NVIDIA Announces Financial Results for Second Quarter Fiscal 2025

[Nomad76]

NVIDIA today reported revenue for the second quarter ended July 28, 2024, of $30.0 billion, up 15% from the previous quarter and up 122% from a year ago. For the quarter, GAAP earnings per diluted share was $0.67, up 12% from the previous quarter and up 168% from a year ago. Non-GAAP earnings per diluted share was $0.68, up 11% from the previous quarter and up 152% from a year ago.

"Hopper demand remains strong, and the anticipation for Blackwell is incredible," said Jensen Huang, founder and CEO of NVIDIA. "NVIDIA achieved record revenues as global data centers are in full throttle to modernize the entire computing stack with accelerated computing and generative AI."





"Blackwell samples are shipping to our partners and customers. Spectrum-X Ethernet for AI and NVIDIA AI Enterprise software are two new product categories achieving significant scale, demonstrating that NVIDIA is a full-stack and data center-scale platform. Across the entire stack and ecosystem, we are helping frontier model makers to consumer internet services, and now enterprises. Generative AI will revolutionize every industry."

During the first half of fiscal 2025, NVIDIA returned $15.4 billion to shareholders in the form of shares repurchased and cash dividends. As of the end of the second quarter, the company had $7.5 billion remaining under its share repurchase authorization. On August 26, 2024, the Board of Directors approved an additional $50.0 billion in share repurchase authorization, without expiration.

NVIDIA will pay its next quarterly cash dividend of $0.01 per share on October 3, 2024, to all shareholders of record on September 12, 2024.

On June 7, 2024, NVIDIA completed a ten-for-one forward stock split. All share and per-share amounts presented have been retroactively adjusted to reflect the stock split.

Outlook
NVIDIA's outlook for the third quarter of fiscal 2025 is as follows:

• Revenue is expected to be $32.5 billion, plus or minus 2%.
• GAAP and non-GAAP gross margins are expected to be 74.4% and 75.0%, respectively, plus or minus 50 basis points. For the full year, gross margins are expected to be in the mid-70% range.
• GAAP and non-GAAP operating expenses are expected to be approximately $4.3 billion and $3.0 billion, respectively. Full-year operating expenses are expected to grow in the mid- to upper-40% range.
• GAAP and non-GAAP other income and expense are expected to be an income of approximately $350 million, excluding gains and losses from non-affiliated investments and publicly-held equity securities.
• GAAP and non-GAAP tax rates are expected to be 17%, plus or minus 1%, excluding any discrete items.

Highlights
NVIDIA achieved progress since its previous earnings announcement in these areas:

Data Center

• Second-quarter revenue was a record $26.3 billion, up 16% from the previous quarter and up 154% from a year ago.
• Announced that the combination of NVIDIA H200 Tensor Core and NVIDIA Blackwell architecture B200 Tensor Core processors swept the latest industry-standard MLPerf benchmark results for inference.
• Revealed that H200 GPU-powered systems are now available on CoreWeave, the first cloud service provider to announce general availability.
• Unveiled an array of Blackwell systems featuring NVIDIA Grace CPUs, networking and infrastructure from top manufacturers such as GIGABYTE, QCT and Wiwynn.
• Reported broad adoption of the NVIDIA Spectrum-X Ethernet networking platform by cloud service providers, GPU cloud providers and enterprises, as well as partners incorporating it into their offerings.
• Released NVIDIA NIM for broad availability to developers globally and announced more than 150 companies are integrating microservices into their platforms to speed generative AI application development.
• Unveiled an inference service with Hugging Face powered by NIM microservices on NVIDIA DGX Cloud to enable developers to deploy popular large language models.
• Introduced an NVIDIA AI Foundry service and NIM inference microservices to accelerate generative AI for the world's enterprises with the Llama 3.1 collection of models.
• Announced Japan advanced its sovereign AI capabilities with its ABCI 3.0 supercomputer, integrating H200 GPUs and NVIDIA Quantum-2 InfiniBand networking.
• Accelerated quantum computing efforts at national supercomputing centers around the world with the open-source NVIDIA CUDA-Q platform.

Gaming and AI PC

• Second-quarter Gaming revenue was $2.9 billion, up 9% from the previous quarter and up 16% from a year ago.
• Announced NVIDIA ACE, a suite of generative AI technologies that bring digital humans to life, now includes NVIDIA Nemotron-4 4B, a small language model for on-device inference, and is available in early access for RTX AI PCs.
• Introduced Project G-Assist, a technology preview demonstrating the power of AI agents to assist gamers and creators in real time.
• Announced new NVIDIA GeForce RTX and DLSS titles, including Indiana Jones and the Great Circle, Dune: Awakening and Dragon Age: The Veilguard, bringing the total number of RTX games and apps to over 600.
• Surpassed 2,000 games on GeForce NOW, expanded the service into Japan and announced launches of Black Myth: Wukong and Star Wars Outlaws.

