Cycle estimate-based kernel selection for dot product quantized s8/u8 kernels

Resolves: COMPMID-3990

Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Change-Id: If840c79209940535450f4ea1cbf6b0ec646a168e
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4866
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>

1 files changed, 26 insertions, 1 deletions

diff --git a/src/core/NEON/kernels/arm_gemm/gemm_hybrid_indirect.hpp b/src/core/NEON/kernels/arm_gemm/gemm_hybrid_indirect.hpp
index eede1a4f76..d9035c8917 100644
--- a/src/core/NEON/kernels/arm_gemm/gemm_hybrid_indirect.hpp
+++ b/src/core/NEON/kernels/arm_gemm/gemm_hybrid_indirect.hpp
@@ -573,7 +573,7 @@ public:
     }
 
     // Estimate cycles for given problem given provided parameters
-    static uint64_t estimate_cycles(const GemmArgs &args, const PerformanceParameters &params) {
+    static uint64_t estimate_cycles(const GemmArgs &args, const PerformanceParameters &params, const OutputStage &os = {} ) {
         // Note: Current hybrid kernels don't actually round up height (they
         // have paths for each possible height).  Might need to make this
         // configurable in future.
@@ -591,6 +591,31 @@ public:
 
         uint64_t total_cycles = mac_cycles;
 
+        // Quantizing kernels with separate quantize need to add in the extra stages.
+        if (std::is_same<OutputStage, Requantize32>::value && SeparateQuantize) {
+            const Requantize32 *qp = reinterpret_cast<const Requantize32 *>(&os);
+
+            // Row sums: need to consider each value in A (batch * multi * M * K)...
+            uint64_t rowsum_bytes = static_cast<uint64_t>(args._nbatches) * args._nmulti * args._Msize * roundup(args._Ksize, strategy::k_unroll());
+
+            // ... but row sums are skipped if B offset==0.
+            if (qp->b_offset == 0) {
+                rowsum_bytes = 0;
+            }
+
+            // Use "prepare bytes per cycle" to store "row sum values per cycle".
+            float rowsum_cycles = static_cast<float>(rowsum_bytes) / params.prepare_bytes_cycle;
+
+            // Requantize: need to consider each value in C (batch * multi * M * N)
+            uint64_t requantize_bytes = static_cast<uint64_t>(args._nbatches) * args._nmulti * args._Msize * args._Nsize;
+
+            // Use "merge bytes per cycle" to store "requantize values per cycle".
+            float requantize_cycles = static_cast<float>(requantize_bytes) / params.merge_bytes_cycle;
+
+            // Recalculate total_cycles with the extra components.
+            total_cycles = mac_cycles + rowsum_cycles + requantize_cycles;
+        }
+
         return total_cycles;
     }