You are viewing a plain text version of this content. The canonical link for it is here.

Posted to discuss-archive@tvm.apache.org by Sunzj via TVM Discuss <no...@discuss.tvm.ai> on 2020/04/07 09:04:03 UTC

[TVM Discuss] [Questions] How to write schedule of vectorized compute for OpenCL


Now the souce code of vectorize for OpenCL looks like:
```
vstore2((vload2(0, ( half*)compute + (ff * 2)) + (vload2(0, pad_temp_shared_local_local + 0) * ((half2)(input1_shared_local_local[0], input1_shared_local_local[0])))), 0, ( half*)compute + (ff * 2));
```


but i want something like:
```
half2 compute;
compute = compute + pad_temp_shared_local_local *  input1_shared_local_local;
```
pad_temp_shared_local_local and input1_shared_local_local are half2 registers.

Could you share how to write schedule to generate such code?





---
[Visit Topic](https://discuss.tvm.ai/t/how-to-write-schedule-of-vectorized-compute-for-opencl/6250/1) to respond.

You are receiving this because you enabled mailing list mode.

To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/142f3eb16e9cd484b2c6be1fe157c72d03825389b63700638f359ac9a19f97da).