@dMatrix
@dMatrix_AI

Purpose-built AI inference for modern data centers. Powering heterogeneous AI infrastructure with digital in-memory compute.

 Santa Clara, CA
dmatrix.ai/
Joined February 2022
845 tokens/sec per user. ~50% lower latency. d-Matrix Corsair + NVIDIA H100 are redefining AI inference with attention-FFN disaggregation. Faster inference. Smarter compute. Read more: d-matrix.ai/accelerating-ai-…
5
11
73
5,002
Meet d-Matrix Squadron. AI inference at data center scale. Up to 1,024 Corsair XPUs. Ultra-low latency. Air-cooled, megawatt-scale compute. Built for the next era of AI inference. Explore Squadron: d-matrix.ai/announcements/sq… #OCP2026 #AIInference
3
6
34
1,053
Big ideas. Bold conversations. NYC energy. ⚡ We're teaming up with @parasail_io and @youdotcom for the AI Engineer New York Speaker VIP Happy Hour! AI innovators, great connections, and conversations shaping what's next. See you in NYC! 🚀 luma.com/6dobkzfh #AIEngineer #AIInference #dMatrix
2
5
403
Next week, we’re heading to #OCPSummit26! Don’t miss d-Matrix: Oct 12, 12 PM — Sid Sheth Oct 14, 1:15 PM — Sai Chalamalasetti 📍 Visit us at Booth C82! Two sessions. One mission: Infinite Inference. opencompute.org/events/ocp-s…
1
5
418
AI infrastructure is evolving — and it takes an ecosystem to move it forward. Great to see @AristaNetworks’ latest announcement on open Ethernet rack-scale infrastructure for AI. As CEO Sid Sheth said: “Delivering interactive AI at scale is a complex tech challenge requiring collaboration across the ecosystem.” Open standards. More choice. Built for what’s next. arista.com/en/company/news/p… #AIInference #dMatrix #AIInfrastructure
1
6
524
We’ve been named one of @FastCompany’s 2026 Next Big Things in Tech in the Foundational AI category. Our Corsair™ AI inference platform and memory-centric architecture are redefining AI inference — delivering up to 10x faster interactive inference and up to 3x better energy efficiency for LLM workloads, without relying on scarce HBM. Proud of the d-Matrix team. Read more: fastcompany.com/next-big-thi… #FCTechAwards #AIInference #dMatrix
1
14
812
Inference is changing, and so are the metrics we use to measure it. In a new interview with @letsdatascience, d-Matrix CTO Sudeep Bhoja shares how to evaluate next-gen inference platforms through real-world performance, efficiency, and the metrics that matter most. The conversation also explores Raptor and the role of heterogeneous infrastructure. Read more: letsdatascience.com/news/d-m… #AIInference #InfiniteInference
1
1
5
391
d-Matrix Founder & CEO @sidsheth joined industry leaders at the @GlobalSemi (GSA) US Executive Forum for “The $700 Billion Bet — What Hyperscalers Want from the Semiconductor Industry.” A timely discussion on what the next phase of AI infrastructure will demand from the semiconductor industry as AI scales.
1
2
5
541
AI Infra Summit went beyond the show floor. We welcomed a group of reporters and industry analysts to the d-Matrix office and lab for a closer look at what we’re building, from our silicon to our server racks. Thank you to everyone who joined us for the lab tour. #InfiniteInference #AIInfraSummit
2
10
519
The path to infinite inference. d-Matrix CEO @sidsheth took the Main Stage at #AIInfraSummit2026, while Satyam Srivastava and Chris Nicol joined discussions on low-latency inference and distributed AI infrastructure. Thanks to everyone who joined us at Booth 722. #AIInference #dMatrix
2
17
517
Heterogeneous inference is here. Stop by Booth 722 at #AIInfraSummit2026 to see how GPUs + d-Matrix are coming together to power AI inference at scale. #AIInference #AIInfrastructure
2
13
449
A great first day at #AIInfraSummit2026. Today, d-Matrix Distinguished Engineer Satyam Srivastava gave a talk on how to close the inference efficiency gap, and Vice President Chris Nicol joined industry leaders for a discussion on operating distributed AI across compute, networking, and cloud-native platforms. Join us tomorrow at Booth 722 and on the main stage to see d-Matrix CEO Sid Sheth give a keynote on “The Path to Infinite Inference” at 4:50pm. #AIInference #AIInfrastructure #dMatrix
1
10
380
Tomorrow at #AIInfraSummit2026, join d-Matrix at Booth 722 and on stage throughout the day. “Solving the Inference Efficiency Gap” with Satyam Srivastava 🕚 Tuesday, September 15 at 11:00 AM 📍 Compute Track “Operating Distributed AI” with Chris Nicol 🕝 Tuesday, September 15 at 2:20 PM 📍 Main Stage “The Path to Infinite Inference” with Sid Sheth 🕔 Wednesday, September 16 at 4:50 PM 📍 Main Stage #AIInference #AIInfrastructure #dMatrix
