MUFU + Zfhmin

VX_config.toml

@@ -34,13 +34,17 @@ VX_CFG_EXT_A_ENABLE = false
 VX_CFG_EXT_ZACAS_ENABLE = false
 
 VX_CFG_EXT_ZICOND_ENABLE = true
+# Zfhmin: half-precision load/store, move and fp16<->fp32/fp64 conversion (no
+# half arithmetic). Requires F.
+VX_CFG_EXT_ZFHMIN_ENABLE = false
 VX_CFG_EXT_TCU_ENABLE = false
 VX_CFG_EXT_DMA_ENABLE = false
 VX_CFG_EXT_DXA_ENABLE = false
 VX_CFG_EXT_TEX_ENABLE = false
 VX_CFG_EXT_RASTER_ENABLE = false
 VX_CFG_EXT_OM_ENABLE = false
 VX_CFG_EXT_RTU_ENABLE = false
+VX_CFG_EXT_MUFU_ENABLE = false
 VX_CFG_VLEN = "expr: $VX_CFG_XLEN * 4"
 
 [pipeline]
@@ -149,6 +153,10 @@ VX_CFG_FSQRT_PE_RATIO= 1
 VX_CFG_FCVT_PE_RATIO = 1
 VX_CFG_FNCP_PE_RATIO = 1
 
+[mufu]
+# MUFU datapath lanes inside the SFU; the SFU's lanes are serialized onto them.
+VX_CFG_NUM_MUFU_LANES = 1
+
 [amo]
 # LR/SC reservation stations per LLC bank; line-indexed in BRAM, eviction yields legal spurious SC failures
 VX_CFG_AMO_RS_SIZE = "expr: max($VX_CFG_NUM_WARPS, up($VX_CFG_NUM_CORES * $VX_CFG_NUM_WARPS / $VX_CFG_L2_NUM_BANKS))"
@@ -447,7 +455,7 @@ VX_CFG_NUM_OCACHES      = "expr: max(1, up($VX_CFG_NUM_OM_CORES / 4))"
 VX_CFG_MISA_STD = "expr: ($VX_CFG_EXT_A_ENABLED << 0) | (0 << 1) | ($VX_CFG_EXT_C_ENABLED << 2) | ($VX_CFG_EXT_D_ENABLED << 3) | (0 << 4) | ($VX_CFG_EXT_F_ENABLED << 5) | (0 << 6) | (0 << 7) | (1 << 8) | (0 << 9) | (0 << 10) | (0 << 11) | ($VX_CFG_EXT_M_ENABLED << 12) | (0 << 13) | (0 << 14) | (0 << 15) | (0 << 16) | (0 << 17) | (0 << 18) | (0 << 19) | (1 << 20) | (0 << 21) | (0 << 22) | (1 << 23) | (0 << 24) | (0 << 25)"
 
 # Custom extensions signature
-VX_CFG_MISA_EXT = "expr: ($VX_CFG_ICACHE_ENABLED << 0) | ($VX_CFG_DCACHE_ENABLED << 1) | ($VX_CFG_L2_ENABLED << 2) | ($VX_CFG_L3_ENABLED << 3) | ($VX_CFG_LMEM_ENABLED << 4) | ($VX_CFG_EXT_ZICOND_ENABLED << 5) | ($VX_CFG_EXT_TCU_ENABLED << 9) | ($VX_CFG_EXT_DXA_ENABLED << 10) | ($VX_CFG_EXT_TEX_ENABLED << 6) | ($VX_CFG_EXT_RASTER_ENABLED << 7) | ($VX_CFG_EXT_OM_ENABLED << 8) | ($VX_CFG_EXT_RTU_ENABLED << 11)"
+VX_CFG_MISA_EXT = "expr: ($VX_CFG_ICACHE_ENABLED << 0) | ($VX_CFG_DCACHE_ENABLED << 1) | ($VX_CFG_L2_ENABLED << 2) | ($VX_CFG_L3_ENABLED << 3) | ($VX_CFG_LMEM_ENABLED << 4) | ($VX_CFG_EXT_ZICOND_ENABLED << 5) | ($VX_CFG_EXT_TCU_ENABLED << 9) | ($VX_CFG_EXT_DXA_ENABLED << 10) | ($VX_CFG_EXT_TEX_ENABLED << 6) | ($VX_CFG_EXT_RASTER_ENABLED << 7) | ($VX_CFG_EXT_OM_ENABLED << 8) | ($VX_CFG_EXT_RTU_ENABLED << 11) | ($VX_CFG_EXT_MUFU_ENABLED << 12) | ($VX_CFG_EXT_ZFHMIN_ENABLED << 13)"
 
