Amazon's stock climbed nearly 3% after Meta agreed to use Amazon's Graviton chips for at least three years, making Meta a top-five Graviton customer. The deal involves hundreds of thousands of Graviton CPUs, which are ARM-based processors designed for AI tasks like real-time reasoning and code generation. Meta's head of infrastructure said the chips provide the performance and efficiency needed at their scale, especially for agentic AI workloads. This move signals a shift from Nvidia chips to alternatives like Graviton and Trainium for cost efficiency, as noted by Amazon CEO Andy Jassy.
NVIDIA's stock has lagged behind peers like Marvell, Micron, and AMD, rising only 7% year-to-date as investors diversify into custom silicon and memory stocks. Meanwhile, NVIDIA has partnered with Oklo Inc. and Los Alamos National Laboratory to develop nuclear-powered AI factories, aiming to make AI more sustainable. The company's market cap remains over $500 billion, but analysts predict a base-case price of $227 over the next year.
Billionaire investor Stanley Druckenmiller sold Meta and added to his positions in Alphabet and Amazon, believing these companies are turning AI into profitable businesses. He expressed concern about Meta's heavy AI spending without clear returns. Alphabet benefits from its own AI chips and growing cloud business, while Amazon's AWS and in-house chips are seen as key to its AI growth.
ServiceNow's stock fell 18% after a revenue miss in Q4, with some investors worried its AI platform isn't meeting expectations. However, analysts argue the drop may be overblown as revenue growth remains steady. Qualcomm is positioned for edge AI growth, with its Snapdragon processors used in phones, cars, and robots, and automotive revenue reaching $1.1 billion in the latest quarter. The company authorized a $20 billion share buyback, signaling management sees the stock as undervalued.
Thinking Machines Lab (TML) gained talent from Meta, including Weiyao Wang who left after eight years. TML recently signed a deal with Google Cloud and has partnerships with Nvidia. The startup's CTO, Soumith Chintala, co-founded PyTorch and is a former Meta employee. TML's headcount is now around 140, and Meta has also poached some TML employees, showing a two-way talent flow.
AI is being used to create sophisticated financial scams and impersonations, and can produce errors called 'AI hallucinations.' Experts advise verifying AI-generated financial or tax information with human professionals, as taxpayers are responsible for any mistakes. Human oversight remains the best defense against AI errors.
Key Takeaways
- Meta signed a multi-year deal to use hundreds of thousands of AWS Graviton CPUs for AI workloads, making Meta a top-five Graviton customer.
- Amazon's stock rose nearly 3% after the Meta deal, highlighting AWS's growing role in AI infrastructure.
- NVIDIA stock is up only 7% year-to-date, lagging behind peers like Marvell, Micron, and AMD as investors diversify.
- NVIDIA partnered with Oklo Inc. and Los Alamos National Laboratory to develop nuclear-powered AI factories.
- Billionaire Stanley Druckenmiller sold Meta and bought Alphabet and Amazon, citing concerns about Meta's AI spending.
- ServiceNow's stock dropped 18% after a Q4 revenue miss, though analysts say the drop may be overblown.
- Qualcomm is positioned for edge AI growth with Snapdragon processors and authorized a $20 billion share buyback.
- Thinking Machines Lab gained talent from Meta, including Weiyao Wang, and has partnerships with Google Cloud and Nvidia.
- AI is being used for financial scams and can produce errors; experts advise verifying AI-generated information with humans.
- SanDisk stock surged 295% year-to-date due to AI-driven demand for memory chips, with Q1 earnings expected on April 30.
Meta deal boosts Amazon custom chips for AI
Amazon's stock rose nearly 3% after Meta agreed to use Amazon's Graviton chips for at least three years. This makes Meta a top-five Graviton customer. The deal highlights Amazon Web Services' growing role in providing AI infrastructure. Amazon CEO Andy Jassy noted a shift from Nvidia chips to alternatives like Graviton and Trainium for cost efficiency. Meta benefits by reducing compute costs for its AI workloads across Facebook and Instagram.
Meta signs deal for millions of Amazon AI CPUs
Meta has signed a deal to use millions of AWS Graviton CPUs, which are ARM-based processors designed for AI tasks. These CPUs handle workloads like real-time reasoning and code generation, which are different from the GPUs used for training AI models. The deal brings more of Meta's business to AWS instead of competitors like Google Cloud. Amazon's Graviton chips compete with Nvidia's new Vera CPU for AI agentic workloads.
Meta expands use of AWS Graviton for AI workloads
Meta is significantly expanding its use of AWS Graviton processors to power its agentic AI workloads. These processors handle complex tasks like code generation and workflow orchestration. The latest Graviton5 chips use advanced 3-nanometer technology for better performance and energy efficiency. Meta's head of infrastructure said the chips provide the performance and efficiency needed at their scale. The deal signals a new era in AI infrastructure development.
Meta to use hundreds of thousands of AWS Graviton chips
Meta will use hundreds of thousands of Graviton chips from Amazon Web Services in a deal lasting at least three years. This follows Meta's $48 billion in AI infrastructure commitments with CoreWeave and Nebius. Graviton chips are CPUs, not GPUs, and are designed for AI workloads like pre-training models. Meta's head of infrastructure said the chips offer the performance and efficiency needed for agentic AI. Meta has used Nvidia GPUs from AWS since 2017.