1
5
334
Big news for d-Matrix and AI inference. We’re collaborating with NVIDIA to bring our next-gen Raptor inference XPUs into NVIDIA's MGX rack-scale infrastructure with NVIDIA NVLink scale-up interconnect. Ultra-low latency inference. Premium-level token services. Built to scale. Read the full announcement: d-matrix.ai/announcements/d-… #NVLinkFusion #AIInfrastructure #AgenticAI #dMatrix
4
15
101
10,790
A great week in Taipei at SEMICON Taiwan 2026. At the Memory Executive Summit, d-Matrix Vice President Chris Nicol joined leaders from across the semiconductor ecosystem to explore how memory is becoming an increasingly important part of AI system architecture. Thank you to everyone who connected with us in Taipei for the thoughtful conversations. #SEMICONTaiwan #AIInference #AIInfrastructure
1
10
675
d-Matrix will be participating in the CASPA 35th Annual Conference & CEO Summit this October. Join CEO @sidsheth for the Edge AI & Intelligent Infrastructure panel alongside Krishna Rangasayee and Melton Chang, moderated by Dylan Patel. 📍 Santa Clara Convention Center 🗓 October 23 🕓 4:20–5:10 PM #AIInfrastructure #EdgeAI
10
454
d-Matrix is excited to join the AI infra Summit in Santa Clara, September 15–17. @sidsheth, Chris Nicol, and Satyam Srivastava will take the stage across the summit to share perspectives on scaling low-latency inference, operating distributed AI infrastructure, and the path to infinite inference. Join us for the sessions and stop by Booth 722 to meet our team and learn more about what we’re building for the next generation of AI inference. #AIInfraSummit #AIInference #AIInfrastructure #dMatrix
1
5
424
A great evening in San Francisco for AI Infrastructure, After Hours, following @anyscalecompute's Ray Summit. Our VIP happy hour, co-hosted with @parasail_io, brought together founders, engineers, and AI infrastructure leaders for thoughtful conversations on inference, heterogeneous systems, and where the industry is headed next. Thank you to the Parasail team and everyone who joined us. #AIInference #AIInfrastructure #HeterogeneousComputing #dMatrix
2
9
728
The trillion-dollar AI chip market is taking shape, and inference is opening up a new competitive landscape. The AI infrastructure landscape is changing fast. @business for Bloomberg looks at @dMatrix_AI, @nvidia, @AMD, @Broadcom, @Google, @Amazon, @Meta, @OpenAI and others are shaping what comes next. As our CEO @sidsheth shared: “Inference is not a one-size-fits-all, so brute-forcing inference with a single chip is not going to work.” The Age of Inference is here. finance.yahoo.com/technology…
1
6
28
1,245
At #HotChips2026, d-Matrix is presenting Raptor™, our 3D DRAM architecture built for the growing memory demands of AI inference. Raptor brings memory closer to compute, delivering massive bandwidth with significantly lower I/O energy than HBM. @ServeTheHome goes inside the architecture and the work behind Raptor: servethehome.com/d-matrix-ra… #AIInference #AIInfrastructure #3DDRAM #dMatrix
6
24
106
8,688
Heterogeneous compute is becoming the standard for AI infrastructure. @CallosumAI just announced their global partner network. d-Matrix is proud to be part of the US ecosystem alongside @Intel, @AMD, and others. No single chip does every job well. The right workload deserves the right silicon. Corsair was built for exactly this: purpose-built for inference, with the memory bandwidth and efficiency to make AI systems practical instead of unsustainable. Read more: callosum.com/blog/global-het… #AIInference #HeterogeneousCompute
1
8
24
1,286
At #HotChips2026, d-Matrix CTO Sudeep Bhoja will join Aayush Ankit of Meta to discuss 3D DRAM-based acceleration for generative inference and the role of memory innovation in the next generation of AI infrastructure. Join them during the Memory Technology tutorial at Hot Chips 2026. 📍 Stanford University 🗓 August 23 🕦 11:30 AM #AIInference #AIInfrastructure #Memory #dmatrix
4
13
1,360
A great day at #ModCon2026 yesterday, where @dmatrix_AI Fellow and Distinguished Architect Satyam Srivastava joined Chris Lattner, Rashid Attar, and Kamran Khan on The Multi-Silicon Stack panel to discuss what it will take to build AI systems across an increasingly diverse compute landscape. Abdul Dakkak of Modular also highlighted how the d-Matrix team integrated Mojo into our MLIR stack and had matrix multiplication running on Corsair™ within days. Thank you to the @Modular team for bringing together the builders and leaders shaping what’s next. #AIInfrastructure #AIInference #HeterogeneousComputing #dMatrix
