AM Intelligen2026-08-25 09:56:46AM Intelligence orders 9,000 Nvidia Vera Rubin systems for deployment in southern IndiaIndian AI infrastructure company AM Intelligence has ordered 9,000 Nvidia Vera Rubin systems and plans to bring them online in southern India next year, according to ChainCatcher. The company’s customer base includes large cloud service providers, AI labs, and institutions developing domestic AI models. The order points to a sizable infrastructure buildout tied to AI computing demand in India, though no financial terms or more specific deployment timeline were disclosed in the source. ChainCatcher’s brief said the systems are scheduled to begin operating next year in the southern part of the country.690
SpaceX2026-08-25 05:35:48SpaceX expands Nvidia partnership for orbital AI data centers, targets first Starmind satellite launch in Q4 2027SpaceX CEO Elon Musk said on Aug. 24 that the company is deepening its partnership with Nvidia to extend AI computing infrastructure into Earth orbit through the Vera Rubin platform. The first batch of Starmind AI satellites is now targeted for launch as early as the fourth quarter of 2027, one year ahead of a previously disclosed 2028 timeline, with a larger rollout planned for 2028. Musk said SpaceX and Nvidia have designed a space-optimized Vera Rubin NVL72 system for launch next year. He also said SpaceX will exclusively use Nvidia chips and adopt the next-generation Vera Rubin platform to support Grok and other AI services. The orbital computing system centers on Nvidia’s Vera CPU, which the company says carries 88 in-house Olympus cores, delivers 1.2 TB/s of memory bandwidth, and completes tasks 1.8 times faster than comparable x86 processors. The article also points to a much broader buildout. Musk previously told investors SpaceX expects to surpass 2 GW of AI compute capacity by the end of 2026 and approach 10 GW by the end of 2027, while Morgan Stanley estimates AI-related capital spending could rise from $53 billion in 2026 to $130 billion in 2027.980
NVIDIA2026-08-25 02:02:15NVIDIA posts first Vera Rubin NVL72 silicon results, showing up to 30x higher agent throughput per megawattNVIDIA has released its first silicon test results for the Vera Rubin NVL72 rack, using DeepSeek-V4-Pro in a coding-focused agent workload rather than a conventional fixed-length inference benchmark. According to the company, Vera Rubin NVL72 delivered up to 30x higher throughput per megawatt than GB300 NVL72 under the AgentX benchmark, while the cost to produce one million tokens fell by as much as 35x. The test framework came from SemiAnalysis and was built to replay full coding sessions with growing context, tool use, and sub-agent generation, reflecting how agent systems behave in production. The company also disclosed that Groq 3 LPX has entered full production as a low-latency inference extension for the Vera Rubin platform. In an Artificial Analysis benchmark, Groq 3 LPX ran Gemma 4 31B with a 100,000-token context window at 3,400 tokens per second, and its peak output in coding workloads reached 6,981 tokens per second. NVIDIA said Vera CPU, a processor built for agent workloads with 88 Olympus cores and 1.2 TB/s of LPDDR5X bandwidth, is also moving into deployment, with SpaceXAI named among early adopters.1160
Nvidia2026-08-25 00:56:52Nvidia Says Groq 3 LPX Racks Enter Full Production, First Deployment Planned for NebiusNvidia senior director Dion Harris said on Monday local time that Groq 3 LPX racks have entered full production and will be deployed at the data center of Nebius, a new cloud provider, with the rollout expected later this year. The move follows Nvidia’s roughly $20 billion technology licensing deal with Groq in December and marks a step toward commercialization of the technology involved. The report also says several core Groq employees have joined Nvidia, while founder Jonathan Ross now serves as Nvidia’s chief software architect. Groq 3 LPX, launched in March, is described as a Vera Rubin platform inference accelerator that integrates 500 megabytes of high-speed SRAM on the chip die to reduce memory bottlenecks. Nvidia said each LPX rack can combine 256 Groq 3 chips and, based on benchmarks it cited, process 3,400 tokens per second. Harris said low-latency chips are not meant to replace GPUs, but to be used where different workloads call for different processors, and cloud providers may price latency-sensitive tiers higher. Nvidia CEO Jensen Huang plans to allocate about one quarter of data-center space for programming applications to Groq chips, with the rest using Vera Rubin systems. Nvidia also expects combined sales of Blackwell and Vera Rubin to reach $1 trillion by 2027.1150
