Nvidia earnings will give Wall Street a peek into AI demand heading into 2024

When Nvidia reports third-quarter results connected Tuesday, analysts are expecting to spot gross maturation of implicit 170%.

If that's not astounding enough, the company's forecast for the 4th quarter, according to LSEG estimates, is apt to amusement an adjacent bigger number: astir 200% growth.

Heading into the Thanksgiving holiday, Wall Street volition beryllium intimately scrutinizing the institution that's been astatine the bosom of this year's artificial quality boom.

Nvidia's banal terms has ballooned 237% successful 2023, acold outpacing immoderate different subordinate of the S&P 500. Its marketplace headdress present sits astatine $1.2 trillion, good supra Meta oregon Tesla. Any denotation connected the net telephone that generative AI enthusiasm is cooling, oregon that immoderate large customers are moving implicit to AMD's processors, oregon that China restrictions are having a detrimental effect connected the concern could spell occupation for a banal that's been connected specified a tear.

"Expectations are precocious starring into NVDA's FQ3'24 net telephone connected Nov-21," Bank of America analysts wrote successful a study past week. They person a bargain standing connected the banal and said they "expect a beat/raise."

However, they flagged China restrictions and competitory concerns arsenic 2 issues that volition seizure capitalist attention. In particular, the emergence of AMD successful the generative AI marketplace presents a caller dynamic for Nvidia, which has mostly had the AI graphics processing portion (GPU) marketplace to itself.

AMD CEO Lisa Su said precocious past period that the institution expects GPU gross of astir $400 cardinal during the 4th quarter, and much than $2 cardinal successful 2024. The institution said successful June that the MI300X, its most-advanced GPU for AI, would commencement shipping to immoderate customers this year.

Nvidia is inactive by acold the marketplace person successful GPUs for AI, but precocious prices are an issue.

"NVDA needs to forcefully antagonistic the communicative its products are excessively costly for generative AI inference," the Bank of America analysts wrote.

Last week, Nvidia unveiled the H200, a GPU designed for grooming and deploying the kinds of AI models that are powering the generative AI explosion, allowing companies to make smarter chatbots and person elemental substance into originative graphical designs.

The caller GPU is an upgrade from the H100, the spot OpenAI utilized to bid its most-advanced ample connection model, GPT-4 Turbo. H100 chips outgo betwixt $25,000 and $40,000, according to an estimation from Raymond James, and thousands of them moving unneurotic are needed to make the biggest models successful a process called "training."

The H100 chips are portion of Nvidia's information halfway group, which saw gross successful the 2nd 4th surge 171% to $10.32 billion. That accounted for astir three-quarters of Nvidia's full revenue.

For the 3rd quarter, analysts expect information halfway maturation to astir quadruple to $13.02 cardinal from $3.83 cardinal a twelvemonth earlier, according to FactSet. Total gross is expected to emergence 172% to $16.2 billion, according to analysts surveyed by LSEG, formerly Refinitiv.

Based connected existent estimates, maturation volition highest successful the 4th fourth astatine astir 195%, LSEG estimates show. Expansion volition stay robust passim 2024 but is expected to decelerate each 4th of the year.

Executives tin expect to tract questions connected the net telephone related to the massive shakeup astatine OpenAI, the creator of the chatbot ChatGPT, which was a large catalyst of Nvidia's maturation this year. On Friday, OpenAI's committee announced the abrupt firing of CEO Sam Altman implicit disputes astir the company's velocity of merchandise improvement and wherever it's focusing its efforts.

OpenAI is simply a large purchaser of Nvidia's GPUs, arsenic is Microsoft, OpenAI's apical backer. Following a chaotic weekend, OpenAI connected Sunday nighttime said erstwhile Twitch CEO Emmett Shear would beryllium starring the institution connected an interim basis, and soon aft that Microsoft CEO Satya Nadella said Altman and ousted OpenAI Chairman Greg Brockman would beryllium joining to lead a caller precocious AI probe team.

Nvidia investors person truthful acold brushed disconnected China-related concerns contempt the imaginable value to the company's business. The H100 and A100 AI chips were the archetypal to beryllium deed by new U.S. restrictions past twelvemonth that aimed to curb income to China. Nvidia said successful September 2022 that the U.S. authorities would still allow it to make the H100 successful China, which accounts for 20% to 25% of its information halfway business.

The institution has reportedly recovered a mode to support selling into the world's second-biggest system portion keeping compliant with U.S. rules. The institution is acceptable to present 3 caller chips, based connected the H100, to Chinese manufacturers, Chinese fiscal media Cailian Press reported past week, citing sources.

Nvidia has historically avoided providing yearly guidance, preferring to look up lone to the adjacent quarter. But fixed however overmuch wealth investors person poured into the institution this twelvemonth and however small other determination is for them to travel this week, they'll beryllium listening intimately to CEO Jensen Huang's code connected the league telephone for immoderate motion that the buzz successful generative AI whitethorn beryllium wearing off.

WATCH: EMJ's Eric Jackson expects a bully study from Nvidia