Professional Visualization

• Second-quarter revenue was $454 million, up 6% from the previous quarter and up 20% from a year ago.
• Introduced generative AI models and NIM microservices for OpenUSD to accelerate workflows and the development of industrial digital twins and robotics.
• Announced major Taiwanese electronics makers are creating more autonomous factories with a new reference workflow that combines NVIDIA Metropolis vision AI, NVIDIA Omniverse simulation and NVIDIA Isaac AI robot development.

Automotive and Robotics

• Second-quarter Automotive revenue was $346 million, up 5% from the previous quarter and up 37% from a year ago.
• Unveiled the world's leaders in robot development, including BYD Electronics, Siemens and Teradyne Robotics, are adopting the Isaac robotics platform for R&D and production.
• Announced Omniverse Cloud Sensor RTX microservices to enable physically accurate sensor simulation to speed development of autonomous machines.
• Won the Autonomous Grand Challenge at the Computer Vision and Pattern Recognition conference in the category of End-to-End Driving at Scale for advances in building physical, generative AI applications for autonomous vehicle development.

View at TechPowerUp Main Site | Source

 

[mb194dc]

Just give some decent use cases to justify the hundreds of billions, of not trillions of dollars once you factor power and building costs in... Spent on LLMs and ML.

I don't see it but I'll be happy to be wrong. 2025 or bit later probably crunch time.

 

[bonehead123]

revenue for the second quarter ended July 28, 2024, of $30.0 billion, up 15% from the previous quarter and up 122% from a year ago

This, but yet such an insignificant $.01 per share dividend...w.T.f.....

Glad I own a buttload of shares, but even so, I still feel like this is total friggin BS..... oh well, I guess Jacket Man needs to buy a few more new jackets (or jets or condos or whatever)

And yes, I own many other tech stocks (large & small) that pay way larger dividends, no wonder sooo many people refer to them as "nGreediya", including me

 

[thesmokingman]

Money printing continues...

 

[chrcoluk]

So in short without the share buyback prices could be 25% lower.

 

[nguyen]

2.9bil revenue in gaming in Q2 huh, Q3 is going to be even higher thanks to Black Myth Wukong and back-to-school PC sale.

 

[Chaitanya]

Money printing continues...

For how long though?

 

[cvaldes]

This, but yet such an insignificant $.01 per share dividend...w.T.f.....

Nvidia doesn't really believe the typical retail investor can get a better ROI from their dividend payout. Nvidia issues the dividend solely to qualify for pension plans, retirement accounts, and certain funds that require that constituent equities issue a dividend.

That's the only reason they pay it out. It's the same with the other Magnificent Seven/major tech companies: AAPL, GOOG, AMZN, META, NFLX, blah blah blah, et al. It's just to qualify the stock for those investment vehicles and portfolios, to keep fund administrators satisfied, not retail investors.

And they're right. If you want a bigger return, diversify by selling off some of your NVDA holdings and stick your earnings somewhere else.

Clamoring for bigger NVDA dividends at this point in the company's trajectory is pretty pointless. Maybe in a few years we can revisit this topic but this is an utterly inane discussion in 2024.

If you want bigger dividend yields from a tech company, go stick your money in INTC (2.55% yield). Ahahahahahahahaha!!!!

 

[thesmokingman]

For how long though?

2 years is my guess. No one can afford 10K H100/H200, so once the big boys get up to saturation, the small guys will buy server time. Then those big boys are concurrently building/ramping their own vertically integrated setups like Tesla and Open AI/MSFT. And then others like Google have been on in-house silicon the whole time. In any case there's only so long any big boy can stomach filling a suppliers coffers at 76% gross margin lol.

 

[cvaldes]

Two years is probably a reasonable guess.

It's important to point out that AI isn't just a chip. A very large part of Nvidia's dominance comes from its software environment. Even TinyBox (makers of small AI servers using consumer GPUs from either AMD or Nvidia) admits themselves in no minced language that Nvidia's software is way better than AMD's.

There are a bunch of companies working on alternatives to CUDA but they aren't going to topple Nvidia overnight. And if there are ten different CUDA alternatives that dilutes the effort to dethrone Nvidia's dominance. A similar parallel can be found in graphics technology: DirectX, Vulkan, Apple Metal, etc. You can't just come up with a CUDA alternative and expect developers to dump CUDA in a month.

And beyond the GPU there's still a bunch of supporting infrastructure on the card, in the server chassis, interconnects between boxes, networking, disk storage, power & cooling, etc. There's a lot more to AI computing than a GPU die rolling off some foundry manufacturing line in Taiwan (TSMC, Samsung, whoever).

