Featured
Latest posts
Featured
Agents Need a Cloud They Can Verify
AI agents keep their memory, credentials and authority on a CPU. TractorOS and confidential VMs on Intel Xeon 6 with Intel TDX, verified by Intel Trust Authority, launch on FarmGPU Cloud by the end of 2026.
888 GB/s & 23.6 PB in 16U: Solidigm's D5-P5336 QLC KV-Cache Tier, Measured in MLPerf® Storage v3.0 Open
QLC Wasn't Supposed to Keep Up With KV Cache. One of Our Clusters Did.
The Fastest 8 Node WEKA Cluster: Solidigm D7-PS1010 and WEKA Power FarmGPU's MLPerf Storage v3.0 Results
Announcing FarmGPU's MLPerf Storage V3.0 Results with Solidigm and WEKA
·
OCP
Scaling AI Clusters at Neoclouds: How FarmGPU Is Helping Build the Open Playbook
Notes from the new OCP FTI workstream FarmGPU is chairing — workgroup goals, 19 community survey responses, the April 2026 hardware-management deep dive, and how FarmGPU is using OCP today.
100x to 280x KV Cache Acceleration: How FarmGPU, Lightbits, and ScaleFlux Are Solving the Long-Context Inference Bottleneck
At GTC 2026, FarmGPU, Lightbits Labs, and ScaleFlux preview a collaborative architecture delivering 100x-280x KV cache acceleration — enabling 3x more inference requests on the same GPUs with 65% lower infrastructure costs.
FarmGPU Takes #1 in ClusterMax 2.0 Storage Benchmarks
FarmGPU ranked #1 in SemiAnalysis ClusterMax 2.0 storage benchmarks, beating every major neocloud and hyperscaler with sub-1-second PyTorch imports.
Building an AI Cluster: Our 17-Day Crash Course in Open Networking
What we learned building a 32-node B200 GPU cluster with OCP-based Ethernet—and why 90% of our time was spent debugging the network
The Great Inversion: AI is Rewriting the Economics of Cloud Infrastructure
The Neoclouds vs Hyperscalers OCP panel discussion didn't just review the current state of cloud infrastructure—it revealed a fundamental, ongoing economic restructuring…
Solidigm and FarmGPU Partner to Redefine Data Storage for the AI Era
Have you ever wondered what it takes to build a world-class AI infrastructure? It's not just about powerful GPUs and cutting-edge…
Bare metal performance with FarmGPU and RunPod Blackwell Instant Clusters
The world of AI development is about to get a massive speed boost. Today, FarmGPU and RunPod are thrilled to jointly announce the immediate availability…
The Neocloud Storage Imperative: Architecting the Data Foundation for the AI Factory
A new class of cloud provider, the "neocloud," has emerged to service the voracious computational demands of the artificial intelligence (AI) revolution.1…
Redefining AI Storage Economics: A Deep Dive into Single-Node Performance and System-Level Optimization
Performance Analysis: From Baseline to Peak Comparing unoptimized, production-optimized, and theoretical maximum throughput View Full Performance Tiers FarmGPU Unoptimized Baseline Standard configuration without Farm…