1

我希望让 qualcomm SNPE（金鱼草神经处理引擎）在我的 linux（不是 Android）板（flightPro w/qualcomm 820.）上工作。它在 cpu 上运行良好。

我已经成功地按照提供的示例将 alexnet 加载到我的 820 板上并在 cpu 模式下运行 snpe (snpe-net-run)。它不在 gpu 模式下运行。

搜索网络和论坛（例如，https ://developer.qualcomm.com/forum/qdn-forums/software/qualcomm-neural-processing-sdk/59207 ）似乎所有（？）linux板都可能缺少opencl完成这项工作所需的驱动程序。

按照例子...

> snpe-net-run --container bvlc_alexnet.dlc --input_list target_raw_list.txt --use_gpu 
The selected runtime is not available on this platform. Continue 
anyway to observe the failure at network creation time.
Aborted

我希望 gpu 能够工作（希望交叉手指比 cpu 快得多！）

4

1 回答 1

1

您需要咨询您的电路板供应商/制造商和您的 Linux BSP 提供商。

在SNPE 产品页面中，820 被列为受支持，但还提到 libOpenCL.so 必须存在于设备上（下面以粗体突出显示）。

Qualcomm 神经处理 SDK 支持 Qualcomm® Snapdragon™ 855、845、820、835、712、675、660、653、652、650、636、632、630、626、625、450、439 和 429 以及 Qualcomm ® QCS605 和 QCS403、Qualcomm® SM6125、Qualcomm® Snapdragon™ 820Am 汽车平台和 Qualcomm Flight。对于 Qualcomm® Adreno™ GPU 支持，设备上必须存在 libOpenCL.so。

For our case, we were using a board with the 626 and an Adreno™ 506 GPU. The board vendor also provided the Linux BSP. When we built the Linux image, it already included a libOpenCL.so under /usr/lib (32-bit) and /usr/lib64 (64-bit).

We were also using another development board from another vendor, and the SNPE SDK was included with the development kit along with instructions on how to set it up onboard.

Basically, it depends on the board and the accompanying BSP. Otherwise, you'll probably need to customize your Linux image to add support for it.

于 2019-07-12T03:06:33.963 回答