You don't just drive to Best Buy, select an AI accelerator off a shelf, bring it back to your cubicle, pop the lid off your PC, stick it into an empty PCIe slot and start running AI code (LLM or something different) in a production setting.

When Nvidia forecasts lower gross margins that'll be a clearer sign that the rest of the market is posing a serious challenge. It will start at the low end though. The bigger threat is one of the other big boys (Amazon, Alphabet, Meta, Microsoft, etc.) coming up with their in-house silicon and whether or not they attempt to market the silicon to other users.

That still leaves the topic of a mature software environment (both vertical and horizontal).

 

[las]

For how long though?

For years. AI is going nowhere anytime soon.
However, eventually, AMD and Intel is going to get more and more competitive. Today, Nvidia is in a league of its own tho.

 

[R0H1T]

For how long though?

As long as the "markets" don't course correct too much! And when/if they do that money will go elsewhere, who knows maybe space is the next AI?

There are a bunch of companies working on alternatives to CUDA but they aren't going to topple Nvidia overnight. And if there are ten different CUDA alternatives that dilutes the effort to dethrone Nvidia's dominance.

Nvidia's biggest threat isn't a "chip company" per se it's in house solutions like from Tesla, AWS, Google, Alibaba, MS(?) & whatever Oracle can come up with. And they are Nvidia's biggest clients!

 

[Bwaze]

2.9bil revenue in gaming in Q2 huh, Q3 is going to be even higher thanks to Black Myth Wukong and back-to-school PC sale.

"Gaming and AI PC
Second-quarter Gaming revenue was $2.9 billion, up 9% from the previous quarter and up 16% from a year ago".

Does anyone really think gaming revenue increase is from GPU sales to gamers? Ada Lovelace will be two years old any time now, and replaced with new generation. Buyers that don't know that most probably aren't the ones buying top of the line GPUs.

I think large majority of that increase is due to AI. Some companies are creating LLM, ML servers using gaming cards. Some people are buying powerful GPUs to run Stable Diffusion and other local AI tools.

But most growth I think is just Nvidia spreading AI income into all sectors, so it looks like an organic growth of a company as a whole, not just a bubble dependant on a single market demand that could evaporate at any moment. They learned that trick well during cryptomining booms, when even Automotive achieved stellar growth, and then fell when the crypto crashed - shareholders tried to sue Nvidia for that behaviour, but were just told they can't prove anything illegal.

Just look at the growth of Professional Visualisation and Automotive now, without any new product that would really explain it. Déjà vu, right?

 

[Vayra86]

"Gaming and AI PC
Second-quarter Gaming revenue was $2.9 billion, up 9% from the previous quarter and up 16% from a year ago".

Does anyone really think gaming revenue increase is from GPU sales to gamers? Ada Lovelace will be two years old any time now, and replaced with new generation. Buyers that don't know that most probably aren't the ones buying top of the line GPUs.

I think large majority of that increase is due to AI. Some companies are creating LLM, ML servers using gaming cards. Some people are buying powerful GPUs to run Stable Diffusion and other local AI tools.

But most growth I think is just Nvidia spreading AI income into all sectors, so it looks like an organic growth of a company as a whole, not just a bubble dependant on a single market demand that could evaporate at any moment. They learned that trick well during cryptomining booms, when even Automotive achieved stellar growth, and then fell when the crypto crashed - shareholders tried to sue Nvidia for that behaviour, but were just told they can't prove anything illegal.

Just look at the growth of Professional Visualisation and Automotive now, without any new product that would really explain it. Déjà vu, right?

Part of the growth is certainly from gamers/gaming, I reckon.

Gaming is a growth market, and even if the PC market is stalling in growth, people still need new GPUs and AMD is bleeding market share. So yes. They get sold to gamers. They also get sold to (semi) pro and enthusiasts that don't necessarily game (Or do), but that's not new either.

 

[Bwaze]

Part of the growth is certainly from gamers/gaming, I reckon.

Gaming is a growth market, and even if the PC market is stalling in growth, people still need new GPUs and AMD is bleeding market share. So yes. They get sold to gamers. They also get sold to (semi) pro and enthusiasts that don't necessarily game (Or do), but that's not new either.

I bet every generation since 90' saw a similar sales curve - hype and rush at the new generation debut, mid generation refresh with Ti cards, and then slowed sales until Nvidia and partners eventually offered discounts for the last couple of months before the new generation.

Unless it was the crypto season, then all bets were off.

And all this time gaming was of course growing market, that's nothing new.

But all of w sudden we have a generation on it's last leg that sells 16% better than last year when it was new-ish? In a market where people spend less and less due to "not really a recession, just your salary is worth less and less"?

Yeah. Right.

 

[TheLostSwede]

A quick breakdown.

x.com

x.com

 

[kondamin]

For years. AI is going nowhere anytime soon.
However, eventually, AMD and Intel is going to get more and more competitive. Today, Nvidia is in a league of its own tho.

