@dMatrix @dMatrix_AI

Our CEO @sidsheth will be live with CNBC’s @jonfortt on Fortt Knox. Date: Monday, September 29, 2025 at 7:30 AM PT Where: Live on YouTube They will discuss today’s biggest AI challenges, the d-Matrix story, and what’s next with 3DIMC™ and Pavehawk™. Watch live:…

d-Matrix is honored to be named a finalist in two categories at the 2025 A.I. Awards: AI Startup of the Year and Most Innovative AI Technology. We’re driving a memory-centric approach to AI inference that makes GenAI more practical and sustainable at scale. More here:…

We will be at the World AI Summit 2025 in Bengaluru, September 25–26. Join us on Day 1, September 25 at 11:00 AM in Hall B for our session: Making GenAI Attainable & Sustainable with Pradip Thaker, PhD, India Country Head & VP of Engineering at d-Matrix. We’ll share how a…

The future of AI inference is memory-first. Our 3DIMC™ tech delivers 10x faster inference speed and 10x greater efficiency, breaking through the memory wall that holds GenAI back. First silicon, Pavehawk™, is live in our labs. Learn more: d-matrix.ai/scaling-ai-inf… #AI…

Incredible milestone from @satyanadella and @Microsoft. Massive GPU scale will drive the next wave of AI training. At d-Matrix, we’re focused on what comes after training — inference acceleration at scale. That’s where performance, efficiency, cost, and sustainability make a…

Satya Nadella @satyanadella

2 weeks ago

Incredible milestone from @satyanadella and @Microsoft. Massive GPU scale will drive the next wave of AI training. At d-Matrix, we’re focused on what comes after training — inference acceleration at scale. That’s where performance, efficiency, cost, and sustainability make a…

612 1K 8K 1.4M 2K

Download Video

If intelligence is the log of compute… it starts with a lot of compute! And that’s why we’re scaling our GPU fleet faster than anyone else. Just last year, we added over 2 gigawatts of new capacity – roughly the output of 2 nuclear power plants. And today we’re going further,…

Our CEO Sid Sheth joined @NextGenInfra_io , @ConvergeDigest, @AvidThink & @WireRoy at #AIInfraSummit to talk about scaling AI efficiently. He shared how JetStream IO + Corsair accelerators enable multi-node AI scaling with ultra-low latency. Watch here 👉…

At #AIInfraSummit, the industry is talking about what’s next in AI inference and d-Matrix is at the center of it. We announced our JetStream I/O accelerator for ultra-low latency scale-out. Here’s what analysts and press are saying about JetStream and our complete inference…

keyboard_arrow_left Previous keyboard_arrow_right Next

Thanks for joining Sid Sheth’s talk on making GenAI attainable and sustainable at AI Infra Summit. Takeaways: latency matters, inference efficiency is the constraint, and system co-design wins. Come by Booth 609 for a hands-on demo. #AIInfraSummit #GenAI #AI #Inference #Compute…

keyboard_arrow_left Previous keyboard_arrow_right Next

Booth 609 at AI Infra Summit is live and showcasing Corsair, Aviator, and JetStream I/O powering generative AI at scale. Stop by and see it in action. #AIInfraSummit #GenerativeAI #AIInference #dMatrix

Today we announced JetStream I/O Accelerators, built for ultra-low latency inference at scale. With Corsair and Aviator, JetStream offers 10x faster performance, 3x better cost-performance, and 3x higher energy efficiency than GPU-based solutions. Read the news:…

We just turned 6! 🎉 The d-Matrix crew plus families came together this weekend for our anniversary summer party filled with games, food, and laughs. Here’s to the people powering the future of AI. #dMatrix #AI #TeamCelebration

keyboard_arrow_left Previous keyboard_arrow_right Next

Sid Sheth, CEO & Co-Founder of D-Matrix, speaks at #AIinfrasummit in Santa Clara on Tue, Sep 9 4:10–4:30 PM. Session: Chip Design: Efficiently Powering the Next Wave of Agentic Reasoning Applications – Making Gen AI Attainable and Sustainable. #AIinfrasummit #AI #Inference

Today at Hot Chips 2025, our CTO, Sudeep Bhoja, presented Corsair – An In-memory Computing Chiplet Architecture at Stanford. Visit our booth to meet the team and see how we are tackling the memory wall with 3D in-memory compute. #AI #Inference #HotChips #Compute #dMatrix

keyboard_arrow_left Previous keyboard_arrow_right Next

By dramatically increasing memory capacity & bandwidth, 3DIMC will power the next era of scalable, efficient AI inference. Next up: Raptor, our next-gen architecture, integrating 3DIMC to break the memory wall. Learn more at d-matrix.ai/scaling-ai-inf…

Excited to announce we’re bringing 3D stacked digital in-memory compute (3DIMC™) to our roadmap. Our first 3DIMC-enabled silicon, Pavehawk™, is now operational in our labs after 2+ years of development. #AI #Inference #Memory #3DIMC #dMatrix

We’re at Hot Chips 2025. Our founder and CTO Sudeep Bhoja will dive into Corsair – An In-memory Computing Chiplet Architecture for Inference Acceleration August 26 | 2:45–4:15 PM | Stanford Join the conversation: hotchips.org #HotChips #AI #dMatrix

We are proud to share that d-Matrix is shortlisted in the 2025 A.I. Awards: 🏆 AI Startup of the Year 🏆 Most Innovative AI Technology A reflection of our team’s work to reimagine compute for the AI era. Onward. 🔗 cloud-awards.com/ai-awards/ai-s… #AI #Innovation #AIAwards

“How do we make AI acceleration radically more efficient?” d-Matrix’s Sree Ganesan shares what it’s like building AI accelerators from inside a startup. A smart, inspiring listen. 🎧 electronicsweekly.com/blogs/electro-… #AI #Inference #WomenInTech

"GPUs dominated training. Inference? That’s a different game." — Sid Sheth, CEO of d-Matrix, on The New Stack Makers 🎧 Inference needs custom hardware, closer memory, & agent-aware compute. A must-listen for anyone in AI infra. 🔗 youtube.com/TheNewStack #AI #Inference