Okay, so is anyone not developing their own AI chip?
Artificial intelligence has been driving a veritable blitzkrieg of buzz as analysts forecast a brave — and perhaps spooky — new future, while Big Tech companies drop a ton of simoleons to get their slice of tomorrow.
Nvidia is the biggest name on the block when it comes to artificial intelligence.
The chipmaker, which has gained nearly $1 trillion in market value this year, launched its new Blackwell GPU architecture, which performs tasks at more than twice the speed of its current offerings.
Bank of America Securities analyst Justin Post noted that Facebook's parent Meta Platforms, Amazon, and Alphabet, the people who own Google, have all announced advancements around in-house chips.
"We aren't suggesting these chips are competitive with Nvidia's AI capabilities, but [cloud-services] companies are featuring their chips to customers as a cost-saving option," the analyst said in a research note.
Post said that with more than $100 billion in capital spending from Big 3 internet companies — excluding Amazon retail capex — and semiconductor gross margins ranging from 45% to 75%, “we think there is an opportunity for the internet sector to benefit from savings from in-house processors optimized for specific high-volume workflows.”
Analyst sees 'crucial role' for AI
Meta Platforms (META) recently released its Meta Training and Inference Accelerator chip, designed to support Meta's Gen AI workflow, recommendation systems, and AI research.
Nick Clegg, Meta's president of global affairs, said the company expected to release its latest AI model, Llama3, next month. The latest Llama will be part of a family of models with different capabilities this year.
Related: Analysts weigh in on Google-parent Alphabet’s stock after cloud event
Meta plans to release two of the smaller Llama3 models soon and the multimodal version in the summer, according to The Information.
Post, who has a buy rating on Meta with a $550 price target, said AI and machine-learning integrations are playing a crucial role in boosting advertising spending for the social-media giant in 2024.
“We believe [artificial intelligence/machine learning] integrations are playing a crucial role in boosting usage and ad spend in 2024, aiding in five key areas: 1) growing content engagement, 2) aiding ad creative generation and enhancement, 3) campaign optimization (Advantage+) 4) enhanced ad measurement, and 5) back-end campaign cost savings for advertisers,” the analyst wrote.
"On the risk side, Meta has less risk of disruption of legacy usage than Google, with potential advantages over most online ad companies who don't have comparable compute infrastructure, [large language models] or user data," Post said.
The analyst said he sensed some hope among investors "that the returns on AI investment are compelling enough to potentially pull some investment from the metaverse."
Meta is also spending billions on buying Nvidia and other AI chips.
Chief Executive Mark Zuckerberg said in January that the company’s “future road map” for AI requires it to build a “massive compute infrastructure.”
Zuckerberg said that by the end of 2024, the infrastructure will include 350,000 H100 graphics cards from Nvidia. Combined with other suppliers, Meta plans to accumulate the equivalent of 600,000 H100 chips this year, he said.
Analyst optimistic about Meta's prospects
In February, Meta, a member of the so-called Magnificent 7 tech stocks, posted adjusted fourth-quarter earnings per share of $5.33 on revenue of $40.1 billion.
Analysts had expected adjusted EPS of $4.94 and revenue of $39 billion.
Related: Analyst revamps Nvidia stock price target as Mag 7 earnings loom
A year earlier, Meta reported earnings of $1.76 a share on revenue of $32.2 billion, meaning that year over year, sales grew 25% and earnings tripled.
The company is scheduled to report first-quarter earnings on April 24.
Piper Sandler analyst Thomas Champion is feeling optimistic about the company's financial prospects, citing potential growth in revenue and earnings before interest, taxes, depreciation, and amortization.
Champion increased his price target for Meta shares to $600 from $525, while maintaining an overweight rating.
The analyst projected a 25% year-over-year increase in first-quarter revenue to $35.7 billion and $19.8 billion in EBITDA, based on recent checks.
More AI Stocks:
- Analyst unveils eye-popping Palantir stock price target after Oracle deal
- Veteran analyst delivers blunt warning about Nvidia's stock
- Analysts revamp Microsoft stock price target amid OpenAI reports
These checks were boosted by the performance of Meta’s Advantage+ portfolio, which aims to help advertisers leverage artificial intelligence to automate their advertising campaigns. and, in particular, the Shopping feature.
Champion said revenue could surpass $36 billion.
He also sees impressions remaining consistent with the fourth quarter, showing 21% year-over-year growth, with pricing staying relatively stable.
For the second quarter, Champion set a revenue target for Meta at between $36 billion and $38.5 billion, a 13% to 20% increase from a year ago.
TheStreet Pro’s Bob Lang said that after the strong earnings move in February, Meta has continued its bullish move on pretty strong turnover and improved indicators.
"We are holding off adding Meta to the portfolio for now, preferring a pullback to start a position," he said. "We believe there is some really strong upside ahead but will wait for a better entry point."
Related: Veteran fund manager picks favorite stocks for 2024