Bubbles pop.

I can see Nvdia drop back to being valuated right around a trillion to well below it when everything grind to a halt

(this being if we don't get an other set of massive QE's and we see massive inflation again it might keep it's 3 trillion valuation but from a dollar that's just worth a third of what it's now)

 

[thesmokingman]

Bubbles pop.

I can see Nvdia drop back to being valuated right around a trillion to well below it when everything grind to a halt

(this being if we don't get an other set of massive QE's and we see massive inflation again it might keep it's 3 trillion valuation but from a dollar that's just worth a third of what it's now)

They're already prepped to move to the real point of AI, the incoming humanoid robotic boom. It's just too bad the leader wants no off the shelve parts and is doing it completely vertical. But they can still profit from the huge market it will bring from every other player which will be pretty freaking huge.

Elon Musk Says Tesla’s Optimus Robot Could Drive Company To $25 Trillion Valuation—Here’s What Experts Think

The robots will not come close to bringing in the kind of money Tesla’s cars do in the near future, and experts said it’s a long shot as to whether Musk could meaningfully launch Optimus by the end of 2025.

www.forbes.com

You don't just drive to Best Buy, select an AI accelerator off a shelf, bring it back to your cubicle, pop the lid off your PC, stick it into an empty PCIe slot and start running AI code (LLM or something different) in a production setting.

When Nvidia forecasts lower gross margins that'll be a clearer sign that the rest of the market is posing a serious challenge. It will start at the low end though. The bigger threat is one of the other big boys (Amazon, Alphabet, Meta, Microsoft, etc.) coming up with their in-house silicon and whether or not they attempt to market the silicon to other users.

That still leaves the topic of a mature software environment (both vertical and horizontal).

When you're building a $500 million dollar cluster Nvidia's moat is not a driving factor, you realize this right?

 

[cvaldes]

When you're building a $500 million dollar cluster Nvidia's moat is not a driving factor, you realize this right?

The number of companies that are capable of building their own AI solution on their own can be counted on one hand. They will do so regardless of Nvidia does/says. Not everyone is allocating a billion in their budget for AI.

It's the other 495 companies in the Fortune 500. JPMorganChase, Walmart, United Health, Visa, ExxonMobil, Mastercard, Procter & Gamble, Johnson & Johnson, Home Depot, Coca-Cola, Merck, et al aren't going tape out AI GPUs anytime soon.

Hell, I'm not aware of any effort by major software players like Oracle or Salesforce in creating their own silicon with the goal of ditching Nvidia/AMD/Intel/Broadcom/etc.

The people at Nvidia already know which big customers is working on developing their own silicon. And not everyone is going to succeed. Likely there will be companies who abandon this endeavor, others may persevere but end up with an in-house solution that doesn't show an differentiating performance characteristics. Building chips is hard. Like really hard. It's also really expensive.

But this is the same with any product that these companies use. It's not like Morgan Stanley is going to start growing trees for harvest to make toilet paper for their offices. Or carpets. Or refrigerators. Hell, most of these companies won't even build their own computers. They buy from the big enterprise vendors (Dell, HP, etc.) for the service and support (both hardware and software).

Only the biggest cloud providers are going to consider rolling their own silicon. And Tesla. Only because Elon Musk is Elon Musk.

And the best AI hardware chip designers that aren't at these companies are working at startups not places like Chevron or McDonald's. At some point much chip design will be done by AI itself but to get there you need to start somewhere. And it's not just chip design anyhow. As I pointed out, there's a lot more to an AI solution than a wafer full of AI dies rolling off a foundry manufacturing line.

Hell, look at Intel. They can't even run a decent semiconductor business and they are The Business. If the experts are having problems competing in the chip industry what are the chances of some mid-tier Fortune 500 company rolling their own in the near future? In fact, if you could acquire all the parts and were given the instructions/assistance on putting it together, how much success can you expect. Look at EVGA? They were one of the best consumer GPU AIB partners and they quit.

 

[las]

Bubbles pop.

I can see Nvdia drop back to being valuated right around a trillion to well below it when everything grind to a halt

(this being if we don't get an other set of massive QE's and we see massive inflation again it might keep it's 3 trillion valuation but from a dollar that's just worth a third of what it's now)

Yeah they do, eventually, we are not even close tho. Nvidia keeps breaking expectations and nothing really changed.

 

[kondamin]

Yeah they do, eventually, we are not even close tho. Nvidia keeps breaking expectations and nothing really changed.

Nvidia might be but the rest of the market is performing well below expectations, people need money to buy clusters when that money isn't there nvidia can't sell much.

 

[las]

Nvidia might be but the rest of the market is performing well below expectations, people need money to buy clusters when that money isn't there nvidia can't sell much.

Nothing in my portfolio is performing below expectations