kernels/kernels/wu_arch_cases/case03_dual_fetch_issue/README.md

Case 03: Dual Fetch Issue

Test Objective

Verify scalar and tensor domains can be active together and both make forward progress through the shared fetch path and split issue paths.

RTL Surface Covered

• Shared fetch arbitration between scalar and tensor schedule streams
• Scalar decode/issue domain
• Tensor decode/issue domain
• Completion markers from both warp classes

Expected Result

All scalar warps write WU_CASE_SCALAR_BASE | wid, all tensor warps write WU_CASE_TENSOR_BASE | wid, and warp 0 writes WU_CASE_PASS.