• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

d-Matrix Announces $110 Million in Funding for Corsair Inference Compute Platform

T0@st

News Editor
Joined
Mar 7, 2023
Messages
2,077 (3.43/day)
Location
South East, UK
d-Matrix, the leader in high-efficiency generative AI compute for data centers, has closed $110 million in a Series-B funding round led by Singapore-based global investment firm Temasek. The goal of the fundraise is to enable d-Matrix to begin commercializing Corsair, the world's first Digital-In Memory Compute (DIMC), chiplet-based inference compute platform, after the successful launches of its prior Nighthawk, Jayhawk-I and Jayhawk II chiplets.

d-Matrix's recent silicon announcement, Jayhawk II, is the latest example of how the company is working to fundamentally change the physics of memory-bound compute workloads common in generative AI and large language model (LLM) applications. With the explosion of this revolutionary technology over the past nine months, there has never been a greater need to overcome the memory bottleneck and current technology approaches that limit performance and drive up AI compute costs.




d-Matrix has architected an elegant DIMC engine and chiplet-based solution to enable inference at a lower total cost of ownership (TCO) than GPU-based alternatives. This new chiplet-based DIMC platform coming to market in 2024 will redefine the category, further positioning d-Matrix as the frontrunner in efficient AI inference. "The current trajectory of AI compute is unsustainable as the TCO to run AI inference is escalating rapidly," said Sid Sheth, co-founder and CEO at d-Matrix. "The team at d-Matrix is changing the cost economics of deploying AI inference with a compute solution purpose-built for LLMs, and this round of funding validates our position in the industry."

"d-Matrix is the company that will make generative AI commercially viable," said Sasha Ostojic, Partner at Playground Global. "To achieve this ambitious goal, d-Matrix produced an innovative dataflow architecture, assembled into chiplets, connected with a high-speed interface, and driven by an enterprise-class scalable software stack. Playground couldn't be more excited and proud to back Sid and the d-Matrix team as it fulfills the demand from eager customers in desperate need of improved economics."

"We're entering the production phase when LLM inference TCO becomes a critical factor in how much, where, and when enterprises use advanced AI in their services and applications," said Michael Stewart from M12, Microsoft's Venture Fund. "d-Matrix has been following a plan that will enable industry-leading TCO for a variety of potential model service scenarios using a flexible, resilient chiplet architecture based on a memory-centric approach."

d-Matrix was founded in 2019 to solve the memory-compute integration problem, which is the final frontier in AI compute efficiency. d-Matrix has invested in groundbreaking chiplet and digital in-memory compute technologies with the goal of bringing to market a high-performance, cost-effective inference solution in 2024. Since its inception, d-Matrix has grown substantially in headcount and office space. They are headquartered in Santa Clara, California with offices in Bengaluru, India and Sydney, Australia. With this Series-B funding, d-Matrix plans to invest in recruitment and commercialization of its product to satisfy the immediate customer need for lower cost, more efficient compute infrastructure for generative AI inference.

About d-Matrix
d-Matrix is a leading supplier of Digital In-Memory Computing (DIMC) solutions that address the growing demand for transformer and generative AI inference acceleration. d-Matrix creates flexible solutions for inference at scale using innovative DIMC circuit techniques, a chiplet-based architecture, high-bandwidth BoW (chiplet) interconnects and a full stack of machine learning and large language model tools and software. Founded in 2019, the company is backed by top investors and strategic partners including Playground Global, M12 (Microsoft Venture Fund), SK Hynix, Nautilus Venture Partners, Marvell Technology and Entrada Ventures.

Visit d-matrix.ai for more information and follow d-Matrix on LinkedIn for the latest updates.

View at TechPowerUp Main Site | Source
 
Joined
Jul 8, 2022
Messages
251 (0.30/day)
Location
USA
Processor i9-11900K
Motherboard Asus ROG Maximus XIII Hero
Cooling Arctic Liquid Freezer II 360
Memory 4x8GB DDR4
Video Card(s) Alienware RTX 3090 OEM
Storage OEM Kioxia 2tb NVMe (OS), 4TB WD Blue HDD (games)
Display(s) LG 27GN950-B
Case Lian Li Lancool II Mesh Performance (black)
Audio Device(s) Logitech Pro X Wireless
Power Supply Corsair RM1000x
Keyboard HyperX Alloy Elite 2
Excellent - another card to add to my collection when prices go way down in 10 years! I can't wait!
 