 [toolchain]
 ASIC = false

VX_types.toml

@@ -56,6 +56,9 @@ VX_MEM_IO_END_ADDR          = "expr: $VX_MEM_USER_BASE_ADDR"
 VX_MEM_OM_BASE_ADDR         = "expr: 0x00000000E0000000 if ($XLEN == 64) else 0xE0000000"
 VX_MEM_OM_END_ADDR          = "expr: $VX_MEM_PAGE_TABLE_BASE_ADDR"
 
+[mufu]
+VX_MUFU_LATENCY = 7
+
 # VM page-table format — RISC-V architectural constants fixed by the SATP mode
 # (SV32 vs SV39, selected by XLEN). Relocated from VX_config.toml.
 [vm]

ci/baselines/perf/core.json

@@ -6,7 +6,7 @@
     },
     "app": "raycast",
     "args": "",
-    "config_hash": "396cee9df6025cdb",
+    "config_hash": "30868719eb39dd56",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -17,7 +17,7 @@
     },
     "app": "raycast",
     "args": "",
-    "config_hash": "262f24032ef14b58",
+    "config_hash": "bbb5e759fb64a2d2",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -28,7 +28,7 @@
     },
     "app": "sgemm",
     "args": "-n128",
-    "config_hash": "ad2b37492d761e8f",
+    "config_hash": "f3cdbb1a54b60bd7",
     "configs": "-DVX_CFG_NUM_WARPS=8 -DVX_CFG_NUM_CORES=2 -DVX_CFG_NUM_THREADS=16 -DVX_CFG_L2_ENABLE",
     "driver": "rtlsim"
   },
@@ -39,7 +39,7 @@
     },
     "app": "sgemm",
     "args": "-n128",
-    "config_hash": "606d011dc6f612d3",
+    "config_hash": "1f13c513ce419a18",
     "configs": "-DVX_CFG_NUM_WARPS=8 -DVX_CFG_NUM_CORES=2 -DVX_CFG_NUM_THREADS=4 -DVX_CFG_L2_ENABLE",
     "driver": "rtlsim"
   },
@@ -50,7 +50,7 @@
     },
     "app": "sgemm",
     "args": "-n128",
-    "config_hash": "9390754d3719fca3",
+    "config_hash": "6c4c16bd5ba783fb",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -61,7 +61,7 @@
     },
     "app": "sgemm",
     "args": "-n128",
-    "config_hash": "740fd84fa2938103",
+    "config_hash": "e654782241f3196f",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -72,7 +72,7 @@
     },
     "app": "sgemmx",
     "args": "-n128",
-    "config_hash": "2e1c44e05f9ac34d",
+    "config_hash": "406f5c17ca35ee9d",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -83,7 +83,7 @@
     },
     "app": "sgemmx",
     "args": "-n128",
-    "config_hash": "024adb49ed96bdd3",
+    "config_hash": "5016dd4b1102f6fc",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -94,7 +94,7 @@
     },
     "app": "sgemv",
     "args": "-m512 -n512",
-    "config_hash": "a3d9df9695acbb95",
+    "config_hash": "b631a0bafc91d264",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -105,7 +105,7 @@
     },
     "app": "sgemv",
     "args": "-m512 -n512",
-    "config_hash": "0e8990b25e64ad7f",
+    "config_hash": "14333df532bad1d6",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -116,7 +116,7 @@
     },
     "app": "softmax",
     "args": "-n64",
-    "config_hash": "9e0c1c9767c760ea",
+    "config_hash": "fdbcac76ee256139",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -127,7 +127,7 @@
     },
     "app": "softmax",
     "args": "-n64",
-    "config_hash": "eb5fed4bfce19d76",
+    "config_hash": "d1489b9fccb46500",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -138,7 +138,7 @@
     },
     "app": "stencil3d",
     "args": "-n32",
-    "config_hash": "6925a5138be930ea",
+    "config_hash": "7514f2e770fda7af",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -149,7 +149,7 @@
     },
     "app": "stencil3d",
     "args": "-n32",
-    "config_hash": "b83a1e3735f79d39",
+    "config_hash": "7b1283f03793abb6",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -160,7 +160,7 @@
     },
     "app": "vecadd",
     "args": "-n16384",
-    "config_hash": "ef0c2548e7c5bf86",
+    "config_hash": "3c79fc19ad8549a0",
     "configs": "-DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -171,7 +171,7 @@
     },
     "app": "vecadd",
     "args": "-n16384",
-    "config_hash": "ce1eb79984a98552",
+    "config_hash": "75b65abb22c3753b",
     "configs": "-DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   }

