operators building large AI computing clusters.

Nvidia previously guided for second-quarter revenue of $91 billion, plus or minus 2%, while excluding data centre computing sales to China. It also projected a non-GAAP gross margin of approximately 75%. The consensus forecast is therefore slightly above the midpoint of the company’s guidance.

A headline beat alone may not be enough to generate a sustained share-price rise. Markets are likely to focus more heavily on Nvidia’s outlook for the October quarter, including demand for Blackwell Ultra systems, the transition towards Vera Rubin and the availability of advanced memory and networking components.

Citi Expects Nvidia Stock to Rise After Earnings

Citi has maintained a Buy rating and a $300 price target for Nvidia while forecasting stronger results than the broader market expects.

The bank estimates July-quarter revenue of approximately $93 billion, around $1 billion above Wall Street’s projection. It expects revenue to rise to $105 billion in the October quarter, roughly $1.5 billion above consensus and 13% higher sequentially.

The bullish forecast is partly based on the expansion of Nvidia’s B300 systems and stronger shipments of 1.6-terabit optical transceivers. These components support high-speed communication between processors inside increasingly large AI data centres.

Citi also expects Nvidia’s data centre revenue to increase by 15% sequentially in the July quarter and by another 14% in the October quarter. Wall Street currently anticipates growth of approximately 13% in both periods.

The bank believes Nvidia has secured sufficient high-bandwidth memory capacity for 2026 and 2027. If confirmed, that could reduce concerns that limited memory supply will restrict shipments of its most advanced AI platforms.

AI Server Prices Reportedly Rising More Than 15%

Nvidia’s earnings arrive as some of its largest customers face reported price increases of more than 15% for servers containing the company’s AI chips.

The increases are expected to affect systems shipped from early 2027, including platforms based on Grace Blackwell and next-generation Vera Rubin processors. The final increase may vary depending on the chip generation and the amount and type of memory installed.

Server manufacturers supplying major data centre operators such as Microsoft, Google and Oracle have reportedly started passing the higher prices to customers. Nvidia has not publicly confirmed the reported changes.

Memory has become a significantly larger part of the cost of an AI server. Server DRAM prices roughly doubled during the first quarter of 2026, while memory can now represent around one-quarter of the cost of a high-end server rack.

The price increases could support Nvidia’s gross margin by transferring some component inflation to customers. However, they could also make data centre projects more expensive and intensify questions about whether cloud providers can continue expanding capital expenditure at the current pace.

Memory Costs Put Nvidia’s Pricing Power to the Test

Nvidia’s ability to raise system prices illustrates the strength of demand for its hardware and software ecosystem. Customers seeking the highest levels of AI training and inference performance still have limited alternatives capable of matching the combination of Nvidia GPUs, networking technology and CUDA software.

Nevertheless, the changing cost structure introduces a new risk. AI investment was initially constrained primarily by GPU availability, power capacity and data centre construction. Memory supply has now emerged as another potential bottleneck.

Higher DRAM and high-bandwidth memory prices could benefit suppliers such as Micron, SK Hynix and Samsung. For Nvidia, however, higher component expenses may pressure margins unless price increases fully offset the additional costs.

Investors will therefore monitor Nvidia’s gross-margin guidance particularly closely. A margin forecast near or above 75% would suggest that the company retains substantial pricing power. A weaker outlook could indicate that component inflation and the transition between Blackwell and Vera Rubin are becoming more expensive.

What Could Move Nvidia Stock After the Report?

A bullish reaction would probably require more than Nvidia simply meeting the consensus estimates. Revenue above $92 billion, October-quarter guidance exceeding $103 billion and stable gross margins could reinforce expectations that AI infrastructure demand remains strong.

Updates on B300 production, Vera Rubin shipments, high-bandwidth memory supply and customer capital spending could also shape the market response. Clear evidence that Nvidia can absorb higher memory costs without weakening demand would support the optimistic case.

A more cautious outlook, slower data centre growth or pressure on gross margins could trigger a negative reaction even if the company reports record results. Investors may also examine Nvidia’s expanding role in financing AI infrastructure, as well as competition from AMD, Broadcom and custom chips developed by major cloud providers.

With options pricing in a move of around 6%, Wednesday’s results could determine whether Nvidia stock returns towards its record high or extends its recent decline below the $205 area.