Compute Express Link

Compute Express Link (CXL) is a new high-speed CPU-to-Device and CPU-to-Memory interconnect designed to accelerate next-generation data center performance.

[Compute Express Link @ComputeExLink]

Join the #CXLConsortium at #SC26 in Chicago for our CXL Exhibitor Forum featuring Anil Godbole (@Intel) on November 17 from 1:30–2:00 p.m. CST. Learn more about #ComputeExpressLink (#CXL) and its role in advancing memory and computing architectures: https://t.co/VEzdvi1Q1Y

[Compute Express Link @ComputeExLink]

How can shared KV Cache improve multi-GPU LLM inference? Join the #CXLConsortium and @MicronTech to explore how #ComputeExpressLink (#CXL) fabric-attached memory enables KV Cache sharing across GPUs. Register for the webinar: https://t.co/PIIgaIN1wx

[Compute Express Link @ComputeExLink]

#CXL 4.0 brings key advancements for next-generation data centers, including doubled bandwidth to 128 GT/s without added latency, new bundled port capabilities, and enhanced memory RAS features. Explore the technology in the CXL 4.0 White Paper: https://t.co/VM9JhCT3Mc

[Compute Express Link @ComputeExLink]

Join #CXLConsortium and member company @MicronTech for a webinar on CXL fabric-attached memory (FAM), shared KV Cache, and practical insights from an open-source multi-GPU inference implementation. Register here: https://t.co/PIIgaIN1wx

[Compute Express Link @ComputeExLink]

#CXL Test Event 012 is set for Dec. 13–17 in South San Francisco, and registration is open for #CXLConsortium members! Join us for official CXL 2.0 and 3.x testing, interoperability evaluation, and the chance to apply for the Integrators List. Learn more: https://t.co/tFQwHkOEWc

[Compute Express Link @ComputeExLink]

AI, HPC, and data-intensive workloads are driving new demands on data center infrastructure. #CXL 4.0 delivers 128 GT/s bandwidth, new bundled port capabilities, enhanced memory RAS, and backward compatibility with previous CXL generations. Learn more: https://t.co/VM9JhCTBBK

[Compute Express Link @ComputeExLink]

#CXLConsortium is returning to @Supercomputing 2026 (#SC26), Nov. 15–20 in Chicago, to showcase how #ComputeExpressLink (#CXL) is advancing AI and HPC applications. Visit the CXL Pavilion (Booth 1218) for live CXL technology demonstrations. Register today: https://t.co/I74HBnwZS0

[Compute Express Link @ComputeExLink]

Join the #CXLConsortium and @SNIA today at the @SNIA Developer Conference for a session on unified fabric management, integrating #ComputeExpressLink (#CXL) Fabric Manager APIs with SNIA Swordfish. Hear from experts at @Micron and @Intel: https://t.co/ZNLjzFEKwJ

[Compute Express Link @ComputeExLink]

#CXLConsortium recently hosted @Meta representatives for a discussion on Vistara and scaling #ComputeExpressLink (#CXL) to millions of servers. Watch the webinar on demand to learn how CXL-based memory expansion can improve efficiency for AI workloads: https://t.co/mf3sVYLnEe

[Compute Express Link @ComputeExLink]

#CXLConsortium and @SNIA will present a unified approach to fabric management at @SNIASDC, integrating #CXL Fabric Manager APIs with SNIA Swordfish. Join John Groves and Richelle Ahlvers on Sept. 30 at 3:35 pm PT to learn more: https://t.co/ZNLjzFEKwJ

[Compute Express Link @ComputeExLink]

Join #CXLConsortium today at 4:10 p.m. at #AIInfraSummit2026 for “Scaling the AI Memory Wall” to hear from CXL Consortium Board Chair Dr. Debendra Das Sharma (@Intel) on advancing memory, architecture, and system design for AI infrastructure. Learn more: https://t.co/6MfSZTMRKD

[Compute Express Link @ComputeExLink]

Join #CXLConsortium tomorrow at 12:45 pm at #AIInfraSummit2026 for “Boost Your AI Performance With CXL,” featuring @Samsung, @AsteraLabs, @Meta, and Primemas to explore how #ComputeExpressLink (#CXL) enables scalable memory and efficient AI infrastructure: https://t.co/crF37THkVS

[Compute Express Link @ComputeExLink]

#ComputeExpressLink (#CXL) 4.0 delivers the performance and scalability needed for next-generation data centers, doubling bandwidth to 128 GT/s with no added latency while enhancing memory RAS and more. Learn more in the CXL 4.0 White Paper: https://t.co/VM9JhCT3Mc

[Compute Express Link @ComputeExLink]

Join the #CXLConsortium “Boost Your AI Performance With CXL” session at #AIInfraSummit2026 to see how CXL enables scalable memory expansion, memory pooling, and more efficient resource utilization across CPUs, GPUs, and accelerators. Learn more: https://t.co/crF37THSLq

[Compute Express Link @ComputeExLink]

#CXL continues to gain momentum across AI and data center infrastructure. New solutions showcased at #FMS2026 highlight how CXL is enabling memory expansion, pooling, and more flexible infrastructure. Read more from @Forbes: https://t.co/XsJL24RYd9

[Compute Express Link @ComputeExLink]

Heading to #AIInfraSummit 2026? #CXLConsortium members will be onsite in the #CXL Pavilion to share how CXL addresses the growing memory demands of #AI infrastructure, with demos from @WeAreLiqid, Primemas and @VIAVISolutions. Read our blog to learn more: https://t.co/crF37THkVS

[Compute Express Link @ComputeExLink]

Missed our latest #CXLConsortium webinar? Catch it on demand! Hear from CXL Consortium members Meta on Vistara, hyperscale memory architectures, and how CXL-based memory expansion can improve efficiency for AI & cloud workloads. Watch now: https://t.co/mf3sVYLnEe

[Compute Express Link @ComputeExLink]

Catch up on the #CXLConsortium’s activities at #FMS2026 by reading our recap blog post highlighting our Best of Show award win, our panel session covering KV Cache optimization, and #CXL 3.x ecosystem momentum. Read the full recap blog: https://t.co/C1qLMXEk4j

[Compute Express Link @ComputeExLink]

Missed #CXLMiniDevCon2026? Presentations from the event are now available for #CXLConsortium members to download. Explore the latest technical insights and innovations shaping the future of #ComputeExpressLink (#CXL), memory, and compute architectures: https://t.co/ob5CF3FGVy

[Compute Express Link @ComputeExLink]

#CXL 4.0 delivers the performance and scalability needed for next-generation data centers, doubling bandwidth to 128 GT/s with no added latency, while adding bundled port capabilities, enhanced memory RAS, and backward compatibility. Read the paper: https://t.co/VM9JhCTBBK