Arm Holdings is making a significant strategic shift by entering the chipmaking market, moving beyond its traditional licensing model. The company is now designing and selling its own chips, starting with the AGI CPU, also referred to as "Arm Advantage," specifically for AI workloads in data centers. This new chip, featuring up to 136 cores, aims for 30% greater power efficiency than competitors and is designed for advanced AI agent tasks, potentially lowering data center costs. Taiwan Semiconductor Manufacturing Company (TSMC) will produce the AGI CPU using 3-nanometer technology, with volume production expected in the second half of 2024.
Early supporters and partners for Arm's new venture include Meta Platforms and OpenAI. Additionally, Arm is collaborating with Amazon Web Services to create custom chips for their data centers. Arm anticipates this new chip business will generate billions in annual revenue, eventually surpassing its current intellectual property sales, with a projection of around $25 billion in total sales within five years. CEO Rene Haas has called this a pivotal moment for the company, indicating a major change in its business strategy to capture more value in the growing AI market.
In the broader AI investment landscape, investor Michael Burry has criticized AI companies like Nvidia and Palantir Technologies, alleging they inflate profits by extending depreciation schedules for GPUs and servers. He claims this accounting practice understates expenses by billions. However, Nvidia CEO Jensen Huang and CoreWeave CEO Michael Intrator have challenged Burry's assertions. Intrator noted that customer contracts typically last five years and that older GPUs, such as the A100 and H100, retain up to 95% of their value due to emerging use cases, contradicting the obsolescence theory. Despite these defenses, Burry's critique, alongside general market concerns, has contributed to selling pressure on AI stocks.
Analysts are also weighing in on AI investments. Bernstein recommends investing in Nvidia and Broadcom, while advising caution on AMD's AI chip stocks, citing strong AI demand causing memory shortages and increased costs. William Blair has introduced a new four-criteria system to evaluate AI stocks, focusing on execution, AI defensibility, pricing model, and organic revenue acceleration. Using this 20-point scale, companies like Everpure (formerly Pure Storage) and DigitalOcean scored highly with 19 out of 20. BlackRock CEO Larry Fink has also shared advice for investors, emphasizing the need for companies and investors to adapt to rapid AI advancements.
Several companies are demonstrating how they are leveraging AI. Xpeng reported its first profit, largely driven by a 121% year-on-year increase in revenue from high-margin services and other solutions, including technical R&D. GETB saw an 8% increase in Annual Recurring Revenue (ARR) to £22.6 million, boosted by new AI features in its SmartVault product, though increased AI development investment led to a decrease in Adjusted EBITDA. Zillow Group plans to highlight its AI-native housing platform strategy, aiming for mid-teens total revenue growth and a 25% net income margin by 2026. CleanSpark is expanding its Texas power capacity by 300 megawatts to support high-density computing for AI, diversifying beyond Bitcoin mining. Furthermore, Erste Group upgraded Meta Platforms stock to 'Buy,' raising its price target to $550 from $470, anticipating a major increase in Meta's AI spending in 2026.
Key Takeaways
- Arm Holdings is strategically shifting to design and sell its own AI data center chips, the AGI CPU, moving beyond its traditional licensing model.
- Arm expects its new chip business to generate billions in annual revenue, aiming for total sales of approximately $25 billion within five years.
- Meta Platforms, OpenAI, and Amazon Web Services are key early partners and collaborators for Arm's new AGI CPU, with TSMC handling production.
- Investor Michael Burry claims AI companies like Nvidia and Palantir inflate profits by extending GPU depreciation schedules, understating expenses.
- Nvidia and CoreWeave dispute Burry's claims, stating customer contracts are long-term and older GPUs like A100 and H100 retain up to 95% of their value.
- Bernstein analysts recommend investing in Nvidia and Broadcom due to strong AI demand but advise caution on AMD's AI chip stocks.
- Xpeng achieved its first profit, primarily driven by a 121% year-on-year increase in high-margin AI-related services and solutions revenue.
- GETB reported an 8% increase in Annual Recurring Revenue to £22.6 million, attributed to new AI features, despite increased AI investment impacting EBITDA.
- Erste Group upgraded Meta Platforms stock to 'Buy' with a $550 price target, citing significant anticipated AI investments in 2026.
