Each few quarters, the controversy round Nvidia strikes. For some time, it was about whether or not AI demand was actual. Then it was about whether or not margins might maintain. Now, heading into its August 26 earnings report, the query is whether or not the subsequent product cycle can maintain an organization already working at this tempo from slowing down.
Financial institution of America thinks it could actually. And the word behind that view is price studying earlier than the earnings date arrives.
Financial institution of America Nvidia earnings preview and $350 value goal
In a word shared with TheStreet on August 7, Financial institution of America analyst Vivek Arya known as Nvidia (NVDA) his prime sector decide forward of the corporate’s fiscal Q2 FY2027 outcomes. He expects income of $94 billion to $95 billion, roughly $3 billion to $4 billion above Nvidia’s personal $91 billion steering. That steering excludes any China knowledge heart compute income, that means precise outcomes might are available larger if modest shipments to that market resume. Third-quarter steering, he says, ought to are available at $107 billion to $108 billion, nicely above the roughly $104 billion Wall Avenue is at the moment modeling.
For context, Nvidia reported Q1 FY2027 income of $81.6 billion in Might, up 85% 12 months over 12 months, with knowledge heart income of $75.2 billion. The Q2 steering of $91 billion implied continued sequential development. BofA’s $94 billion to $95 billion estimate would lengthen that momentum additional.
However Arya is not actually writing concerning the subsequent quarter. “The graduation of Vera Rubin next-generation chip deliveries marks the start of an prolonged improve cycle spanning a number of quarters,” he wrote. That is the argument. Not only a beat. A brand new cycle.
Arya has a $350 value goal on Nvidia, representing roughly 56% upside from $223.96 on the time of the word. He factors out that the inventory is buying and selling at about 16 instances ahead earnings, its lowest valuation in roughly a decade, even because the earnings trajectory continues to rise, as TheStreet reported.
Why Nvidia Vera Rubin might set off a multi-quarter improve cycle
Nvidia confirmed at GTC Taipei in June that Vera Rubin has entered full manufacturing. The platform pairs Rubin GPUs with the brand new Vera CPU and is predicted to be accessible from cloud companions within the second half of 2026. AWS, Google Cloud, Microsoft and Oracle are already making ready deployments, with OpenAI, Anthropic and SpaceX among the many first clients. Nvidia is focusing on sufficient capability to require 2 gigawatts of energy for the buildout, as TheStreet reported.
GPU spot rental costs are close to all-time highs, Arya notes within the word. The B200 is working at about $5.66 per hour, the H100 at $2.80 per hour, the A100 at $1.64. That knowledge level issues as a result of it addresses one of many persistent doubts concerning the AI commerce. If clients can nonetheless lease compute at these costs and earn money from it, they’ve each cause to maintain shopping for the subsequent era of {hardware}. The priority about return on funding fades when the rental market is that this robust.
Extra Nvidia:
The Vera CPU is the a part of Rubin that Financial institution of America finds most fascinating past the GPU story. An earlier word from the financial institution known as it “the only biggest new addition because the GPU.” Arya’s present word tasks Vera CPU gross sales within the second half of fiscal 2027 at roughly $20 billion, with an annual run price of $50 billion or extra by fiscal 2028. That trajectory, if it holds, would make Nvidia the biggest server CPU vendor.
Nvidia gross margin and reminiscence price inflation outlook for 2027
Reminiscence price inflation has develop into one of many louder considerations about Nvidia’s margins. DRAM now makes up 40% to 50% of whole manufacturing prices, up from 15% to twenty% traditionally. The fear is that as Nvidia strikes to extra memory-intensive architectures, these rising prices eat into its famously excessive gross margins sooner than the corporate can value round them.
Arya’s word pushes again on that straight. For Vera Rubin compute racks particularly, the reminiscence price improve quantities to about 60 foundation factors of gross margin stress versus Blackwell Extremely. Gross margins are anticipated to settle at 73% to 74% over time, down modestly from about 75% now. That’s not the type of structural margin erosion that will break the funding case.
The larger quantity is on the pod stage. Full AI pods, which bundle in additional reminiscence and storage, might see as much as 500 foundation factors of margin impression. However Arya expects that blend to remain small initially. Nvidia’s long-term provide agreements with SK Hynix and its pricing energy, given GPU rental charges close to all-time highs, give the corporate room to cross by prices quite than take in them.
Nvidia confirmed at GTC Taipei in June that Vera Rubin has entered full productionOta/Getty Pictures
Nvidia OpenAI round financing and free money movement defined
Nvidia has dedicated roughly $70 billion in direct fairness stakes to ecosystem companions. That features $30 billion to OpenAI, as much as $10 billion to Anthropic, and $5 billion to Ilya Sutskever’s Secure Superintelligence. Some traders have questioned whether or not these preparations are round, primarily Nvidia financing the purchasers who purchase Nvidia chips.
Arya’s word addresses this straight. In opposition to the $70 billion in direct investments, Nvidia is predicted to generate roughly $470 billion in free money movement throughout 2026 and 2027. The $70 billion represents about 15% of that. There may be room to maintain returning roughly 50% of free money movement to shareholders whereas making these investments, in response to Benzinga.
The $250 billion backstop tied to an OpenAI and SB Vitality campus in Ohio is a special type of dedication. It’s not upfront money. It’s a contingent assure that solely triggers if OpenAI defaults on the lease, with publicity back-loaded to 2028 and past. When Nvidia is predicted to be producing $300 billion to $500 billion per 12 months in free money movement by then, the chance profile appears to be like completely different than the headline quantity suggests.
NVDA inventory valuation at decade low and the BofA bull case
At 16 instances ahead earnings, Nvidia is at its least expensive in a couple of decade. The financial institution’s EPS projections put Nvidia at greater than $13 per share by 2027 and greater than $25 by 2030, assuming the AI knowledge heart market develops alongside Arya’s mannequin. That mannequin assumes Nvidia holds greater than 70% share of a market the financial institution sees rising previous $1.7 trillion in AI knowledge heart programs.
The dangers are actual. AMD is gaining floor in AI accelerators. The foremost cloud firms are constructing extra of their very own customized chips. China export restrictions stay an overhang. Hyperscaler capital spending might get extra uneven if the returns on AI infrastructure disappoint. None of these go away.
However the setup Financial institution of America is describing is an organization with this stage of earnings energy sitting at a valuation that does not replicate it. Whether or not August 26 confirms that or not nearly misses the purpose. The larger query is whether or not Rubin delivers what Arya’s word says it would. If it does, one earnings report is the beginning of one thing, not the factor itself.
