VΛST Data

SkyPilot now works natively with @VAST_Data storage 🤝 Training runs often start with a dead period - copying data while GPUs sit idle. With VAST + SkyPilot, you can skip that entirely. • Mount petabyte-scale data directly. No staging, no prefetch pipelines, no idling • Stream at NVMe speeds to SkyPilot-managed nodes across any Kubernetes, Slurm, or neocloud • Switch compute providers without touching your storage config Read the blog by the @Vast_Data team: vastdata.com/blog/instant-d…

Andrew meets with John, where they discuss Supermicro and @VAST_Data's collaboration on exciting storage projects such as the CNode-X AI Data Platform. Learn more about Supermicro and VAST Data's AI and Storage solutions: hubs.la/Q048Hj1j0 #Supermicro #VASTData #NVIDIA

EP03 of the IT Talk: Is Edge AI About To Go Mainstream?, @Vast_Data VAST Forward Recap x.com/i/broadcasts/1…

VAST FWD will feature insights from NVIDIA Founder and CEO Jensen Huang on NVIDIA’s journey building AI infrastructure in collaboration with VAST. From training to inference to agent-based systems, Jensen outlines how enterprises are putting AI to work across their organizations.

Check out my latest article: Kubernetes Isn’t the Problem. Data Pipelines Are. linkedin.com/pulse/kubernet… via @LinkedIn @theCUBE @SiliconANGLE

Your legacy storage is squandering your flash. VAST Data is here to help you take it back. While the rest of the industry faces long delays waiting for flash allocation, VAST customers are 6x-ing their capacity with the drives they already own. Today, VAST launches VAST Amplify to eliminate the "replication tax" and turn fragmented silos into a high-performance AI data plane. 3% overhead. Up to 8:1 reduction. Zero wait time. Efficiency is the new supply. Learn more by visiting: vastdata.com/usecase/ssd-sh…

@NYSE @JeffDenworth @calebsilver @PeopleIncco @Amberdataio Join us at our inaugural global user conference and spend two days shaping, building, and scaling the future of AI: vastdata.com/vast-forward?u…

Today on NYSE Live | @JeffDenworth of @VAST_Data discusses AI goals for upcoming conference. Hear what @calebsilver of @PeopleIncco has to say on the Investopedia Survey. Plus, @Amberdataio COO says its new intelligence platform was years in the making! twitter.com/i/broadcasts/1…

Stop by Booth #3402 for your chance to win big — our #SC25 raffle kicks off today!

As Microsoft Ignite gets underway, VAST Data is proud to announce a new partnership with Microsoft Azure to set the cloud standard for modern AI. When the VAST AI Operating System runs natively in Azure, customers will be able to extend their AI foundation instantly, avoid data movement, and run GPU-accelerated workloads with consistent performance across environments. VAST is working closely with the Azure team as Microsoft advances its AI infrastructure roadmap. As the company looks ahead, this work reinforces VAST’s role in supporting emerging AI architectures with an operating system built for scale, performance, and simplicity. Read the full announcement: vastdata.com/press-releases…

The VAST AI Operating System is the data foundation for Texas Advanced Computing Center (TACC)'s Horizon, the highest performing academic supercomputer in the world, underscoring a 2+ year strategic partnership that spans TACC’s Stampede3 and VISTA. As the flagship of the National Science Foundation’s new Leadership-Class Computing Facility, the decision to extend VAST to Horizon was based on proven outcomes with Stampede3 and VISTA around performance, scale, resiliency and ease of administration. The VAST AI OS is designed to converge AI and HPC infrastructure, feeding GPU-dense training and inference side-by-side with traditional simulation workloads, allowing labs to run next-gen AI and classic HPC pipelines without re-architecting data paths, and ultimately unlocking faster time-to-results – with less time moving data and more time running models, training AI, and publishing discoveries: vastdata.com/press-releases…

We’re bringing the energy (and the swag) to #SC25. Stop by Booth 3402 for live sessions, limited-edition shirts, and a chance to see how AI infrastructure should really scale.

