perf(dflash, gfx1100): dedicated MQ4-V2 verify kernels, split-K residual, and launch fusion — merge-sort xt 209 → 285 tok/s

crates/hipfire-arch-qwen35/Cargo.toml

@@ -42,3 +42,38 @@ required-features = ["deltanet"]
 name = "qwen_dense_tp2_parity"
 path = "examples/qwen_dense_tp2_parity.rs"
 required-features = ["deltanet"]
+
+[[example]]
+name = "test_dflash_gdn_pre_gfx1100"
+path = "examples/test_dflash_gdn_pre_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_dflash_draft_collapse_gfx1100"
+path = "examples/test_dflash_draft_collapse_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_dflash_hidden_scatter_gfx1100"
+path = "examples/test_dflash_hidden_scatter_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_dflash_snapshot_bulk_gfx1100"
+path = "examples/test_dflash_snapshot_bulk_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_mq_f16_projection_producers_gfx1100"
+path = "examples/test_mq_f16_projection_producers_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_mq_f16_residual_producers_gfx1100"
+path = "examples/test_mq_f16_residual_producers_gfx1100.rs"
+required-features = ["deltanet"]
+
+[[example]]
+name = "test_qwen35_fa_batch_fusion_gfx1100"
+path = "examples/test_qwen35_fa_batch_fusion_gfx1100.rs"
+required-features = ["deltanet"]