- William Blair introduced a new four-criteria system for evaluating AI stocks, identifying Everpure and DigitalOcean as top performers.
Arm enters chipmaking with new AI processor AGI CPU
Arm Holdings is changing its business model by designing and selling its own chips, starting with the AGI CPU for AI workloads in data centers. This move aims to capture more value in the growing AI market, moving beyond its traditional licensing of chip designs. The AGI CPU, featuring up to 136 cores, is designed for efficient AI agent tasks and could significantly lower data center costs. Meta and OpenAI are among the early supporters, with Taiwan Semiconductor Manufacturing Company handling production. Arm plans for volume production in the second half of the year.
Arm launches new AI chip, expects billions in revenue
Arm Holdings announced a new AI data center chip called 'Arm Advantage' that aims to be 30% more power-efficient than competitors. This marks a significant strategic shift for Arm, which previously only licensed chip designs. The company expects this new chip to generate billions in annual revenue. Arm is also collaborating with Amazon Web Services to create custom chips for their data centers. The new chip is expected to be available in the second half of 2024.
Arm unveils new AI chip, expects billions in revenue
Arm Holdings has launched a new artificial intelligence data center chip named AGI CPU, which it expects to generate billions in revenue and signifies a major strategic change for the company. This chip is designed for advanced AI tasks that require minimal oversight, moving beyond typical chatbot functions. CEO Rene Haas called this a pivotal moment for Arm. Meta Platforms and OpenAI are early partners, and Taiwan Semiconductor Manufacturing Co. will produce the chip using advanced 3-nanometer technology. Volume production is planned for the second half of the year.
Arm to sell its own chips for AI revenue
Arm Holdings is entering the chip manufacturing market by selling its own chips for the first time, starting with the AGI CPU. Meta Platforms is the first major customer for this new chip, which features up to 136 cores. Arm projects that revenue from its new chip business will surpass its current intellectual property sales, aiming for total sales of around $25 billion within five years. This strategic shift allows Arm to capture more profit by producing the chips itself, rather than just licensing designs. The company anticipates significant growth and earnings increases in the coming years.
Michael Burry's AI stock claims challenged by Nvidia and CoreWeave
Investor Michael Burry claims AI companies like Nvidia and CoreWeave inflate profits by extending depreciation schedules for GPUs. He argues that chips designed for short lifespans are treated as if they last much longer, understating expenses. Nvidia and CoreWeave dispute this, stating that customer contracts are long-term and older GPUs retain significant value due to new applications. CoreWeave's CEO highlighted that five-year contracts and strong resale values for GPUs like A100 and H100 contradict Burry's obsolescence theory. Despite these defenses, AI stocks have seen significant drops, partly due to Burry's critique and broader market concerns.
Michael Burry's AI stock claims challenged by Nvidia and CoreWeave
Michael Burry has criticized AI companies, including Nvidia and Palantir Technologies, for allegedly inflating profits by extending depreciation schedules for their GPUs and servers. He claims this accounting practice understates expenses by billions. Nvidia CEO Jensen Huang and CoreWeave CEO Michael Intrator have pushed back against Burry's claims. Intrator stated that customer contracts typically last five years and that older GPUs like the A100 and H100 retain 95% of their value due to emerging use cases. While Burry's critique has amplified selling pressure on AI stocks, broader concerns about AI spending and valuations also contribute to the sector's decline.
Xpeng achieves first profit driven by AI services
Xpeng reported its first profit, largely due to a significant increase in revenue from services and other solutions, which grew by over 121% year-on-year. This segment, which includes technical R&D services and carbon credit trading, has a much higher profit margin than Xpeng's vehicle sales. While the vehicle business also saw improved gross margins, the profit milestone was primarily driven by this shift in revenue mix towards higher-margin technology and software services. Xpeng is positioning itself as a technology company beyond just car manufacturing, with strong cash reserves to invest in areas like smart driving and robotics.
GETB reports revenue growth with AI investment
GETB announced an 8% increase in Annual Recurring Revenue (ARR) to £22.6 million, driven by strong performance in the US tax market with its SmartVault product and new AI features. However, Adjusted EBITDA decreased to £0.3 million due to increased investment in AI development and migration automation. While Workiro's ARR saw a slight decline, the overall growth in ARR highlights the company's focus on AI-powered solutions.
