Arm is now a merchant silicon vendor. Not just an IP company.
@renehaas237’s rationale: Agentic AI era needs more CPUs in the same GW power envelope. Arm’s legacy is power-efficient CPUs
@Arm $ARM
In an exclusive first look at our new Austin lab, @CNBC highlights Arm’s entry into production silicon with the Arm AGI CPU.
In the video, they explore how the Arm AGI CPU is built for AI inference in data centers and the growing role of CPUs in scaling AI infrastructure.https://t.co/tHVAw7OROM
#ArmEverywhere
I wasn’t able to attend the $ARM event today, but here are my quick thoughts:
WHY IT MATTERS:
∙Business model transformation. IP + CSS + finished silicon changes the revenue profile entirely. Arm goes from royalty taker to chip seller.
∙The agentic AI thesis is real. Arm’s math: agentic workloads drive 4x more CPU cores per GW (30M to 120M). CPUs are the bottleneck in agent-driven data centers. I said at Davos the AI bottleneck is the full infrastructure stack. This fits.
∙@Meta’s involvement is the headline. As I told CNBC today, if Arm captures even 5% of Meta’s $115-$135B annual capex, that is a top-line changer
∙Clear market gap. You need 1,000 engineers and $500M to build a custom Arm server CPU. Most companies can’t do that. Arm fills the white space between hyperscaler custom silicon (Graviton, Axion, Cobalt) and off-the-shelf x86.
∙The x86 critique was sharp. Awad called out legacy overhead and multi-threading inefficiencies. Showed benchmarks with SMT on and off for x86. The x86 numbers got worse. Direct shot at Intel and AMD
∙Huge financial upside from a profit dollar standpoint.
Today.
A defining moment for AI compute.
In a few hours, Arm CEO Rene Haas takes the stage at #ArmEverywhere.
Don’t miss it. https://t.co/sagxWAOMKH
📅 10:00am PDT
🎥 Live on X
Money right here from @awscloud when it comes to custom silicon.
Not just first for Amazon, but first of any hyperscaler to have the XPU, CPU, and DPU all custom on the same board.
I expect Google to do the same.
Also, great showcase for @Arm both in CPU and DPU!
Arm just contributed its Foundation Chiplet System Architecture to OCP. I talked with Mohamed Awad, GM of Arm Infrastructure, about:
• Power-limited AI datacenters
• Arm IP → CSS → Total Design → FCSA
• What's standardized and what's open
https://t.co/WuHjKVhHzj
The money isn’t the hard part. It’s speed. The market is moving so fast. Helping focus investment on what really matters has always been the mission of OCP and of @Arm#OCP
🗓️ Happening tomorrow at #OCPSummit25!
Mohamed Awad shares how energy-efficient compute, chiplet-based design, and open collaboration are transforming AI and cloud infrastructure — and shaping the next era of the data center.
📍 Tuesday | 09:05am https://t.co/slj0y1N1og
In the age of AI, power efficiency isn’t just a competitive advantage. It’s a baseline requirement.
🆕 That’s why close to 50% of the compute shipped to top hyperscalers in 2025 will be Arm-based.
This is what reshaping the cloud landscape looks like: https://t.co/wKYr2PBkaq
Here's to a great #GTC25! Did you catch these Arm-based announcements during NVIDIA CEO, Jensen Huang's keynote?
➡️ Blackwell Ultra, using the GB300 NVL72, which connects 72 Blackwell Ultra GPUs and 36 Arm Neoverse-based Grace CPUs
➡️ Vera Rubin - as part of NVIDIA's future roadmap, packing 88 Arm-based cores
➡️ New desktop systems coming from leading vendors based on the new Arm-based DGX Spark personal AI supercomputer
➡️ GM's future self-driving fleet using Arm-based NVIDIA Drive AGX technology
➡️ A sneak peek of more robotics on Jetson!
As AI innovation continues at pace, the Arm compute platform is providing the foundation for the next generation of AI applications from the edge to the cloud, covering PC, autonomous vehicles, data centers and more. We're proud to partner with NVIDIA to make this a reality.
Announcing The Stargate Project
The Stargate Project is a new company which intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. We will begin deploying $100 billion immediately. This infrastructure will secure American leadership in AI, create hundreds of thousands of American jobs, and generate massive economic benefit for the entire world. This project will not only support the re-industrialization of the United States but also provide a strategic capability to protect the national security of America and its allies.
The initial equity funders in Stargate are SoftBank, OpenAI, Oracle, and MGX. SoftBank and OpenAI are the lead partners for Stargate, with SoftBank having financial responsibility and OpenAI having operational responsibility. Masayoshi Son will be the chairman.
Arm, Microsoft, NVIDIA, Oracle, and OpenAI are the key initial technology partners. The buildout is currently underway, starting in Texas, and we are evaluating potential sites across the country for more campuses as we finalize definitive agreements.
As part of Stargate, Oracle, NVIDIA, and OpenAI will closely collaborate to build and operate this computing system. This builds on a deep collaboration between OpenAI and NVIDIA going back to 2016 and a newer partnership between OpenAI and Oracle.
This also builds on the existing OpenAI partnership with Microsoft. OpenAI will continue to increase its consumption of Azure as OpenAI continues its work with Microsoft with this additional compute to train leading models and deliver great products and services.
All of us look forward to continuing to build and develop AI—and in particular AGI—for the benefit of all of humanity. We believe that this new step is critical on the path, and will enable creative people to figure out how to use AI to elevate humanity.
"The flexibility of Axion is essential if we want to more quickly and broadly meet the infrastructure needs of our customers and partners — especially in the fast-moving age of AI." - Partha Ranganathan & Mark Lohmeyer @GoogleCloud - https://t.co/hS6xt6bdQL
🤝 We're taking our partnership with @GoogleCloud to the next level.
Axion processors, built #onArm Neoverse V2, are the newest example of how we're working together to optimize for workload performance and efficiency.
@AwadMo shares the details. 👇
By building #onArm Neoverse, @GoogleCloud's Axion processors are well-suited for the age of AI, benefitting from:
⚡ Higher performance
🔋 Lower power consumption
⬆️ Greater scalability
@AwadMo shares why. 👇
Full interview: https://t.co/k8M0bWbWpB
OK - looking for more details on @Google Axion? Check out the below. I try to cover:
- the reach of @arm in the cloud
- details on Axion (what's available)
- impact to @AmpereComputing
- longer term impact
https://t.co/hxWj58bO6T
📢 Custom Arm-based silicon from @GoogleCloud is here!
Axion, built on Armv9 Neoverse V2, will bring performance & energy efficiency gains to AI inferencing & general-purpose workloads, like web and app servers.
Congrats to the team! 👏 https://t.co/oQha28uEri #GoogleCloudNext
📸 Leaders from technology, industry and academic institutions joined us at #MWC24 to celebrate the new AI-RAN Alliance.
We're looking forward to our extensive collaboration, sharing expertise to advance research and innovation in radio access networks.
🆕 Introducing two new Arm Neoverse Compute Subsystems, Neoverse CSS V3 & CSS N3.
Built on our 3rd generation of Neoverse technology, V3 and N3 bring generation-on-generation performance improvements to drive custom solutions and accelerate AI workloads: https://t.co/TgYYP0XROl
Congratulations to the @Microsoft team that believed in this vision and to the entire team at @Arm who moved mountains to make this happen! The Azure Cobalt 100, Microsoft's first custom silicon for the cloud built #onArmNeoverse CSS, looks amazing! https://t.co/sPr1QiM9gd