Skip to content

AcceleratorSelection

Source code in docs_src/capo_deadline/types/accelerator_selection.py 
14
15
16
17
18
class AcceleratorSelection(TypedDict, closed=True):
    name: "capo_deadline.types.accelerator_name.AcceleratorName"
    """<p>The name of the chip used by the GPU accelerator.</p> <p>The available GPU accelerators are:</p> <ul> <li> <p> <code>t4</code> - NVIDIA T4 Tensor Core GPU (16 GiB memory)</p> </li> <li> <p> <code>a10g</code> - NVIDIA A10G Tensor Core GPU (24 GiB memory)</p> </li> <li> <p> <code>l4</code> - NVIDIA L4 Tensor Core GPU (24 GiB memory)</p> </li> <li> <p> <code>l40s</code> - NVIDIA L40S Tensor Core GPU (48 GiB memory)</p> </li> <li> <p> <code>rtx-pro-server-6000</code> - NVIDIA RTX PRO Server 6000 GPU (96 GiB memory)</p> </li> </ul>"""
    runtime: "capo_deadline.types.accelerator_runtime.AcceleratorRuntime"
    """<p>Specifies the runtime driver to use for the GPU accelerator. You must use the same runtime for all GPUs in a fleet. </p> <p>You can choose from the following runtimes:</p> <ul> <li> <p> <code>latest</code> - Use the latest runtime available for the chip. If you specify <code>latest</code> and a new version of the runtime is released, the new version of the runtime is used.</p> </li> <li> <p> <code>grid:r580</code> - <a href="https://docs.nvidia.com/vgpu/19.0/index.html">NVIDIA vGPU software 19</a> </p> </li> <li> <p> <code>grid:r570</code> - <a href="https://docs.nvidia.com/vgpu/18.0/index.html">NVIDIA vGPU software 18</a> </p> </li> <li> <p> <code>grid:r535</code> - <a href="https://docs.nvidia.com/vgpu/16.0/index.html">NVIDIA vGPU software 16</a> </p> </li> </ul> <p>If you don't specify a runtime, Amazon Web Services Deadline Cloud uses <code>latest</code> as the default. However, if you have multiple accelerators and specify <code>latest</code> for some and leave others blank, Amazon Web Services Deadline Cloud raises an exception.</p> <important> <p>Not all runtimes are compatible with all accelerator types:</p> <ul> <li> <p> <code>t4</code> and <code>a10g</code>: Support all runtimes (<code>grid:r580</code>, <code>grid:r570</code>, <code>grid:r535</code>)</p> </li> <li> <p> <code>l4</code> and <code>l40s</code>: Only support <code>grid:r570</code> and newer</p> </li> <li> <p> <code>rtx-pro-server-6000</code>: Only supports <code>grid:r580</code> </p> </li> </ul> <p>All accelerators in a fleet must use the same runtime version. You cannot mix different runtime versions within a single fleet.</p> </important> <note> <p>When you specify <code>latest</code>, it resolves to <code>grid:r580</code> for all currently supported accelerators.</p> </note>"""

name instance-attribute 

name: AcceleratorName

The name of the chip used by the GPU accelerator.

The available GPU accelerators are:

  • t4 - NVIDIA T4 Tensor Core GPU (16 GiB memory)

  • a10g - NVIDIA A10G Tensor Core GPU (24 GiB memory)

  • l4 - NVIDIA L4 Tensor Core GPU (24 GiB memory)

  • l40s - NVIDIA L40S Tensor Core GPU (48 GiB memory)

  • rtx-pro-server-6000 - NVIDIA RTX PRO Server 6000 GPU (96 GiB memory)

runtime instance-attribute 

runtime: AcceleratorRuntime

Specifies the runtime driver to use for the GPU accelerator. You must use the same runtime for all GPUs in a fleet.

You can choose from the following runtimes:

If you don't specify a runtime, Amazon Web Services Deadline Cloud uses latest as the default. However, if you have multiple accelerators and specify latest for some and leave others blank, Amazon Web Services Deadline Cloud raises an exception.

Not all runtimes are compatible with all accelerator types:

  • t4 and a10g: Support all runtimes (grid:r580, grid:r570, grid:r535)

  • l4 and l40s: Only support grid:r570 and newer

  • rtx-pro-server-6000: Only supports grid:r580

All accelerators in a fleet must use the same runtime version. You cannot mix different runtime versions within a single fleet.

When you specify latest, it resolves to grid:r580 for all currently supported accelerators.