Joined
Jan 5, 2006
Messages
18,585 (2.70/day)
System Name AlderLake
Processor Intel i7 12700K P-Cores @ 5Ghz
Motherboard Gigabyte Z690 Aorus Master
Cooling Noctua NH-U12A 2 fans + Thermal Grizzly Kryonaut Extreme + 5 case fans
Memory 32GB DDR5 Corsair Dominator Platinum RGB 6000MT/s CL36
Video Card(s) MSI RTX 2070 Super Gaming X Trio
Storage Samsung 980 Pro 1TB + 970 Evo 500GB + 850 Pro 512GB + 860 Evo 1TB x2
Display(s) 23.8" Dell S2417DG 165Hz G-Sync 1440p
Case Be quiet! Silent Base 600 - Window
Audio Device(s) Panasonic SA-PMX94 / Realtek onboard + B&O speaker system / Harman Kardon Go + Play / Logitech G533
Power Supply Seasonic Focus Plus Gold 750W
Mouse Logitech MX Anywhere 2 Laser wireless
Keyboard RAPOO E9270P Black 5GHz wireless
Software Windows 11
Benchmark Scores Cinebench R23 (Single Core) 1936 @ stock Cinebench R23 (Multi Core) 23006 @ stock
Getting tired of "AI" news......
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,486 (2.40/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Getting tired of "AI" news......
Maybe try emailing Jen-Hsun Huang and see if he can do something about it?
A new proprietary Nvidia name is the least he should be able to provide.
 
Joined
Sep 15, 2021
Messages
169 (0.15/day)
It's not that irrelevant because if nVidia has real competition in AI focused hardware, it could result in them not being able to cruise by and make all the money in that market, thus increasing the importance of gaming focused hardware for them and possibly result in better pricing/performance for hardware enjoyed by the average TPU enjoyer.
 
Joined
Apr 13, 2022
Messages
1,141 (1.22/day)
It's not that irrelevant because if nVidia has real competition in AI focused hardware, it could result in them not being able to cruise by and make all the money in that market, thus increasing the importance of gaming focused hardware for them and possibly result in better pricing/performance for hardware enjoyed by the average TPU enjoyer.

Gaming with dedicated physical hardware on PCs it's already been killed the corpse just hasn't started to stink yet. But it is DONE. Consoles will kick about for a bit longer and in the case of Nintendo maybe a lot longer. But for the PC owning anything is long since dead and we are just in the last cash grab before it.

PC gaming is not the master race or a special snow flake, it's been the leading toxin of all that's bad in gaming and pushed those trends forward. It's going to happen, and there is nothing any of us can do about it. And as normal when PC gaming pushes cloud onto the rest of gaming people will scream master race as PC once again ruins things for everyone.
 
Joined
Oct 6, 2021
Messages
1,605 (1.43/day)
Gaming with dedicated physical hardware on PCs it's already been killed the corpse just hasn't started to stink yet. But it is DONE. Consoles will kick about for a bit longer and in the case of Nintendo maybe a lot longer. But for the PC owning anything is long since dead and we are just in the last cash grab before it.

PC gaming is not the master race or a special snow flake, it's been the leading toxin of all that's bad in gaming and pushed those trends forward. It's going to happen, and there is nothing any of us can do about it. And as normal when PC gaming pushes cloud onto the rest of gaming people will scream master race as PC once again ruins things for everyone.
nJust a year with a lot of abnormally bad AAA games and people start commenting on these desperate outbursts... Cloud Gaming is doomed to failure or a very small niche.

Master Race is alive and stronger (also expensive) than ever. lol
 
Joined
Dec 29, 2010
Messages
3,788 (0.75/day)
Processor AMD 5900x
Motherboard Asus x570 Strix-E
Cooling Hardware Labs
Memory G.Skill 4000c17 2x16gb
Video Card(s) RTX 3090
Storage Sabrent
Display(s) Samsung G9
Case Phanteks 719
Audio Device(s) Fiio K5 Pro
Power Supply EVGA 1000 P2
Mouse Logitech G600
Keyboard Corsair K95
It's not that irrelevant because if nVidia has real competition in AI focused hardware, it could result in them not being able to cruise by and make all the money in that market, thus increasing the importance of gaming focused hardware for them and possibly result in better pricing/performance for hardware enjoyed by the average TPU enjoyer.
Dojo is gonna surprise Jensen. The irony is that Nv gpus are not even that great at NN, but its just the best we have. And thus everyone whose anyone in AI is compute constrained. It's so bad that Tesla went out and designed their own silicon for NN, aka Dojo which is ramping up now. And Tesla knows gpus with them bringing a 10K H100 cluster online recently. And the kicker is they are still compute constrained lol, hence Dojo.
 

