I have a Orin Devkit with 32GB memory. Is there a simple apps/program to check available compute resource - 200 sparse TOPS?
What will be the electrical power required to hit the 200 TOPs, is it 40W and will devkit thermal transfer plate (TTP) handle it?
What is the difference between the dense and sparse TOPs?
Sparse TOPs is double the Dense TOPs. Sparsity is only supported in Ampere, and this is a new feature with Orin. It is described in the tech brief in section 3rd Generation Tensor Cores and Sparsity on page 5. A further breakdown can be found in our Jetson AGX Orin Series Data Sheet of both Sparse and Dense for the total AI Performance, the GPU alone, and the DLA alone.
For more details on the sparsity feature please refer to this Accelerating Inference with Sparsity Using the NVIDIA Ampere Architecture and NVIDIA TensorRT | NVIDIA Technical Blog
With the increasing adoption of Sparse neural networks in AI applications, we believe comparing Orin’s INT8+Sparsity compute of 275 TOPs to Xavier’s 32 TOPs is a fair comparison and it highlights the performance ceiling that each platform offers when all the features of each platform is enabled.
We will have more performance data in coming future.
Thanks Kay
The Orin Dev kit distributed to the users - Can it go up to 275 TOPs? There is a 32 GB and 64GB version.
Is there a command to confirm the actual mounted memory/DRAM?
Does Devkit can handle the heat load if it pushed towards 275 TOPs?
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.