ci/baselines/perf/dxa.json

@@ -6,7 +6,7 @@
     },
     "app": "dxa_copy_mcast",
     "args": "",
-    "config_hash": "6a815f855f6e0d05",
+    "config_hash": "1d98c347f39baa17",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE",
     "driver": "rtlsim"
   },
@@ -17,7 +17,7 @@
     },
     "app": "dxa_copy",
     "args": "-d1",
-    "config_hash": "626c5f6244584b1f",
+    "config_hash": "f0ace578558d01a1",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -28,7 +28,7 @@
     },
     "app": "dxa_copy",
     "args": "-d1",
-    "config_hash": "7c32a11d123bfb13",
+    "config_hash": "9a3f1b02710f7458",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -39,7 +39,7 @@
     },
     "app": "sgemm2_dxa_mcast",
     "args": "-n128",
-    "config_hash": "822d0ca194266a27",
+    "config_hash": "4d3b7f000ce31563",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -50,7 +50,7 @@
     },
     "app": "sgemm2_dxa_mcast",
     "args": "-n128",
-    "config_hash": "87ea544e77719254",
+    "config_hash": "4dffe3629adfe705",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -61,7 +61,7 @@
     },
     "app": "sgemm2_dxa",
     "args": "-n128 -t4 -m1",
-    "config_hash": "e55b61598f5f10ea",
+    "config_hash": "6aac48586a19e5d1",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -72,7 +72,7 @@
     },
     "app": "sgemm2_dxa",
     "args": "-n128 -t4 -m1",
-    "config_hash": "21da2f2737bf03dd",
+    "config_hash": "15066287b6cf674a",
     "configs": "-DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -83,7 +83,7 @@
     },
     "app": "sgemm_tcu_wg_dxa_mcast",
     "args": "-m 128 -n 128 -k 64",
-    "config_hash": "20304652a1bfb94c",
+    "config_hash": "6f3f943387268cd9",
     "configs": "-DVX_CFG_EXT_TCU_ENABLE -DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_ALU_BLOCKS=1 -DVX_CFG_NUM_LSU_BLOCKS=1 -DVX_CFG_NUM_FPU_BLOCKS=1 -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -94,7 +94,7 @@
     },
     "app": "sgemm_tcu_wg_dxa_mcast",
     "args": "-m 128 -n 128 -k 64",
-    "config_hash": "bf4520d008c07dc9",
+    "config_hash": "99f4567a3f3d04a6",
     "configs": "-DVX_CFG_EXT_TCU_ENABLE -DVX_CFG_EXT_DXA_ENABLE -DVX_CFG_NUM_ALU_BLOCKS=1 -DVX_CFG_NUM_LSU_BLOCKS=1 -DVX_CFG_NUM_FPU_BLOCKS=1 -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -105,7 +105,7 @@
     },
     "app": "sgemm_tcu_wg_dxa",
     "args": "-m 128 -n 128 -k 64",
-    "config_hash": "2a4e017d98740636",
+    "config_hash": "a45156b86aef0add",
     "configs": "-DVX_CFG_ISSUE_WIDTH=4 -DVX_CFG_EXT_TCU_ENABLE -DVX_CFG_TCU_WGMMA_ENABLE -DVX_CFG_EXT_DXA_ENABLE -DWGMMA_NRC=16 -DVX_CFG_NUM_ALU_BLOCKS=1 -DVX_CFG_NUM_LSU_BLOCKS=1 -DVX_CFG_NUM_FPU_BLOCKS=1 -DVX_CFG_NUM_WARPS=8 -DVX_CFG_NUM_THREADS=32",
     "driver": "rtlsim"
   },
@@ -116,7 +116,7 @@
     },
     "app": "sgemm_tcu_wg_sp_dxa",
     "args": "-m 128 -n 128 -k 128",
-    "config_hash": "f05069d7cf0ee3ff",
+    "config_hash": "63afe161fb9d2aac",
     "configs": "-DVX_CFG_ISSUE_WIDTH=4 -DVX_CFG_EXT_TCU_ENABLE -DVX_CFG_TCU_WGMMA_ENABLE -DVX_CFG_TCU_SPARSE_ENABLE -DVX_CFG_EXT_DXA_ENABLE -DWGMMA_NRC=32 -DVX_CFG_NUM_ALU_BLOCKS=1 -DVX_CFG_NUM_LSU_BLOCKS=1 -DVX_CFG_NUM_FPU_BLOCKS=1 -DVX_CFG_NUM_WARPS=8 -DVX_CFG_NUM_THREADS=32",
     "driver": "rtlsim"
   }