Toothless

Tech, Games, and TPU!
Supporter
Joined
Mar 26, 2014
Messages
9,567 (2.47/day)
Location
Washington, USA
System Name Veral
Processor 7800x3D
Motherboard x670e Asus Crosshair Hero
Cooling Corsair H150i RGB Elite
Memory 2x32 Corsair Dominator
Video Card(s) Powercolor 7900XTX Red Devil
Storage Crucial P5 Plus 1TB, Samsung 980 1TB, Teamgroup MP34 4TB
Display(s) Acer Nitro XZ342CK Pbmiiphx, 2x AOC 2425W, AOC I1601FWUX
Case Fractal Design Meshify Lite 2
Audio Device(s) Blue Yeti + SteelSeries Arctis 5 / Samsung HW-T550
Power Supply Corsair HX850
Mouse Corsair Nightsword
Keyboard Corsair K55
VR HMD HP Reverb G2
Software Windows 11 Professional
Benchmark Scores PEBCAK
AI is useless, people think it will revolutionize healthcare, for the rich it will, for people like me, my health insurance company will deny all my claims until I am bent over in excruciating pain screaming, as happened earlier this year, my Doctor tried like hell, but Doctors and AI have no power in the realm of Greed and Envy

fuck'em all boys.
Plenty of other applications for AI..
 
Joined
Dec 29, 2010
Messages
3,788 (0.75/day)
Processor AMD 5900x
Motherboard Asus x570 Strix-E
Cooling Hardware Labs
Memory G.Skill 4000c17 2x16gb
Video Card(s) RTX 3090
Storage Sabrent
Display(s) Samsung G9
Case Phanteks 719
Audio Device(s) Fiio K5 Pro
Power Supply EVGA 1000 P2
Mouse Logitech G600
Keyboard Corsair K95
for rich people too?
I didn't respond on your rant but you're barking up the wrong tree. Just because kids are cheating using chatbots doesn't mean AI is useless or for rich people. Poor form man.
 
Joined
Dec 29, 2010
Messages
3,788 (0.75/day)
Processor AMD 5900x
Motherboard Asus x570 Strix-E
Cooling Hardware Labs
Memory G.Skill 4000c17 2x16gb
Video Card(s) RTX 3090
Storage Sabrent
Display(s) Samsung G9
Case Phanteks 719
Audio Device(s) Fiio K5 Pro
Power Supply EVGA 1000 P2
Mouse Logitech G600
Keyboard Corsair K95
care to give me a specific example where it has helped a poor person with sources provided?

Cap Lying GIF by Twitch
Are you seriously an educator?
 
Joined
Aug 14, 2013
Messages
2,373 (0.58/day)
System Name boomer--->zoomer not your typical millenial build
Processor i5-760 @ 3.8ghz + turbo ~goes wayyyyyyyyy fast cuz turboooooz~
Motherboard P55-GD80 ~best motherboard ever designed~
Cooling NH-D15 ~double stack thot twerk all day~
Memory 16GB Crucial Ballistix LP ~memory gone AWOL~
Video Card(s) MSI GTX 970 ~*~GOLDEN EDITION~*~ RAWRRRRRR
Storage 500GB Samsung 850 Evo (OS X, *nix), 128GB Samsung 840 Pro (W10 Pro), 1TB SpinPoint F3 ~best in class
Display(s) ASUS VW246H ~best 24" you've seen *FULL HD* *1O80PP* *SLAPS*~
Case FT02-W ~the W stands for white but it's brushed aluminum except for the disgusting ODD bays; *cries*
Audio Device(s) A LOT
Power Supply 850W EVGA SuperNova G2 ~hot fire like champagne~
Mouse CM Spawn ~cmcz R c00l seth mcfarlane darawss~
Keyboard CM QF Rapid - Browns ~fastrrr kees for fstr teens~
Software integrated into the chassis
Benchmark Scores 9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999
Joined
Feb 11, 2020
Messages
239 (0.14/day)
That didn't take long for the custom replacements to show up. It's the Bitcoin saga all over again. Now to watch GPU demand start to weaken.
 
Joined
Dec 29, 2010
Messages
3,788 (0.75/day)
Processor AMD 5900x
Motherboard Asus x570 Strix-E
Cooling Hardware Labs
Memory G.Skill 4000c17 2x16gb
Video Card(s) RTX 3090
Storage Sabrent
Display(s) Samsung G9
Case Phanteks 719
Audio Device(s) Fiio K5 Pro
Power Supply EVGA 1000 P2
Mouse Logitech G600
Keyboard Corsair K95
That didn't take long for the custom replacements to show up. It's the Bitcoin saga all over again. Now to watch GPU demand start to weaken.
psst... :laugh:



 
Joined
Jun 18, 2021
Messages
2,534 (2.06/day)
but its just the best we have

*it's just the most cost effective thing we have.

There are better devices but somewhat generic gpus are the most cost effective all things considered (software stack, supply availability, able to be repurposed into something else - not even talking about dumping consumer cards on the used market, the big data center gpu chips can also be partitioned and made available as generic cloud compute).
 
Top