tldrdailynews.com logo
Tag

Vera Rubin

All articles tagged with #vera rubin

Nvidia Unveils DSX POD: Containerized Vera Rubin AI Factories Cut Deployment to 14 Days
technology2 hours ago

Nvidia Unveils DSX POD: Containerized Vera Rubin AI Factories Cut Deployment to 14 Days

Nvidia announced the DSX POD, a containerized AI factory containing seven Vera Rubin racks, designed to cut data center deployment times from six months to 14 days. The system, weighing 28,000 kg, is fully assembled with liquid cooling and cabling at the factory before truck delivery. While Nvidia did not disclose pricing, analysts estimate a single rack costs over $7 million. The move aims to accelerate revenue generation for customers by reducing onsite construction and integration efforts, with global deliveries expected in early 2027.

via CNBC|
#ai-infrastructure#data-centers#gtc-berlin
Nvidia, Meta, and Google Standardize Open AI Factory Designs at 2026 OCP Summit
technology9 days ago

Nvidia, Meta, and Google Standardize Open AI Factory Designs at 2026 OCP Summit

Nvidia, Meta, and Google have jointly advanced open standards for AI data center design at the 2026 OCP Global Summit in San Jose. The collaboration focuses on the Vera Rubin platform, featuring cable-free liquid-cooled racks and high-density power distribution. These open specifications aim to accelerate the deployment of agentic AI by improving serviceability and energy efficiency across the industry.

via Yahoo Finance|
#ai-infrastructure#nvidia#ocp-global-summit
Nvidia’s Groq 3 LPX enters full production with 3,400 tok/s benchmark
technology1 month ago

Nvidia’s Groq 3 LPX enters full production with 3,400 tok/s benchmark

Nvidia has placed its Groq 3 LPX rack into full production, marking the commercialization of its $20 billion acquisition of Groq. Independent benchmarks by Artificial Analysis show the system achieving 3,431 tokens per second on Google’s Gemma 4 31B model with a 100,000-token context. This performance is four times faster than Cerebras’ current offering. The hardware uses SRAM-heavy architecture to minimize latency for agentic AI workloads, pairing with Nvidia’s Vera Rubin GPUs to separate prefill and decode phases. Nebius is the first customer to deploy the racks, which are expected to be online later this year.

via The Register|
#ai-inference#benchmarks#groq
SpaceX Picks Nvidia’s Vera Rubin for AI, Musk Hails Nvidia as AI Leader
business2 months ago

SpaceX Picks Nvidia’s Vera Rubin for AI, Musk Hails Nvidia as AI Leader

Elon Musk announced that SpaceX will exclusively use Nvidia’s Vera Rubin AI architecture by year-end, with Musk praising Nvidia as making the best AI computer. The move reinforces Nvidia’s strong customer base (SpaceX being a major client), but analysts expect it to have limited impact on Nvidia’s stock given the massive ongoing AI data-center spending by hyperscalers and SpaceX’s own capital expenditure.

via Yahoo Finance|
#ai#business#elon-musk
SpaceX Compute Push Could Put Nvidia on a $1 Trillion Revenue Path
technology2 months ago

SpaceX Compute Push Could Put Nvidia on a $1 Trillion Revenue Path

Analysts say SpaceX's pledge to run its AI compute on Nvidia's Vera Rubin GPUs and to scale toward 20 GW by 2027 could unlock outsized Nvidia revenue: even a small slice (2 GW) could add roughly $100 billion in incremental revenue, with the potential for Nvidia to reach about $1 trillion in cumulative Vera Rubin chip revenue by 2027. Nvidia's own data-center business is a large portion of its revenue, and the story hinges on SpaceX delivering on its targets; SpaceX's Q2 results and high capex underscore the opportunity, but this is not Nvidia guidance.

via 24/7 Wall St.|
#ai-compute#ben-reitzes#nvidia
Nvidia commits $5B to scale Ilya Sutskever’s breakthrough AI project
technology2 months ago

Nvidia commits $5B to scale Ilya Sutskever’s breakthrough AI project

Nvidia is investing about $5 billion in Safe Superintelligence (SSI) to access Nvidia’s Vera Rubin hardware, enabling SSI to roughly tenfold its computing power over the next year under milestone-based terms; SSI remains secretive with no product yet, as the deal deepens Nvidia’s ties to a high-profile AI venture amid similar large investments in the sector.

via Financial Times|
#artificial-intelligence#ilya-sutskever#nvidia
NVIDIA Holds Key Support as Investors Await Vera Rubin Chips
business3 months ago

NVIDIA Holds Key Support as Investors Await Vera Rubin Chips

Nvidia edged around $199 after a tech-sector selloff, holding near a key support level as investors weigh how long the current AI demand cycle lasts and whether the Vera Rubin chips will sustain leadership. The stock has lagged peers this year but remains reasonably valued (forward earnings around 19x) and is returning roughly half of its free cash flow to shareholders via buybacks and dividends. A breakout hinges on Vera Rubin’s arrival and durability in the second half of 2026, suggesting the stock could trade in a new range rather than rally immediately.

via Barron's|
#ai-chips#business#buybacks
Micron Bulls See Big Upside, But Is the Rally Truly Justified?
investing4 months ago

Micron Bulls See Big Upside, But Is the Rally Truly Justified?

