Skip to content

Releases: oneapi-src/oneCCL

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.12

27 Mar 12:52
Compare
Choose a tag to compare

What's New

  • Performance improvements for scaleup for all message sizes for AllReduce, Allgather, and Reduce_Scatter.
  • Optimizations also include small message sizes that appear in inference apps.
  • Performance improvements for scaleout for Allreduce, Reduce, Allgather, and Reduce_Scatter.
  • Optimized memory usage of oneCCL.
  • Support for PMIx 4.2.6.
  • Bug fixes.

Removals

  • oneCCL 2021.12 removes support for PMIx 4.2.2

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.11.2

21 Dec 18:16
8d18c7b
Compare
Choose a tag to compare

This update provides bug fixes to maintain driver compatibility for Intel® Data Center GPU Max Series.

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.11.1

29 Nov 15:50
Compare
Choose a tag to compare

This update addresses stability issues with distributed Training and Inference workloads on Intel® Data Center GPU Max Series.

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.11

28 Nov 16:07
Compare
Choose a tag to compare
  • Added point to point blocking communication operations for send and receive.

  • Performance optimizations for Reduce-Scatter.

  • Improved profiling with Intel® Instrumentation and Tracing Technology (ITT) profiling level.

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.10

20 Jul 09:50
b4c31ba
Compare
Choose a tag to compare
  • Improved scaling efficiency of the Scaleup algorithms for ReduceScatter
  • Optimized performance of oneCCL scaleup collectives by utilizing the new embedded Data Streaming Accelerator in Intel® 4th Generation Xeon Scalable Processors (formerly code-named Sapphire Rapids)

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.9

11 Apr 14:29
0db8329
Compare
Choose a tag to compare

• Optimizations across the board including improved scaling efficiency of the Scaleup algorithms for Alltoall and Allgather
• Add collective selection for scaleout algorithm for device (GPU) buffers

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.8

19 Dec 15:44
bfa1e99
Compare
Choose a tag to compare

• Provides optimized performance for Intel® Data Center GPU Max Series utilizing oneCCL.
• Enables support for Allreduce, Allgather, Reduce, and Alltoall connectivity for GPUs on the same node.

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.7.1

01 Dec 17:33
debdc21
Compare
Choose a tag to compare

Intel® oneAPI Collective Communications Library 2021.7.1 has been updated to include functional and security updates. Users should update to the latest version as it becomes available.

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.7

13 Oct 13:32
9e66a9e
Compare
Choose a tag to compare

Intel(R) oneAPI Collective Communications Library (oneCCL) 2021.6

26 Aug 16:06
0ab9bf9
Compare
Choose a tag to compare
  • Added support for Intel® ITT (Intel® Instrumentation and Tracing Technology) profiling
  • Added WSL2 (Windows Subsystem for Linux 2) support
  • Moved dependency check for Level Zero from compile-time to runtime
  • Bug fixes