Major Features and Improvements
Intel® Extension for TensorFlow* extends the official TensorFlow capabilities, allowing TensorFlow workloads to run on Intel® Data Center GPU Max Series, Intel® Data Center GPU Flex Series, and Intel® Xeon® Scalable Processors. This release includes the following major features and improvements:
-
New Install Channel: New install channel is provided, to solve the package size limitation of Pypi.
pip install --upgrade intel-extension-for-tensorflow[xpu] -f https://developer.intel.com/itex-whl-weekly -
Toolkit Support: Supports Intel® oneAPI Base Toolkit 2024.2.
-
Updated Support: The Intel® Extension for TensorFlow* has been upgraded to support oneDNN 3.4.3.
-
Expreimental Support: Continues to provide experimental support for Intel® Arc™ A-Series GPUs on Windows Subsystem for Linux 2 with Ubuntu Linux installed and native Ubuntu Linux.
Bug Fixes
- Fixes device memory leak issues exposed by
ZeroLike,SetOneDnnLayout,GetDeviceInfoandSegmentReduce. - Fixes potential host memory leak issue.
- Fixes accurancy issue exposed by
Softmax. - Fixes performance regression issue exposed by
AddV2WithSoftmax. - Fixes
SYCL ESIMD feature not support on hostissue.
Known Issues
- TensorList limitation: TensorList is not supported with NextPluggableDevice by TensorFlow 2.15.
- Allocation limitation of WSL: A maximum size of single allocation allowed on a single device is set on the Windows Subsystem for Linux (WSL2), which may cause Out-of-Memory error. Users can remove the limitation with environment variable
UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1 - FP64 support: FP64 is not natively supported by the Intel® Data Center GPU Flex Series platform. If you run any AI workload with the FP64 kernel on that platform, the workload will exit with an exception as
'XXX' Op uses fp64 data type, while fp64 instructions are not supported on the platform. - GLIBC++ mismatch: A
GLIBC++version mismatch may cause a workload exit with the exception,Can not find any devices. To check runtime environment on your host, please run itex/tools/python/env_check.py.Try running env_check.py script to confirm.