SpaceXAI will be adopting Nvidia's Vera Rubin platform for AI supercomputing to power its Grok AI, using what the company called the "first CPU built for AI agents" to scale up its AI systems, as per a blog post from Nvidia on Monday.
The private space exploration technology company will also be using the Vera Rubin platform for orbital computing in space satellites such as its Starmind AI satellite.
"SpaceXAI's planned first-generation Starmind AI satellite will be based on the optimized NVIDIA Vera Rubin NVL72 rack-scale system, extending the same accelerated computing architecture powering next-generation AI factories on Earth into space," the post said.
ALSO READ: Infosys, Wipro ADRs Tick Higher As AI Trade Turns Sour. What's The Nvidia Factor Here
The Elon Musk founded company will combine Nvidia's accelerated computing, NVLink interconnect technology, Spectrum-X Ethernet networking and BlueField data processing and software to create an integrated architecture which is expected to provide high performance and energy efficiency at scale while also reducing cost per token.
“Vera gives us the CPU performance and memory bandwidth to run enormous amounts of orchestration, code and data processing while keeping GPUs doing what they do best,” Mike Nicolls, president of SpaceXAI said.
“That means higher-performance AI agents and more useful work from every watt of compute," he added.
Nvidia's Vera is geared around managing and speeding up CPU-intensive work with regards to model inference (processing new data to make a decision or generate output). It covers a wide variety of proceses from tool use and code execution to data processing, orchestration and simulation.
It uses 88 NVIDIA-designed Olympus cores, Nvidia Spatial Multithreading technology and high-bandwidth LPDDR5X memory, which can provide up to 1.2TB/s of bandwidth.
Accoridng to the company, this facilitates 1.8 times faster task completion in comparison to x86 CPUs with regards to agentic AI, reinforcement learning and processing data.
“Agentic AI requires a new kind of computing system, one built not only to generate answers, but to take action,” Ian Buck, vice president of hyperscale and high-performance computing at Nvidia.
“Vera gives AI agents the CPU performance to act in real time: executing code, processing data and coordinating complex tasks. SpaceXAI is taking this architecture from massive AI factories to the next frontier of computing in orbit," Buck added.
Nvidia's Groq 3 LPX interactive AI inference accelerator, an extension of the Vera Rubin platform, entered full production as per another blog post from the company, hitting 3,400 output tokens per second.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.