You might like
We had a great week at #MUG25!
Our second talk of #MUG25 Day 1 comes from @enfabrica_ 's own Boris Shpolyansky as he talks about shared memory pool usage for #AI/#ML/#AIinference workloads! #enfabrics #DeepLearning #MVAPICH #Ethernet #SharedMemory #CXL #GPU
And that's a wrap #MUG25 @mvapich! @SHRIJEET2
Shrijeet Mukherjee of @enfabrica_ continues the topic of networks for #HPC/#ML/#AI Workloads with his presentation of Enfabrica's Accelerated Compute Fabric and network and bandwidth elasticity! #HPC_Networks #MUG25 #DeepLearning #AI_Networking #MPI #MVAPICH #MUG2025
Our second talk of #MUG25 Day 1 comes from @enfabrica_ 's own Boris Shpolyansky as he talks about shared memory pool usage for #AI/#ML/#AIinference workloads! #enfabrics #DeepLearning #MVAPICH #Ethernet #SharedMemory #CXL #GPU
🚀 Excited to share that Enfabrica will be at the AI Infra Summit! 📍 Visit us at Booth #622 to explore how Enfabrica’s approach to memory and networking unlocks the performance needed for the next wave of AI. We look forward to connecting with innovators, builders, and…
Happy Friday! "Enfabrica's new Ethernet-based AI memory fabric system drops AI inference cost per user per token by up to 50%." Another great piece of coverage on our EMFASYS announcement. Thank you @duanebenson at @AllAboutCircuit for the write up. Read the full article…
It’s been an exciting week here at Enfabrica. On Tuesday, we proudly unveiled EMFASYS, our industry-first AI memory fabric system, and we’ve been energized by the response. Huge thanks to Timothy Prickett Morgan over at @TheNextPlatform for covering the announcement and helping…
Enfabrica says that its CXL memory godbox can cut the cost of AI inference in half on Nvidia GPU systems. $NVDA #Enfabrica nextplatform.com/2025/07/29/ski…
LLM inference clouds, you now have an ace in the hole. Enfabrica is announcing today the release of EMFASYS, the industry’s first RDMA Ethernet-based AI memory fabric system, built to dramatically improve the token economics of large-scale, agentic, memory-bound AI inference…
🚀 ACF SuperNIC, meet momentum. Enfabrica’s 3.2 Tbps ACF-S chip “Millennium” has 4 times the I/O bandwidth of today’s highest performing AI backend NIC…and it’s redefining how data movement can be accelerated in both training and inference clusters. This is more than a chip —…
Enfabrica's ACF SuperNIC silicon ("Millennium") is in the house! And already: - it has been built into a dozen system/rack setups for software validation and GPU server benchmarking - it has passed 100 million packets through its full-chip SuperNIC datapath between PCIe <> ACF-S…
🎉 Enfabrica has been named to the #Futuriom50 for the second year in a row! 🎉 We are happy to be named as one of the strongest private companies in cloud and communications infrastructure. A huge thank you to our incredible team and congratulations to all the other companies…
The #Futuriom50 is LIVE Our annual report on top trends and private companies dives into #AI infrastructure, #distributedcloud, #cloudsecurity, #SASE, and more. futuriom.com/articles/news/…
#ICYMI We are expanding in Hyderabad, India! We are happy to announce the opening of Enfabrica (India) Private Limited and its new R&D Hub based in Hyderabad. The opening of the new office and R&D center in a critical technology and AI innovation hub is an important investment…
In case you missed our talk at @hotchipsorg, catch the replay of Enfabrica's presentation in the Networking Silicon track, "ACF-S: An 8 Terabit/sec SuperNIC for High-Performance Data Movement in AI and Accelerated Compute Networks." Shrijeet Mukherjee and Thomas Norrie took the…
Hot Chips 2024 presentations are now publicly available at hc2024.hotchips.org
Today, we are happy to announce two transformative updates: 🔹 The General Availability of Enfabrica’s ACF SuperNIC — Our groundbreaking chip is redefining GPU networking for AI and accelerated computing. With support for scaling 500K+ GPU clusters and delivering unmatched…
We had a very successful Supercomputing '23 this week in Denver. Customers and partners all interacted with our ACF-S composer and created their own blueprint for a supercomputer with our technology. Check it out here: bit.ly/3utNa4R
youtube.com
YouTube
Enfabrica's ACF-S Supercomputer Composer
Courtesy of @Nasdaq - a live shot of Times Square just this past Friday! To all our incredible employees, partners and investors, thank you for your role in getting @enfabrica_ to where we are today. We couldn't do it without you and this is just the beginning!
Enfabrica looks really cool. blog.enfabrica.net/press-release-…
blog.enfabrica.net
Press Release: Networking Chip Startup Enfabrica Emerges from Stealth Mode to Solve Scalability and…
Reveals Groundbreaking Accelerated Compute Fabric Devices and Solutions at MemCon 2023, Enabling 50% Lower Cost of Compute and 50X Memory…
[Merci de retweeter] Est ce que les personnes qui aimeraient prolonger plus de deux unités nucléaires en Belgique pourraient l'écrire sur ce fil Twitter ? Moi je pense qu'il faut certainement prolonger aussi Tihange 1 et envisager la prolongation de Tihange 2 et Doel 3.
He has done excellent research work to design new neural network architectures that can adapt to new tasks and that have long-term memory capabilities, see for example: hdl.handle.net/2268/230984 hdl.handle.net/2268/248560 2/2
Et pour en rajouter une couche, elle a l'audace d'insinuer que sortir du nucléaire participe à la neutralité climatique. Je n'ai qu'une seule chose à vous dire : vous me faites honte, vous faites la honte de tout un pays. 2/2
From all of us at GigaIO, happy holidays, and a New Year filled with peace and joy (and compostable computing😉).
United States Trends
- 1. Texans 77,8 B posts
- 2. Aaron Rodgers 31,2 B posts
- 3. Tomlin 23,3 B posts
- 4. CJ Stroud 17,9 B posts
- 5. Arthur Smith 4.992 posts
- 6. #HereWeGo 20,7 B posts
- 7. Christian Kirk 3.745 posts
- 8. #HOUvsPIT 3.764 posts
- 9. #HTownMade 4.530 posts
- 10. Nico Collins 3.006 posts
- 11. Anderson 31,3 B posts
- 12. Woody Marks 2.338 posts
- 13. Jonnu Smith 2.702 posts
- 14. TJ Watt 2.367 posts
- 15. Muth 1.487 posts
- 16. Sheldon Rankins 1.619 posts
- 17. DK Metcalf 2.839 posts
- 18. Echols 2.063 posts
- 19. Davis Mills 1.310 posts
- 20. Pick 6 14,8 B posts
You might like
Something went wrong.
Something went wrong.