SC22 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Birds of a Feather Archive

A Look into the Compute Express Link™ (CXL™) Device Ecosystem

Authors: Kurtis Bowman (Advanced Micro Devices (AMD) Inc), Sandeep Dattaprasad (Astera Labs), Cheolmin Park (Samsung), Jerry Lotto (Synopsys)

Abstract: Compute Express Link™ (CXL™) maintains memory coherency between the CPU memory space and memory on CXL attached devices. CXL enables a high-speed, efficient interconnect between the CPU, platform enhancements, and workload accelerators such as GPUs, FPGAs, and other purpose-built accelerator solutions.

This BoF session will feature a panel of experts from the CXL Consortium to discuss available CXL devices and what devices the industry can expect to see in the next year. The experts will also explore the new features in the CXL 3.0 specification and the new usage models it will enable.

Long Description: Compute Express Link™ (CXL™) maintains memory coherency between the CPU memory space and memory on CXL attached devices. This enables fine-grained resource sharing for higher performance with heterogeneous processing, memory disaggregation, memory pooling, persistent memory, and emerging memory media. Last year at SC’21, CXL Consortium members showcased the first public demonstrations of CXL, proving to the industry that CXL’s vision to enable a new ecosystem of high-performance, heterogeneous computing is now a reality. There were also multi-vendor demos to illustrate interoperability between vendor solutions.

As CXL products are nearing release into the market, with the completion of the first CXL Consortium Pre-FYI compliance event, the panel of experts will discuss the type of CXL product devices that are currently available today and the CXL devices we expect to see in the next year. The panelists will also introduce the newly released CXL 3.0 specification, which expands capabilities to increase scale and optimize resource utilization and examine the usage models CXL 3.0 will enable.

We invite the HPC community to engage directly with the CXL experts behind the technology and the solutions to get their questions answered about this emerging ecosystem as well as to provide their view to the Consortium on how the specification should continue to evolve to meet the community’s requirements. Attendees will gain deeper insight into the CXL device ecosystem and how CXL can solve the data demands from emerging applications such as Artificial Intelligence and Machine Learning.


Back to Birds of a Feather Archive Listing