Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[SYCL][NFC][Doc] Update sycl_ext_intel_device_info backend limitations #16792

Merged
merged 2 commits into from
Feb 6, 2025
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 1 addition & 16 deletions sycl/doc/extensions/supported/sycl_ext_intel_device_info.md
Original file line number Diff line number Diff line change
Expand Up @@ -62,7 +62,7 @@ The device ID can be obtained using the standard get\_info() interface.

A new device descriptor will be added which will provide the device Universal Unique ID (UUID).

This new device descriptor is currently only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior would be to expose the UUIDs of all supported devices which enables detection of total number of unique devices.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved
The DPC++ default behavior would be to expose the UUIDs of all supported devices which enables detection of total number of unique devices.


## Version ##
Expand Down Expand Up @@ -100,11 +100,6 @@ The UUID can be obtained using the standard get\_info() interface.

A new device descriptor will be added which will provide the PCI address in BDF format. BDF format contains the address as: `domain:bus:device.function`.

This new device descriptor is only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved

**Note:** The environment variable SYCL\_ENABLE\_PCI must be set to 1 to obtain the PCI address.


## Version ##

All versions of the extension support this query.
Expand Down Expand Up @@ -140,8 +135,6 @@ The PCI address can be obtained using the standard get\_info() interface.

A new device descriptor will be added which will provide the physical SIMD width of an execution unit on an Intel GPU. This data will be used to calculate the computational capabilities of the device.

This new device descriptor is only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.


## Version ##

Expand Down Expand Up @@ -180,8 +173,6 @@ A new device descriptor will be added which will provide the number of execution

This new device descriptor will provide the same information as "max\_compute\_units" does today. We would like to have an API which is specific for Intel GPUs.

This new device descriptor is only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.


## Version ##

Expand Down Expand Up @@ -355,8 +346,6 @@ Then the number of hardware threads per EU can be obtained using the standard ge

A new device descriptor will be added which will provide the maximum memory bandwidth. If the device is a subdevice, then the maximum bandwidth of the subdevice is returned.

This new device descriptor is only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved


## Version ##

Expand Down Expand Up @@ -397,8 +386,6 @@ Beware that when other processes or threads are using this device when this call
is made, the value it returns may be stale even before it is returned to the
caller.

This new device descriptor is only available for devices in the Level Zero platform, and the matching aspect is only true for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved


## Version ##

Expand Down Expand Up @@ -435,7 +422,6 @@ Then the free device memory can be obtained using the standard get\_info() inte

A new device descriptor is added which provides the maximum clock rate of device's global memory.

This new device descriptor is not available for devices in the OpenCL platform, and the matching aspect is false for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved

## Version ##

Expand Down Expand Up @@ -472,7 +458,6 @@ Then the memory clock rate can be obtained using the standard get\_info() interf

A new device descriptor is added which provides the maximum bus width between device and memory.

This new device descriptor is not available for devices in the OpenCL platform, and the matching aspect is false for those devices. The DPC++ default behavior is to expose GPU devices through the Level Zero platform.
KseniyaTikhomirova marked this conversation as resolved.
Show resolved Hide resolved

## Version ##

Expand Down