Looks like a good value prop for Intel. Hopeful they can carry the Gaudi learnings into their datacenter discrete GPU lineup. Should learn about this this week at Vision. $INTC

Early testing shows highly competitive performance from Intel Gaudi 3, with a strong economic advantage on IBM Cloud. More detailed analysis coming soon.

Now we are hearing from Sachin Katti, head of the NEX group at Intel. Talking through real-world implementations of all of this technology, including: Showing performance data from our Signal65 Gaudi 3 testing, published yesterday. Read here: signal65.com/research/ai/in…

We continue to see areas of pure performance AND performance-per-dollar leadership for the @Intel Gaudi 3 AI accelerators, and now that it's on IBM Hybrid Cloud & Infrastructure, availability is better than ever. Read our full report: signal65.com/research/ai/si…

.Intel Gaudi 3 shows highly competitive inference performance to NVIDIA GPUs. When considering IBM Hybrid Cloud & Infrastructure pricing, Gaudi 3 achieves a notable economic advantage for a wide range of AI workloads. The performance + economics + cloud availability is well suited for enterprise AI

Is your AI/HPC performance being choked by slow data access? 👀 Our latest market report in partnership with Hammerspace (led by analyst Mitch Lewis) unveils Tier 0, unlocking shared NVMe for up to 10x faster speeds & 15% more GPU utilization! 🚀 Curious how? Learn

Exciting news! We're thrilled to announce that Michael Dell 🇺🇸, Founder, Chairman, and CEO of Dell Technologies, will deliver the Show Opening Keynote for The Six Five Summit: AI Unleashed 2025. This fully virtual event, taking place June 16-19, is free to attend and will explore the

Next stop San Francisco for #CFD23 Tech Field Day - Cloud Field Day #CFD24 10 22-23

Is your cloud data strategy driving value or just driving up your bill? ☁️ 💰 On Tech Field Day Podcast, Alastair Cooke, Mitch Lewis, & Vriti Magee discuss smarter scaling, cloud costs, & data abundance Watch 👉 buff.ly/6FfgMRu #CFD23 #TFDPodcast #CloudData #Scalability

Last week AMD took the covers off the new Instinct MI355X AI accelerator and Signal65 got hands on in AMD labs for testing. A single 8-GPU MI355X server can run LoRA MLPerf nearly 10% faster than four-node, 32-MI300X GPU config! Read the full report: signal65.com/research/ai/am…

From inventing the "GPU" in 1999 to pioneering real-time ray tracing, NVIDIA has clearly transformed PC gaming. Our latest article dives deep into how DLSS, RTX, and AI-powered features are reshaping what's possible in gaming. Read the full report: signal65.com/research/nvidi…

If you thought fine tuning large AI models was strictly for data center GPUs, then you might be interested to learn about Phison Electronics USA aiDAPTIV+ that uses high-speed storage and desktop GPUs to bridge the gap bw edge and cloud AI compute. Read the report: signal65.com/research/ai/co…

🚀 Mew data: Cost Efficient On-Premises AI Processing with Phison Electronics USA aiDAPTIV+ ✅ Up to 85% cost savings vs traditional AI infrastructure ✅ Enhanced security with on-premises processing ✅ Break memory constraints that stop standard configs Report: signal65.com/research/ai/co…

We hosted two Phison Electronics USA directors in a discussion of the company's aiDAPTIV+ technology and our Signal65 report covering its capabilities and performance. Some great context here! youtube.com/watch?v=w9nDtG…

The latest Signal65 testing shows major advantages for Arm based Graviton4 on AWS. Graviton4 achieved advantages in both performance and price/performance across 4 distinct workloads: - Redis - Nginx - XGBoost - Llama Read the full report: signal65.com/research/arm-n…

A new benchmark report from Signal65 just dropped! In this analysis, Arm-based Amazon Web Services Graviton4 instances didn’t just outperform x86 on key general compute and AI workloads, but they did so at a much better price-performance ratio. ⬇️⬇️⬇️ signal65.com/research/arm-n…

☁️ ARM can win in the cloud: AWS Graviton4 delivers 34% faster XGBoost training than x86 instances while offering better cost efficiency. Signal65's latest research shows why 50%+ of AWS's new CPU capacity is now Arm based. Full analysis: signal65.com/research/arm-n…

$ARM continues to make significant gains in cloud. $AMZN AWS now seeing 50%+ CPU instances ARM based.

With all the Arm and Intel talk, I thought this was interesting testing. Amazon Web Services Graviton versus X86.

Ready to kick things off here at #CFD24 Tech Field Day - Cloud Field Day #CFD24 10 22-23