6
22
4,022
The future of AI infrastructure is heterogeneous. Our work with @infinity_ai_ demonstrates how to unlock the full performance of the memory-centric Corsair™ accelerator and deploy heterogeneous compute in production faster. Qwen3 is the latest example of what’s possible. Learn more: infinity.inc/research/dmatri… #AIInference #AIInfrastructure #dMatrix
1
4
12
1,779
As AI systems evolve, heterogeneous infrastructure is emerging as a critical path forward. Join d-Matrix Fellow and Distinguished Architect, Satyam Srivastava, at #ModCon2026 where he will speak on the shift toward heterogenous AI systems. 📍 Grand Hyatt San Francisco 🗓 August 18 🔗 modular.com/modcon
6
630
Our interns, mentors, and managers took a break from the office for a fun evening at Topgolf ⛳ Thanks to everyone who came out and made it such a great time! #dMatrix #LifeAtdMatrix #InternLife
1
6
681
A milestone worth celebrating. Thank you, @Nasdaq, for recognizing d-Matrix’s acquisition of @WallarooAI on the Nasdaq Tower in Times Square. We’re excited about what’s ahead as we help customers move AI from models to production at scale. #AI #Inference #EnterpriseAI #dMatrix
8
4
34
3,341
We’re excited to welcome @Wallarooai to d-Matrix. By adding AI deployment and orchestration software to our purpose-built inference platform, we’re making it easier for customers to deploy and scale heterogeneous AI inference workloads from server to rack scale. This is another major step in our end-to-end silicon-to-software strategy. Read the full announcement: d-matrix.ai/announcements/d-… #AI #Inference #AIInfrastructure #MLOps
1
1
11
1,318
The future of AI starts with better infrastructure. Join our Vice President, Dr. Chris Nicol, at #SEMICONTaiwan 2026 as he shares how innovations in memory and AI inference are driving the next wave of semiconductor advancement. 📍 Memory Executive Summit 🗓 September 1 🕜 1:30 PM semicontaiwan.org/en/memory_…
3
330
Bringing new AI models to production shouldn't take months. We're excited to partner with @infinity_ai_ to accelerate performant, full-model inference on d-Matrix Corsair. Together, we're combining hardware-aware model optimization with the d-Matrix inference platform to help bring production AI models online faster, reducing developer effort and enabling full-model inference in weeks instead of months. Learn more: infinity.inc/research/dmatri… #AI #AIInference #LLM #AIInfrastructure
1
5
16
1,146
AI inference is entering a new era. d-Matrix CEO @sidsheth and CTO Sudeep Bhoja discuss how heterogeneous inference with Corsair™ and @nvidia GPUs is accelerating token generation, reducing latency, and powering the next generation of AI infrastructure. Watch the full interview. piped.video/watch?v=kHhrhICZ… Thanks @furrier , @theCUBE , @GemmaAllenSays , @bjbaumann2014 for the great conversation. #AIInference #HeterogeneousCompute #AIInfrastructure #LLM
1
6
354
The future of AI inference is rack scale. A full rack of SMC X-14 servers powered by d-Matrix Corsair™ accelerators and connected with JetStream™ networking. Purpose-built for high-performance, low-latency, power-efficient AI inference. Real infrastructure. Real engineering. Real systems. And yes...real racks have wires. 😉 #AI #AIInference #RackScale #AIInfrastructure #HPC #DataCenter #dMatrix
3
26
1,790
We're humbled to share that Corsair just won the AI Processor Innovation Award as part of this year's AI Breakthrough Awards. The industry is moving swiftly to heterogeneous compute. We built d-Matrix exactly for this shift and have begun shipping in volume. 🔗 d-matrix.ai/announcements/d-… #AI #Inference
2
6
9
717
Corsair™ is now in full production. Products are beginning to ship in volume to priority hyperscalers, neoclouds, and frontier AI labs. Thanks to our ecosystem partners TSMC, @AristaNetworks, @Broadcom, @Supermicro, @gimletlabs, and @AlchipTECH for helping make this milestone possible. Read more: d-matrix.ai/announcements/d-… #AIInference #AIInfrastructure #HeterogeneousCompute #DataCenterAI #Corsair
1
2
10
1,224
How do you build AI infrastructure when no single accelerator wins every workload? Join Satyam Srivastava (d-Matrix) and Tom St. John (Gimlet Labs) at TPC26 on June 3 in Baltimore for “Bold New World of Heterogeneous AI Computing.” tpc26.org/tpc26-sessions/ #TPC26 #AIInfrastructure #dMatrix #AIHardware #AI
1
1
3
604
