Forums - Snapdragon 8 Gen 1 Support

8 posts / 0 new
Last post
Snapdragon 8 Gen 1 Support
kamen.kirov
Join Date: 16 Dec 21
Posts: 3
Posted: Tue, 2022-04-12 05:39

I am running snpe-platform-validator tool from SNPE ver 1.61 on the Oneplus 10 pro device which has SD 8 Gen 1 SoC. I get this output:

SNPE is not supported on this SoC

Any idea when this support will come?

  • Up0
  • Down0
weihuan
Join Date: 12 Apr 20
Posts: 270
Posted: Sun, 2022-04-17 18:52

Dear customer,

Which backend you want to test, CPU, GPU, DSP?

Could you please leave more helpful messages here?

BR.

Wei

  • Up0
  • Down0
kher_vijay
Join Date: 11 Apr 22
Posts: 1
Posted: Fri, 2022-04-29 03:34

I am running snpe benchmarks tool from SNPE ver 1.61.0 for CPU & GPU runtime using dlc file (derived using snpe-caffe-to-dlc)on the Samsung Galaxy tab S8(WiFi)device which has SD 8 Gen 1 SoC. CPU runtime is working fine but we are get this output while GPU:

ERROR - snpebm.snpebm_device: ['The selected runtime is not available on this platform. Continue anyway to observe the failure at network creation time.'] ERROR - snpebm.snpebm_device: ['error_code=500; error_message=Target runtime is not available. error_code=500; error_message=Target runtime is not available. No viable runtimes available.; error_component=Host Runtime; line_no=409; thread_id=-382967328; error_component=Host Runtime; line_no=267; thread_id=-302631820']

If I am setting the "CpuFallback" to "true" then this error will disappear but there is no marginal difference in CPU & GPU timing. Can anyone help with this issue, why GPU is not working when we execute the DLC file on SD 8 Gen 1 SoC? Same thing is working  on Samsung Galaxy tab S7 device which has Qualcomm SM8250 Snapdragon 865 5G+ SoC.

  • Up0
  • Down0
weihuan
Join Date: 12 Apr 20
Posts: 270
Posted: Sat, 2022-05-21 20:16

Dear customer,

It seems like some model layer don't support on GPU runtime and then fallback to CPU in executing flow. Could you please help to check the SNPE DLC info with tool of snpe-dlc-info if all model layers can be implemented on GPU.

BR.
Wei

  • Up0
  • Down0
kamen.kirov
Join Date: 16 Dec 21
Posts: 3
Posted: Fri, 2022-06-10 03:36

It doesn't matter what backend I choose, the program responds with the same message. So obviously the whole SOC platform is not supported yet by SNPE.

  • Up0
  • Down0
pablog
Join Date: 28 Feb 21
Posts: 11
Posted: Fri, 2022-06-10 04:01

I have the same issue with with hdK8450 board. it does not recognized SNPE. Maybe missing libs from Qualcomm.

  • Up0
  • Down0
kamen.kirov
Join Date: 16 Dec 21
Posts: 3
Posted: Tue, 2022-06-14 06:07

SNPE 1.62 supports 8 Gen 1. Even though it's not mentioned in the Change log.  Just verified it on Oneplus 10 Pro device.

  • Up0
  • Down0
pablog
Join Date: 28 Feb 21
Posts: 11
Posted: Tue, 2022-06-14 09:54

Nice to know it, Thanks!!

But now I am facing another problem. DSP is not working:

 

// Adding ADSP path
ADSP_LIBRARY_PATH="$(pwd)/dsp;/system/lib/rfsa/adsp;/system/vendor/lib/rfsa/adsp;/dsp;/vendor/lib/;/vendor/rfs/msm/adsp;/vendor/rfs/mdm/adsp" 
 
./snpe-platform-validator --runtime dsp                                                                                                                    
 
PF_VALIDATOR: DEBUG: Calling PlatformValidator->setRuntime
PF_VALIDATOR: DEBUG: Calling PlatformValidator->RuntimeCheck
PF_VALIDATOR: DEBUG: Setting up QnnBackend
DspTransport.openSession qnn_open failed, 0x00000072
IDspTransport: Unknown rpc status 0x00000072
DspTransport failed,cannot open session, error 0xffffffff
terminating with uncaught exception of type DlSystem::RuntimeException: QnnBackend_initialize failed: 1
/usr/local/google/buildbot/src/android/ndk-release-r19/external/libcxx/../../external/libcxxabi/src/abort_message.cpp:73: abort_message: assertion "terminating with uncaught exception of type DlSystem::RuntimeException: QnnBackend_initialize failed: 1" failed
Aborted 
GPU seems to be recognized
./snpe-platform-validator --runtime gpu                                                                                                                
PF_VALIDATOR: DEBUG: Calling PlatformValidator->setRuntime
PF_VALIDATOR: DEBUG: Calling PlatformValidator->RuntimeCheck
PF_VALIDATOR: DEBUG: Building and running a simple Vector addition gpu program.
Unit Test on the runtime GPU: Passed.
SNPE is supported for runtime GPU on the device.
PF_VALIDATOR: DEBUG: Calling PlatformValidator->IsRuntimeAvailable
Runtime GPU Prerequisites: Present.
 

Note: tested on v16.2 and v16.3. Hardware: hdK8450

 

  • Up0
  • Down0
or Register

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.