Elon Musk2026-08-24 17:29:50Musk says Nvidia’s space-optimized Vera Rubin platform will reach scale in 2028According to ChainCatcher, Elon Musk said Nvidia’s space-optimized Vera Rubin chip platform is expected to reach meaningful scale in 2028. He also said the Vera Rubin NVL72 system is scheduled to be launched into orbit in the fourth quarter of next year. The update centers on Nvidia’s Vera Rubin platform and its space-focused deployment timeline, with 2028 given as the point when the platform will reach considerable scale. Separately, the NVL72 system was identified as the configuration set for orbital launch, and the target window was specified as Q4 next year. No additional technical specifications, deployment figures, or mission details were disclosed in the source item.1070
Nvidia2026-08-24 15:07:27Nvidia Says SpaceXAI Will Deploy Vera CPU and Extend Vera Rubin to Orbital ComputingNvidia said on Aug. 24 that SpaceXAI will deploy its Vera CPU to speed up next-generation agentic AI applications. Vera is the first CPU built specifically for AI agents, according to Nvidia, and it features 88 in-house Olympus cores, spatial multithreading and LPDDR5X memory with bandwidth of 1.2 TB/s. Nvidia said the chip can deliver up to 1.8 times faster task completion than x86 CPUs in agentic AI, reinforcement learning and data processing workloads. SpaceXAI President Mike Nicolls said Vera’s CPU performance and memory bandwidth are enough to support large-scale task orchestration, code execution and data processing while keeping GPUs running efficiently. SpaceXAI also plans to expand the AI infrastructure behind Grok using Nvidia’s Vera Rubin platform and move toward gigawatt-scale compute capacity. Vera Rubin combines Nvidia accelerated computing, NVLink interconnect, Spectrum-X networking and BlueField data processing technology, with the goal of lowering per-token cost and improving overall AI factory efficiency. Nvidia also said SpaceXAI will send an optimized Vera Rubin NVL72 system into space, where the first-generation Starmind AI satellite will use the rack-scale architecture. Nvidia described that as its first landmark deployment extending accelerated computing from terrestrial data centers to orbital computing.1350
Nvidia2026-08-23 00:36:46Nvidia AI server prices set to rise by more than 15% for many systems, affecting Vera Rubin and Grace BlackwellSome of Nvidia’s large customers have been told to expect higher prices for servers built with the company’s AI chips, according to a Bloomberg-cited Odaily report. The increase is tied to persistently rising memory chip costs, and in many cases server prices are expected to climb by more than 15%. The reported price changes are expected to apply to systems scheduled to begin shipping in early 2027. Products affected include Vera Rubin and Grace Blackwell. The exact size of the increase will depend on the chip generation and the memory configuration used in each system. Server manufacturers that build systems for major data center operators including Microsoft, Google, and Oracle have also recently notified customers of the planned increases. The report did not disclose additional pricing details beyond the expected threshold and the factors that will determine final adjustments.1090
Nvidia2026-08-22 18:57:43Nvidia to raise prices on some AI servers by more than 15% next yearSome of Nvidia’s major customers have been told that prices for servers equipped with the company’s artificial intelligence chips will rise by more than 15% in most cases, according to ChainCatcher, citing people familiar with the matter. The price increase is set to take effect early next year. Systems affected by the change include models built with Nvidia’s flagship Vera Rubin and Grace Blackwell chips. The update points to higher costs for select AI server buyers as Nvidia continues supplying advanced hardware tied to its latest chip lineup. No additional details on specific customers or exact pricing schedules were disclosed in the report.970