Our CEO @sidsheth joined @GregShove on the Supercompanies podcast to talk about how we're not just building inference chips — we're using AI to do it. 🎧 piped.video/o5vVWphzmsw
3
404
AI coding agents are exploding in capability. The infrastructure behind them needs to evolve just as fast. GPU-only systems are increasingly hitting latency, efficiency, and scaling walls for agentic coding workloads. Heterogeneous and disaggregated inference pipelines change the equation: • Lower latency • Better efficiency • Speculative decoding acceleration • Scalable infrastructure for AI coding agents The future of AI coding requires purpose-built inference architectures. d-matrix.ai/where-heterogene… #AI #Inference #AgenticAI #LLM #AIInfrastructure #SpeculativeDecoding
2
268
Tomorrow at #IMAPSMemorySummit, Max (Sunghwan) Min from d-Matrix presents: “Low-Latency Digital In-Memory Compute Packaging for LLM AI Inference: SRAM, DRAM and Beyond” 🗓 May 14 | 4:35 PM 📍 Hyatt Regency, Santa Clara Come see how DIMC architecture is reshaping AI inference hardware. ⚡️ imaps.org/page/MemorySummitS… #AIInference #DIMC #dMatrix #Semiconductor
1
2
425
Building the future of AI inference takes a village. Fortunate to have @Infineon as a key partner on this journey. Together, we’re advancing low-latency, power-efficient AI inference infrastructure for the next generation of interactive AI. Read more: infineon.com/market-news/202… #AI #AIInference #GenerativeAI #LLM #DataCenter #Semiconductors #Inference #MachineLearning #PowerEfficiency #AIInfrastructure
3
258
May the throughput be with you. AI doesn’t win at training. It wins at inference. Every prompt. Every response. Every real-time decision. That’s where latency matters. That’s where efficiency matters. We’re building systems designed for this moment—high throughput, low latency, and power efficiency at scale. Built for inference. Ready for production. #AI #Inference #AIInfrastructure #MayThe4th
1
1
5
383
Join us at TiEcon 2026 this Thursday as our Founder & CEO, Sid Sheth, joins the Unicorn Panel to discuss what truly scales in AI today. From inference to infrastructure, the conversation is shifting from hype to real-world execution. 📍 Santa Clara Convention Center 📅 April 30, 2026 ⏰ 1:30–2:00 PM PST Register here: tiecon.org/speakers-2026/ #TiEcon2026 #AIInfrastructure #Inference #dMatrix
1
2
411
@dmatrixai's Chris Nicol joins the panel on what composable infrastructure really means for AI at scale. CCIB | April 29–30 ⤵️ See you there. #AI #AIInfrastructure #Composable #GenAI
3
231
AI inference is a pipeline, not a single workload. Speculative decoding paired with memory-centric architecture delivers lower latency, better GPU utilization, and faster responses at scale. The future of AI infrastructure is built by matching the right hardware to the right phase of inference. Read more: d-matrix.ai/how-speculative-… #AIInference #SpeculativeDecoding #GenAI #dMatrix
4
14
823
Proud to receive Andes Technology's Most Valuable Customer award at the RISC-V Now! conference today. The RISC-V ecosystem is a core part of how d-Matrix is building the fastest, most efficient inference accelerator for the Age of AI Inference — and Andes has been a key partner in making that possible. Thank you to Frankwell Lin and the entire Andes team. Photo (left to right): d-Matrix Founder & CEO Sid Sheth, Andes Chairman & CEO Frankwell Lin, and d-Matrix Chief AI SW Architect Satyam Srivastava. #AIInference #RISCV #LowLatency
4
367
We’re at RISC-V Now. Join Satyam Srivastava (d-Matrix) on solving low-latency, efficient inferencing in the datacenter. 4/21 | San Jose, CA Register: eventbrite.com/e/risc-v-now-… #RISCV #AIInference #AIInfrastructure #Datacenter #ML
1
248
Great to see our CEO Sid Sheth speaking at HumanX. “It’s not a money problem — it’s an architecture problem.” Scaling AI requires heterogeneous systems where specialized silicon works together to deliver real-time inference at scale. The Age of Inference is here. #AI #Inference #HumanX
1
1
7
445
Today at HumanX, d-Matrix CEO Sid Sheth will share why the GPU-only era is ending and what comes next for AI inference infrastructure. “The GPU-Only Era Is Over” April 9 | 2:25–2:40 PM Grove Theater humanx.co/speakers/sid-sheth #HumanX #AIInference
1
1
4
319
Today we announced the acquisition of GigaIO’s data center business, expanding d-Matrix’s rack-scale AI infrastructure capabilities. AI has entered the Age of Inference. Inference is now a systems problem — and we’re building what comes next. d-matrix.ai/announcements/ac… #AIInference #AIInfrastructure #AgeOfInference
2
6
288
Load more