Micron Technology trades near $900 with a forward multiple around 8–9x after a strong Q2 beat and guidance, as NVIDIA certifies MU as an HBM4 supplier for the Vera Rubin AI project. The company’s Q3 guide ($33.5B revenue, $19.15 EPS, ~81% gross margin) and a 30% dividend raise plus $650M in buybacks bolster the bull case that AI-memory demand could keep margins elevated. Yet insiders have been net sellers while the stock has surged well over the past year, underscoring remaining cyclicality in memory stocks. A high-profile target of $1,750 contrasts with consensus around ~$939, signaling a wide split on valuation. If demand stays robust, 8–9x forward earnings could look cheap; if AI demand slows or memory cycles turn, the stock could face a meaningful pullback.

via 24/7 Wall St.|
#ai-memory#hbm4#investing
NVIDIA Maps the AI-Driven PC and Robotics Era at Computex 2026
technology4 months ago

NVIDIA Maps the AI-Driven PC and Robotics Era at Computex 2026

NVIDIA kicked off Computex 2026 with Jensen Huang emphasizing agentic AI across hardware and software, announcing Vera Rubin full production, Vera CPUs, and RTX Spark chips with Windows-on-Arm support (N1X-based laptops/desktops), plus a Windows-based DGX Station. The company showcased robotics platforms Cosmos 3 and Isaac GROOT, rolled out Nemotron 3 Ultra as an open model, and highlighted Cadence tooling to speed chip design, all tied to a Taiwan-focused ecosystem and DSX for AI factories. The keynote framed compute as data and pledged energy-efficient, high-throughput systems designed for agents.

via ServeTheHome|
#agentic-ai#computex-2026#nvidia
NVIDIA Vera Rubin Reaches Full Production to Power Global Agentic AI Factories
technology4 months ago

NVIDIA Vera Rubin Reaches Full Production to Power Global Agentic AI Factories

NVIDIA says Vera Rubin has entered full production, with OEMs and global suppliers scaling Vera Rubin-based POD-scale AI factory systems that unify Vera Rubin NVL72 GPUs, Vera CPU, Groq 3 LPX, Vera BlueField-4 storage, and Spectrum-X Ethernet switches to deliver roughly 10x agent-throughput at scale versus the prior Grace Blackwell platform. The system uses Spectrum-X co-packaged optics for high-density, secure multi-tenant workloads, supported by DSX tooling and DOCA security to accelerate deployment. Production shipments are set to begin this fall, with a broad ecosystem across 30 countries and hundreds of partners.

via NVIDIA Newsroom|
#agentic-ai#bluefield-4#pod-scale
NVIDIA Unveils Vera Rubin: A Seven-Chip AI Factory for Agentic AI
technology7 months ago

NVIDIA Unveils Vera Rubin: A Seven-Chip AI Factory for Agentic AI

NVIDIA announced the Vera Rubin platform, a POD-scale AI supercomputer that combines seven production chips (NVL72 Rubin GPUs with 72 GPUs and 36 Vera CPUs connected by NVLink 6, plus Groq 3 LPX, BlueField-4 STX, Spectrum-6 SPX, and ConnectX-9) to run every phase of AI—from pretraining to agentic inference—across massive AI factories. The system promises major efficiency gains (one-fourth the GPUs of the Blackwell platform, up to 35x higher inference throughput per megawatt, and up to 10x more revenue potential for trillion-parameter models) and includes DSX power provisioning, storage, and networking optimizations, with ecosystem support and availability through major cloud providers and OEMs in the second half of 2026.

via NVIDIA Newsroom|
#agentic-ai#ai-factory#nvidia
Micron Faces HBM4 Shift as Nvidia's Vera Rubin Draws Supply From Rivals
technology7 months ago

Micron Faces HBM4 Shift as Nvidia's Vera Rubin Draws Supply From Rivals

Micron, a key Nvidia memory supplier, could lose share as Nvidia's Vera Rubin reportedly sources HBM4 from SK Hynix and Samsung, sidelining Micron in the AI data-center memory supply; while demand for Micron's other memory chips may persist and memory prices could rise on persistent shortages, the hit may be temporary as Vera Rubin ramps.

via The Motley Fool|
#hbm4#memory-chips#micron-technology
Nvidia’s Vera Rubin: a modular, power-efficient AI data-center rack
technology7 months ago

Nvidia’s Vera Rubin: a modular, power-efficient AI data-center rack

Nvidia unveiled Vera Rubin, its next AI rack-scale system set to ship in H2 2026, boasting about 10x the performance per watt of the Grace Blackwell platform while consuming roughly twice the total power. The 72 Rubin GPUs and 36 Vera CPUs reside in a fully liquid-cooled, modular rack designed for quick swaps and easier upgrades. Competition from AMD’s upcoming Helios and in-house cloud chips from Amazon and Alphabet adds pressure, but Meta plans to deploy Rubin by 2027. Pricing is reportedly around $3.5–$4 million per rack, signaling strong pricing power as AI data-center demand grows. Analysts show a Strong Buy with a notable upside potential.

via TipRanks|
#ai-hardware#data-centers#nvda