ci/baselines/perf/graphics.json

@@ -6,7 +6,7 @@
     },
     "app": "gfx_draw3d",
     "args": "-w 128 -h 128 -t box.cgltrace -r box_ref_128.png",
-    "config_hash": "1e2937b20579ad3f",
+    "config_hash": "71e067fd7765f075",
     "configs": "-DVX_CFG_EXT_TEX_ENABLE -DVX_CFG_EXT_RASTER_ENABLE -DVX_CFG_EXT_OM_ENABLE -DVX_CFG_NUM_CORES=2 -DVX_CFG_NUM_THREADS=4 -DVX_CFG_L2_ENABLE",
     "driver": "rtlsim"
   },
@@ -17,7 +17,7 @@
     },
     "app": "gfx_draw3d",
     "args": "-w 128 -h 128 -t box.cgltrace -r box_ref_128.png",
-    "config_hash": "edc924bc4f246971",
+    "config_hash": "ac94e8341d7c171e",
     "configs": "-DVX_CFG_EXT_TEX_ENABLE -DVX_CFG_EXT_RASTER_ENABLE -DVX_CFG_EXT_OM_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -28,7 +28,7 @@
     },
     "app": "gfx_draw3d",
     "args": "-w 128 -h 128 -t box.cgltrace -r box_ref_128.png",
-    "config_hash": "d682cc8c6f99c06b",
+    "config_hash": "00c3aa3428bf4e86",
     "configs": "-DVX_CFG_EXT_TEX_ENABLE -DVX_CFG_EXT_RASTER_ENABLE -DVX_CFG_EXT_OM_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -39,7 +39,7 @@
     },
     "app": "gfx_om",
     "args": "-w 128 -h 128 -r whitebox_128.png",
-    "config_hash": "46a66220c1d9956b",
+    "config_hash": "a99eb5442dfe8bd5",
     "configs": "-DVX_CFG_EXT_OM_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -50,7 +50,7 @@
     },
     "app": "gfx_om",
     "args": "-w 128 -h 128 -r whitebox_128.png",
-    "config_hash": "a7554f6fa88474dc",
+    "config_hash": "b1631e4a43822c5a",
     "configs": "-DVX_CFG_EXT_OM_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -61,7 +61,7 @@
     },
     "app": "gfx_raster",
     "args": "-w 128 -h 128 -t triangle.cgltrace -r triangle_ref_128.png",
-    "config_hash": "e2e1879d7d6f9aaf",
+    "config_hash": "2c54468d2adf668e",
     "configs": "-DVX_CFG_EXT_RASTER_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -72,7 +72,7 @@
     },
     "app": "gfx_raster",
     "args": "-w 128 -h 128 -t triangle.cgltrace -r triangle_ref_128.png",
-    "config_hash": "6e48f6b249533cc5",
+    "config_hash": "1f50896f410bc97f",
     "configs": "-DVX_CFG_EXT_RASTER_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   },
@@ -83,7 +83,7 @@
     },
     "app": "gfx_tex",
     "args": "-i toad.png -r toad_ref_f0.png -f 0 -g 0",
-    "config_hash": "c8c6b30537e93e2f",
+    "config_hash": "b72e550c7f6c7aca",
     "configs": "-DVX_CFG_EXT_TEX_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -94,7 +94,7 @@
     },
     "app": "gfx_tex",
     "args": "-i toad.png -r toad_ref_f0.png -f 0 -g 0",
-    "config_hash": "a4fc4d97e6e80586",
+    "config_hash": "15f81408ca91fc9f",
     "configs": "-DVX_CFG_EXT_TEX_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   }

ci/baselines/perf/raytracing.json

@@ -6,7 +6,7 @@
     },
     "app": "rt_raycast",
     "args": "",
-    "config_hash": "9a4f28d2555289b5",
+    "config_hash": "048a348462ce8158",
     "configs": "-DVX_CFG_EXT_RTU_ENABLE -DVX_CFG_NUM_CORES=2 -DVX_CFG_NUM_THREADS=4 -DVX_CFG_L2_ENABLE",
     "driver": "rtlsim"
   },
@@ -17,7 +17,7 @@
     },
     "app": "rt_raycast",
     "args": "",
-    "config_hash": "aef578263250d585",
+    "config_hash": "bbab935bba11ff85",
     "configs": "-DVX_CFG_EXT_RTU_ENABLE -DVX_CFG_NUM_THREADS=16",
     "driver": "rtlsim"
   },
@@ -28,7 +28,7 @@
     },
     "app": "rt_raycast",
     "args": "",
-    "config_hash": "c7b45fe5c1828bc5",
+    "config_hash": "f2e45a7a7018b151",
     "configs": "-DVX_CFG_EXT_RTU_ENABLE -DVX_CFG_NUM_THREADS=4",
     "driver": "rtlsim"
   }