Keith Nellis takes the stage tomorrow at OCP Global Summit. Join us from 8:54–9:09 AM at SJCC, Concourse Level – Expo Hall Theater for session: “Efficiently Powering the Next Wave of Agentic Reasoning Applications — Making GenAI Attainable and Sustainable.” Keith will share how d-Matrix is advancing memory-centric inference to deliver breakthrough performance and efficiency for real-world GenAI workloads. Visit us at Booth C11 to see Corsair, JetStream, and SquadRack in action. #OCPGlobalSummit #GenAI #Inference #dMatrix #AIInfrastructure #DataCenters
d-Matrix
Semiconductor Manufacturing
Santa Clara, California 14,005 followers
Making Generative AI Commercially Viable
About us
To make AI inference commercially viable, d-Matrix has built a new computing platform from the ground up: Corsair™, the world’s most efficient compute solution for AI inference at datacenter scale. We are redefining Performance and Efficiency for AI Inference at scale.
- Website
-
http://www.d-matrix.ai
External link for d-Matrix
- Industry
- Semiconductor Manufacturing
- Company size
- 51-200 employees
- Headquarters
- Santa Clara, California
- Type
- Privately Held
- Founded
- 2019
- Specialties
- AI, semiconductor, Inference, AI chips, and Blazing fast inference
Locations
-
Primary
5201 Great America Pkwy
Santa Clara, California 95054, US
Employees at d-Matrix
Updates
-
Lots of energy at the OCP Global Summit 2025! Stop by Booth C11 to see our latest AI inference innovations — including SquadRack™, JetStream™, and Corsair™, purpose-built for scalable, low-latency generative AI. #OCPGlobalSummit #dMatrix #Corsair #JetStream #SquadRack #AIInference #GenerativeAI #DataCenters
-
Meet SquadRack™, the industry’s first rack scale, purpose-built solution for AI inference at datacenter scale. Developed in collaboration with Arista Networks, Broadcom, and Supermicro, SquadRack introduces a disaggregated, standards based reference architecture for ultra-low latency, batched inference. Powered by our Corsair compute-memory accelerators and JetStream IO accelerators, SquadRack configurations are customizable and make it easy to deploy in existing data centers. We’re showcasing it this week at #OCP2025, Booth C11. Visit us and learn how d-Matrix is redefining how inference scales one rack at a time. Read the full press release here: https://lnkd.in/gnM2QayJ #OCP2025 #AIInference #AIInfrastructure #RackScale #DataCenter #dMatrix #Corsair #JetStream #SquadRack #Arista #Broadcom #Supermicro Jas Tremblay Vik Malyala Vijay Vusirikala
-
-
We’re heading to OCP Global Summit this week in San Jose, CA! Stop by Booth C11 from October 14–16 to see how d-Matrix is redefining AI inference at scale. Come meet the team, explore our latest hardware and software innovations, and learn how we’re powering the next generation of AI data centers. #OCP2025 #AIInfrastructure #InferenceAcceleration #DataCenter #dMatrix
-
-
#Hiring At d-Matrix, we’re redefining how generative AI inference is deployed and scaled in the data center. We're looking for an AI Software Applications Engineer to help customers unlock the full potential of our AI hardware platform. In this customer-facing, highly technical role, you’ll work across the stack—from model deployment to performance tuning—making generative AI faster, more efficient, and easier to run at scale. If you're passionate about solving complex AI challenges at the intersection of ML software and next-gen hardware, let’s connect. Apply: https://lnkd.in/gRyAr3v8 #AIInference #GenerativeAI #MLInfrastructure #AIAccelerators #SoftwareEngineering #CustomerSuccess
-
-
Last week at the World AI Summit 2025 in Bengaluru, our India Country Head & VP of Engineering, Pradip Thaker, presented on how d-Matrix is redefining AI inference. His session highlighted the shift from GPU-centric architectures—built for training—to new solutions purpose-built for inference at scale. With memory-centric architectures, we’re breaking through the memory wall by integrating compute and memory, delivering massive gains in both capacity and bandwidth. We’re proud to have contributed to the important discussions on inclusivity and responsibility that shaped this year’s Summit #WorldAISummitBLR #WorldAI #eletsAISummit #AISummit #AISummitBLR | Elets Technomedia | IT BT Department, Karnataka | Karnataka Digital Economy Mission (KDEM) | NxtGen Cloud Technologies Pvt Ltd
-
-
Last week at the World AI Summit in Bengaluru, d-Matrix was honored with the Most Promising AI Startup Award. The recognition comes as we mark more than three years of building in India. This award highlights the strength of our d-Matrix India team and the global collaboration that powers every innovation at d-Matrix. We’re proud to be shaping the future of AI inference together. Kota Vamsi Krishna Darsi, our first India employee, received the trophy on behalf of leadership, joined by teammates Senthil, Malav, Santosh, Ranjith, Adeeb, Pradip, and Dilip. #WorldAISummitBLR #WorldAI #eletsAISummit #AISummit #AISummitBLR | Pradip Thaker | Elets Technomedia | IT BT Department, Karnataka | Karnataka Digital Economy Mission (KDEM) | NxtGen Cloud Technologies Pvt Ltd
-
-
Interview starts in 30 minutes (7:30 AM PT). Join our CEO, Sid Sheth in conversation with Jon Fortt (Fortt Knox / CNBC’s Closing Bell: Overtime) on YouTube. We’ll dig into: • The toughest challenges in AI—scaling compute while managing cost and energy • The d-Matrix origin story and Sid’s entrepreneurial journey • What’s next for inference with 3D stacked Digital In-Memory Compute (3DIMC™) Watch here: https://lnkd.in/gRDHkPeR #AI #GenerativeAI #Inference #ForttKnox #dMatrix
-
Earlier today, our CEO Sid Sheth sat down with Jon Fortt on Fortt Knox / CNBC’s Closing Bell: Overtime. If you missed it, watch the recording in the post below. #AI #GenerativeAI #Inference #ForttKnox #dMatrix
I speak with d-Matrix CEO Sid Sheth about AI acceleration and inference, the challenges of AI energy consumption, and his entrepreneurial journey.
Sid Sheth, d-Matrix CEO: A Fortt Knox Conversation
www.linkedin.com
-
We’re excited to share that our CEO, Sid Sheth, will be joining Jon Fortt host of Fortt Knox and CNBC’s Closing Bell: Overtime, for a live conversation. Date: Monday, September 29, 2025 at 7:30 AM PT Format: Livestream on YouTube The discussion will cover: - Today’s toughest challenges in AI, including scaling compute while tackling costs and energy constraints - The d-Matrix origin story and Sid’s entrepreneurial journey - What’s next for inference with 3D stacked Digital In-Memory Compute (3DIMC™) and more This is an opportunity to hear how d-Matrix is making generative AI accessible, commercially viable, and sustainable for enterprises worldwide. Join the livestream here: https://lnkd.in/gRDHkPeR #AI #GenerativeAI #Inference #ForttKnox #dMatrix
-