Running SQL Server on @awscloud? Amazon R8i instances powered by Intel Xeon 6 deliver higher performance and better economics than R6i, helping you get more from every SQL Server license.

Learn more: ms.spr.ly/6017aadDZ

Benchmarks from @ProwessConsult show Intel Xeon 6 processors deliver up to 3.5x higher MySQL and 3.6x higher PostgreSQL throughput over 3rd Gen Intel Xeon systems on @RedHat OpenShift.

Read the technical brief: ms.spr.ly/6017avxed

Your AI workflows shouldn't be tied to your desk. With AI Playground Home Agent and Slack/Telegram, you can launch @PyTorch-powered image generation and content creation from your phone while the compute stays local on your Intel-powered AI PC.

See it in action at PyTorch t.co/Vsvt0BZVXr

The bottleneck in agentic AI often sits outside the model. Retrieval, orchestration, memory, and tool use land heavily on the CPU. Organizations that treat CPU and GPU like a relay team instead of separate systems tend to improve coordination across the full stack.

At @PyTorch kan Wang will highlight how we’re scaling heterogeneous systems to help developers optimize the entire stack, not just the GPU.

Learn how at ms.spr.ly/6018avvSq

Intel and AMD are working together to simplify AI development on x86.

By bringing @PyTorch CPU optimizations together through oneDNN, we’re creating a more consistent software path for developers across Intel and AMD processors.

Learn more: ms.spr.ly/6011avvs3

AI is only as strong as the infrastructure behind it.

At @IBM #TechXchange2026, discover how Intel Xeon 6 and IBM technologies help organizations scale AI inference and enterprise workloads across cloud and hybrid environments.

Learn more: ms.spr.ly/6011avP3S

How much setup stands between you and your first physical AI project?

Potentially just a few minutes.

Guy Tamir walks through two ways to install Intel's Physical AI Studio: a Docker setup for the fastest path or a native installation if you want more control.

Check out ps://t.co/pbDXkEkuHC

Hosting an 816B-parameter model on air-cooled hardware?

Intel delivers Day 0 support for @Meta's Muse Spark Open using four 160GB Intel GPU D-series cards alongside Intel Xeon processors for flexible, memory-intensive enterprise AI. It is the most capable U.S. open-weights ore: ms.spr.ly/6013at3KN

Build with performance, portability, and open-source innovation. At @PyTorch Conference 2026, discover how Intel technologies work with PyTorch to support agentic AI and high-performance model serving through engines like vLLM.

Learn more: ms.spr.ly/6012asIbI

Day 0 OpenVINO support for @OpenBMB's MiniCPM-V 4.7 is here!

Run multimodal AI workloads for document understanding, OCR, and video analysis across Intel CPUs, GPUs, and NPUs.

Learn more: ms.spr.ly/6017asxHJ

Modernize your robotics infrastructure and scale with confidence using practical, validated architectural guidance. @RedHat and Intel's joint reference architecture provides the blueprint.

Try it today: ms.spr.ly/6016asqqY

Open-source AI moves forward when builders learn, share, and create together.

Join Intel and @huggingface in San Francisco on Oct. 16 to kick off Open Source AI Week and connect with the community shaping what comes next.

See event details here: ms.spr.ly/6015asqSX

You’ve trained your physical AI model; what's next?

Guy Tamir walks through how OpenVINO Physical AI handles the workflow after training, from benchmarking and model export to inference, deployment packages, and the ongoing loop between a robot’s observation and the actions
Watch and try it out: ms.spr.ly/6017aombZ

Bay Area AI builders:

Open source demonstrates what's possible when builders collaborate to develop, improve, and share new technologies.

That's why we're celebrating Open Source AI Week with @huggingface in San Francisco.

We look forward to connecting with fellow builders and utions that continue to move open-source AI forward.

Event details: ms.spr.ly/6015aU2Ad

The next challenge for physical AI isn't just better models. It's deploying them at scale.

See how OpenVINO Physical AI helps address latency, reliability, safety, hardware constraints, and costs for VLA models like π0.5: ms.spr.ly/6014aomjK

EmbeddingGemma v2 brings high-quality embedding generation to everyday devices.

With @Google’s LiteRT-LM pipeline and OpenVINO backend, we can offer Day 0 support for EmbeddingGemma v2 for developers running models on Intel NPUs.

Learn more at ms.spr.ly/6011aomjv

Fast, accurate AI interactions depend on more than powerful models.

See how @RedHat OpenShift AI and Intel Xeon 6 processors use intelligent routing to accelerate multi-turn LLM inference and keep chatbot experiences running smoothly: ms.spr.ly/6011aoEgt

Conventional wisdom says massive AI models need specialized hardware. But mixture-of-experts workloads are often bound to memory, not compute. Intel Xeon offers an efficient, practical alternative for running large-scale MoE models.

Learn more: ms.spr.ly/6013aWNKt

Search results took a huge step forward at OpenSearchCon NA.

During our keynote, Intel and @OpenSearchProj announced Scalable Vector Search (SVS) integration for Open Search, allowing for smarter intent matching alongside efficient data compression that lets teams scale with cover what’s new: ms.spr.ly/6019aWRcD

What if the bottleneck isn't your GPU?

As AI workloads move from tokens to frames, scaling efficiently requires rethinking how each stage of the pipeline is deployed and optimized. See how a disaggregated approach delivered up to 28% higher throughput with the same hardware ms.spr.ly/6019al28b