William Blair rates AI stocks using new criteria
Analysts at William Blair have developed a new four-criteria system to evaluate AI stocks: ability to execute, AI defensibility, pricing model, and likelihood of organic revenue acceleration. Using a 20-point scale, they identified top performers like Everpure (formerly Pure Storage) and DigitalOcean, both scoring 19 out of 20. Other companies like Snowflake and MongoDB also received strong ratings. The system aims to identify companies best positioned for success in the evolving AI landscape, focusing on factors beyond just hype.
Zillow highlights AI platform strategy at investor summit
Zillow Group is set to discuss its AI-native housing platform strategy at an upcoming AI Summit for Investors. CEO Jeremy Wacksman emphasized that Zillow's platform simplifies the moving process by integrating proprietary data and AI into core workflows. The company plans to leverage its position as the most visited real estate site in the US to enhance consumer experiences and professional tools. Zillow also reaffirmed its financial outlook for 2026, targeting mid-teens total revenue growth and a 25% net income margin at mid-cycle.
BlackRock CEO Larry Fink advises on AI investing
BlackRock CEO Larry Fink has issued his annual letter to shareholders, focusing on advice for investors navigating the era of artificial intelligence. The letter, titled 'A Letter To Our Shareholders,' highlights the transformative potential of AI and its impact on the investment world. Fink stresses the importance for companies and investors to adapt to rapid AI advancements. The core message encourages strategic considerations for investing in a landscape increasingly shaped by artificial intelligence.
Bernstein favors Nvidia and Broadcom, cautious on AMD AI chips
Bernstein analysts recommend investing in Nvidia and Broadcom while advising caution on AMD's AI chip stocks. They note that strong AI demand is causing memory shortages and increasing costs, impacting various tech sectors. Despite these challenges, Bernstein believes demand for Nvidia and Broadcom remains strong, with potential for further stock growth. For AMD, Bernstein seeks more evidence of genuine customer demand beyond supply constraints before recommending the stock.
CleanSpark expands Texas power capacity for AI compute
CleanSpark has acquired a second Texas site, adding 300 megawatts of power capacity approved by ERCOT. This expansion supports the company's move into high-density computing for AI and high-performance workloads, potentially diversifying beyond Bitcoin mining. The new capacity reinforces CleanSpark's strategy of using power-dense infrastructure for both mining and AI compute. This expansion increases the importance of power market conditions and energy policies for the company's future growth and execution.
Erste Group upgrades Meta stock, citing AI investments
Erste Group has upgraded Meta Platforms (META) stock to 'Buy' from 'Hold,' citing the company's significant AI investments and pricing strategies. Analyst Hans Engel anticipates a major increase in Meta's spending in 2026 to achieve its AI objectives. Based on this outlook, Engel raised his price target for META stock to $550 from $470, highlighting Meta as a strong candidate for high profit margin stocks.
Sources
- Arm pivots to chipmaking with AI chip launch: what this could mean for ARM
- Arm unveils new AI chip, expects it to add billions in annual revenue
- Arm unveils new AI chip, expects it to add billions in annual revenue
- Arm to Sell Its Own Chips for First Time in Bid for AI Revenue
- Was Michael Burry Right About AI Stocks for the Wrong Reason?
- Was Michael Burry Right About AI Stocks for the Wrong Reason?
- Xpeng posts first profit as business shifts toward AI-driven services
- GETB: ARR up 8% to £22.6m, EBITDA down to £0.3m, driven by SmartVault growth and AI investment
- Analysts' new AI stock rating system reveals these winners and losers
- Zillow Group to highlight AI-native housing platform strategy at AI Summit for Investors
- Larry Fink’s Advice in the Era of Artificial Intelligence
- ‘Own Nvidia, Broadcom,’ But Avoid AMD Stock – Bernstein Weighs In on AI Chip Stocks
- Texas Power Expansion For AI Compute Could Be A Game Changer For CleanSpark (CLSK)
- Erste Group upgrades Meta (META) to buy, highlights the company’s AI investments
Comments
Please log in to post a comment.