NVIDIA stock lags as other semiconductor stocks rally
NVIDIA stock is up only 7% year-to-date while peers like Marvell, Micron, and AMD have surged. Investors are diversifying their AI investments into memory, custom silicon, and networking stocks. NVIDIA's large market cap and China export controls are headwinds. The stock would need to rise 50% to reach $300, which is not expected soon. Analysts predict a base-case price of $227 for NVIDIA over the next year.
SanDisk stock surges 295% as AI drives memory demand
SanDisk stock has risen 295% year-to-date due to AI-driven demand for memory chips. The company reports Q1 earnings on April 30, with Wall Street expecting huge growth. SanDisk spun off from Western Digital in February and has beaten earnings estimates for four straight quarters. AI infrastructure spending is boosting demand for enterprise SSDs and NAND flash. The stock is near its analyst target, so future gains depend on strong results and guidance.
Billionaire Druckenmiller sells Meta, buys Alphabet and Amazon
Billionaire investor Stanley Druckenmiller sold Meta Platforms and added to his positions in Alphabet and Amazon. He believes these companies are turning AI into profitable businesses. Druckenmiller is concerned about Meta's heavy spending on AI without clear returns. Alphabet benefits from its own AI chips and growing cloud business. Amazon's AWS and in-house chips are seen as key to its AI growth.
ServiceNow stock drops 18 percent on AI concerns
ServiceNow's stock fell 18% after a revenue miss in its Q4 earnings report. Some investors worry that its AI platform is not meeting expectations. However, analysts argue the drop may be overblown because revenue growth remains steady. They believe ServiceNow's AI platform is gaining traction and has strong long-term prospects.
NVIDIA partners with Oklo and LANL for nuclear AI factories
NVIDIA has partnered with Oklo Inc. and Los Alamos National Laboratory to develop nuclear-powered AI factories. The collaboration combines NVIDIA's AI and computing technology with Oklo's advanced nuclear reactors. This aims to make AI more sustainable and efficient. NVIDIA's stock has performed well, with a market cap over $500 billion and strong returns over the past five years.
Thinking Machines Lab gains talent from Meta
Weiyao Wang left Meta after eight years to join Thinking Machines Lab (TML). TML recently signed a deal with Google Cloud and has partnerships with Nvidia. The startup has attracted several former Meta employees, including its CTO Soumith Chintala, who co-founded PyTorch. TML's headcount is now around 140. Meta has also poached some TML employees, showing a two-way talent flow.
Strategist says AI picks and shovels trade still strong
SlateStone Wealth's Kenny Polcari says weakness in AI enabler stocks offers a buying opportunity. He believes recent layoffs in tech reflect hiring excesses, not lower demand for AI. Polcari suggests that beaten-down software names deserve a fresh look from investors.
AI is a useful tool but watch out for fraud and errors
Artificial intelligence is being used to create sophisticated financial scams and impersonations. AI can also produce errors called 'AI hallucinations' that give incorrect information. Experts advise verifying any AI-generated financial or tax information with human professionals. Taxpayers using AI for tax returns should be careful, as they are responsible for any mistakes. Human oversight is the best defense against AI errors.
Qualcomm positioned for edge AI growth
Qualcomm is well positioned for the shift to edge AI, where AI runs on devices instead of in the cloud. Its Snapdragon processors are used in phones, cars, and robots. Automotive revenue reached $1.1 billion in the latest quarter. Qualcomm trades at a lower valuation than peers like Broadcom and Marvell. The company authorized a $20 billion share buyback, signaling management sees the stock as undervalued.
Amazon reveals new massive AI opportunity
Amazon may be building one of the most important AI businesses in big tech. The company's chip story and AWS demand signals point to significant upside. The video discusses Amazon's AI opportunities and how spending risks could affect the stock's future.
Sources
- Amazon custom chips get a boost from Meta, giving the cloud giant another path to win in AI
- In another wild turn for AI chips, Meta signs deal for millions of Amazon AI CPUs
- Meta Bets on AWS Graviton for AI
- Meta will adopt hundreds of thousands of AWS Graviton chips in latest AI infrastructure grab
- Why Isn't NVIDIA Stock at $300 While Other Semiconductor Stocks Rally?
- SanDisk Climbs 295% YTD as AI Fuels Memory Mania: Can Q1 Results Justify the Run?
- Billionaire Stanley Druckenmiller Sells Meta and Buys These 2 Other Magnificent AI Stocks Instead
- ServiceNow's 18% Drop: AI Fears Continue, But May Be Overblown
- NVIDIA (NVDA) Partners With Oklo, LANL to Develop Nuclear-Powered AI Factories
- Meta's loss is Thinking Machines gain
- The AI ‘picks and shovels’ trade still has legs: Strategist
- AI is a wonderful tool, but beware of fraud and errors
- Is Qualcomm The Next AI Stock To Surge?
- This "Magnificent Seven" Stock Just Revealed a New Massive AI Opportunity
Comments
Please log in to post a comment.