In the latest MLOps Community Podcast, @Dpbrinkm sits down with Andy Pernsteiner (Field CTO at @VAST_Data) to unpack what really happens behind the scenes when companies try to scale GPU infrastructure without losing their minds—or their budgets.

In today's news: VAST Data and @googlecloud have partnered to address the challenge of hybrid AI data sprawl with the first fully managed service of the VAST AI Operating System on Google Cloud. Enterprises want to run AI where it performs best but fragmented storage and costly migrations often stand in the way. This expanded partnership makes data placement a choice rather than a constraint. The VAST AI OS leverages the VAST DataSpace to allow enterprises to seamlessly connect on-prem clusters with Google Cloud, eliminating complex migrations and instantly making data available wherever AI runs. The result: a fully managed AI OS on Google Cloud that unifies hybrid data with intelligent streaming, removes migration drag, and keeps GPUs fed – so organizations can deploy AI faster at lower cost with consistent, compliant data close to compute resources: vastdata.com/press-releases…

I'm going to paraphrase @VAST_Data's CEO Renen Hallak on this one: "As we build our company, we aspire to the example set by great companies built before us. We try to follow their path, from one milestone to the next. From selling $1M, to $10M, to $100M, to $1B. From burning cash, to breaking even, to profitability. Each milestone is an additional step toward VAST Data being one of the greats. And then, one day, we realize we have stumbled upon a milestone our predecessors have never reached. We realize that going forward, there is no precedence and there is no other choice but to forge our own path." Today's announcement with CoreWeave is testament to a vibrant partnership between two AI infrastructure pioneers, and a signal to the world that the tomorrow's intelligence will not come from yesterday's technology. We are honored to work with their team, and humbly accept the challenge to help CW soar to even greater heights. reuters.com/technology/nvi…

The VAST AI Operating System is the primary data foundation for the next generation of AI. VAST Data and @CoreWeave are proud to announce a $1.17 billion commercial agreement to extend the strategic partnership. This expanded collaboration reinforces CoreWeave's commitment to the VAST AI OS and advances the shared mission to redefine the data and compute architecture for AI. The unified infrastructure delivers instant access to massive datasets, breakthrough performance, and cloud-scale economics for both training and inference workloads. Building a new class of intelligent data architecture for mission-critical industries together. Read the full announcement: vastdata.com/press-releases…

A new Frontier AI Factory in Memphis, co-built with @togethercompute, launches in early 2026! Designed to power demanding AI workloads — featuring @nvidia GB200 NVL72 systems, @DellTech’s liquid-cooled Integrated Rack Scalable Systems, and @VAST_Data's AI storage platform.

The future of intelligence is super-human, requiring a "conference of conferences" where the world's knowledge is processed at machine speed. This hyper-efficient communication is the next frontier but it demands a foundational data model capable of supporting real-time, machine-to-machine data flow. In this episode of the Shared Everything podcast, VAST Data CTO Alon Horev and @nvidia Senior VP Kevin Deierling join Nicole Hemsoth Prickett to dive deep into the architecture that makes this vision real. They discuss how VAST is extending GPU memory to provide the massive context needed for agentic workflows, enabling 5x speed-up by eliminating the need to recompute data. Listen to the full episode and learn how to build the foundation for real-time, super-human intelligence --> lnkd.in/gN5UD6zJ

Recomputing KV Cache for long prompts or repeated inputs adds latency and wastes GPU resources. The latest NVIDIA Dynamo release solves this by offloading KV Cache to CPU RAM, SSDs, or remote storage—cutting redundant computation and speeding up responses. Storage providers @VAST_Data and @WekaIO have validated this approach, proving efficient KV Cache offload at scale. And the #opensource LMCache project has integrated its management layer into Dynamo, further extending caching capabilities. 👉 Read the blog: nvda.ws/4gQRXln