Port oss f16_fast_gemv into fbcode (#3610)

Summary:
X-link: facebookresearch/FBGEMM#688


This diff content includes:
1. Port OSS FastGEMV `fp16` kernel into fbcode and expose to python as a step 1 - `torch.ops.fbgemm.f16_fast_gemv`
https://github.com/wangsiping97/FastGEMV/blob/1fdff6f74aade033c02727a419afd6a4b4bfbc3f/fast_gemv.cu#L14
2. Add `fp16_oss_fast_gemv` to quantize ops benchmark script
3. Add two simple tests for custom op`torch.ops.fbgemm.f16_fast_gemv` to test
     - `torch.compile()` able
     -  correctness

Perf numbers compared with `f16_baseline,fp16_oss_fast_gemv,cuda_lite,marlin_bf16i4,machete_bf16i4`

======================

### Benchmark Results

| **M** | **N** | **K** | **Method** | **Elapsed Time (ms)** | **TFLOPS** | **GB/s** |
| --- | --- | --- | --- | --- | --- | --- |
| 1 | 1280 | 8192 | bf16_baseline | 0.024 | 0.860 | 861.042 |
| 1 | 1280 | 8192 | fp16_oss_fast_gemv | 0.019 | 1.126 | 1127.391 |
| 1 | 1280 | 8192 | cuda_lite_fp8 | 0.015 | 1.357 | 679.032 |
| 1 | 1280 | 8192 | marlin_bf16i4 | 0.027 | 0.768 | 192.612 |
| 1 | 1280 | 8192 | machete_bf16i4 | 0.026 | 0.810 | 203.219 |
| 1 | 8192 | 1024 | bf16_baseline | 0.018 | 0.952 | 953.176 |
| 1 | 8192 | 1024 | fp16_oss_fast_gemv | 0.010 | 1.763 | 1765.033 |
| 1 | 8192 | 1024 | cuda_lite_fp8 | 0.014 | 1.198 | 600.054 |
| 1 | 8192 | 1024 | marlin_bf16i4 | 0.015 | 1.144 | 287.150 |
| 1 | 8192 | 1024 | machete_bf16i4 | 0.014 | 1.187 | 298.096 |
| 1 | 7168 | 8192 | bf16_baseline | 0.073 | 1.609 | 1608.983 |
| 1 | 7168 | 8192 | fp16_oss_fast_gemv | 0.069 | 1.697 | 1697.308 |
| 1 | 7168 | 8192 | cuda_lite_fp8 | 0.044 | 2.679 | 1340.093 |
| 1 | 7168 | 8192 | marlin_bf16i4 | 0.033 | 3.590 | 898.436 |
| 1 | 7168 | 8192 | machete_bf16i4 | 0.039 | 3.017 | 755.147 |
| 1 | 8192 | 3584 | bf16_baseline | 0.045 | 1.312 | 1312.239 |
| 1 | 8192 | 3584 | fp16_oss_fast_gemv | 0.026 | 2.268 | 1134.843 |
| 1 | 8192 | 3584 | cuda_lite_fp8 | 0.026 | 2.271 | 1136.151 |
| 1 | 8192 | 3584 | marlin_bf16i4 | 0.021 | 2.808 | 703.164 |
| 1 | 8192 | 3584 | machete_bf16i4 | 0.024 | 2.460 | 615.990 |

Note that currently the precision with fast_gemv kernel and cuda_lite not matched yet. so need fp8 to coming in for a fairer result.


heuristic sweep results from the 4 problem sizes we care about:

P1722806148


**Next step:**
Need fp8 mixed precision support for fast gemv kernel which is what we want

Differential Revision: D68470488

fbgemm_gpu/experimental/gen_ai/CMakeLists.txt

@@ -33,7 +33,10 @@ endif()
 # CUDA-specific sources
 file(GLOB_RECURSE experimental_gen_ai_cpp_source_files_cuda
   src/quantize/cutlass_extensions/*.cu
-  src/quantize/cutlass_extensions/**/*.cu)
+  src/quantize/cutlass_extensions/**/*.cu
+  src/quantize/fast_gemv/*.cu
+  src/quantize/fast_gemv/**/*.cu
+  src/quantize/fast_gemv/**/*.cuh)
 
 # HIP-specific sources
 file(GLOB_RECURSE experimental_gen_ai_cpp_source_files_hip

fbgemm_gpu/experimental/gen_ai/bench/quantize_ops.py

@@ -361,6 +361,38 @@ def cuda(self) -> bool:
         return True
 
 
+@register_quantize_op
+class Fp16OSSFastGemv(QuantizeOpBase):
+    """
+    FP16 OSS fast gemv kernel.
+    """
+
+    def quantize(self, x, w):
+        # dummy quantize
+        return x, w
+
+    def compute(self, x, w):
+        out = torch.ops.fbgemm.fp16_fast_gemv(x, w)
+        return out
+
+    def quantize_and_compute(self, x, w):
+        x, w = self.quantize(x, w)
+        return self.compute(x, w)
+
+    @property
+    def name(self) -> str:
+        return "fp16_oss_fast_gemv"
+
+    @property
+    def hip(self) -> bool:
+        # This implementation is specific to cublas.
+        return False
+
+    @property
+    def cuda(self) -> bool:
+        return True
+
+
 @register_quantize_op
 class FP8CublasRowwiseGemm(QuantizeOpBase):
     """

fbgemm_gpu/experimental/gen_ai/src/quantize/fast_gemv/fp16_fast_gemv.cu

@@ -0,0 +1,96 @@
+/* This source code is copied from https://github.com/wangsiping97/FastGEMV
+
+MIT License
+
+Copyright (c) 2023 Siping Wang
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+*/
+
+#include <ATen/ATen.h>
+#include <ATen/cuda/CUDAContext.h>
+#include <c10/core/ScalarType.h>
+#include <c10/cuda/CUDAGuard.h>
+
+#include "include/fast_gemv.cuh"
+
+namespace fbgemm_gpu {
+
+dim3 get_best_block_dim(int m, int n, int k) {
+  if (m == 1 && n == 1280 && k == 8192) {
+    return dim3(128, 4);
+  } else if (m == 1 && n == 8192 && k == 1024) {
+    return dim3(32, 8);
+  } else if (m == 1 && n == 7168 && k == 8192) {
+    return dim3(256, 1);
+  } else if (m == 1 && n == 8192 && k == 3584) {
+    return dim3(64, 2);
+  } else {
+    // Default block dimensions
+    return dim3(32, 4);
+  }
+}
+
+at::Tensor fp16_fast_gemv(at::Tensor X, at::Tensor W) {
+  // X: M x K
+  // W: N x K
+  auto m = X.size(0);
+  auto n = W.size(0);
+  auto k = W.size(1);
+
+  TORCH_CHECK(X.is_cuda() && X.is_contiguous());
+  TORCH_CHECK(W.is_cuda() && W.is_contiguous());
+
+  // the block_dim values are sweeped results from the heuristics tuning script
+  dim3 block_dim = get_best_block_dim(m, n, k);
+
+  if (n % block_dim.y != 0) {
+    throw std::runtime_error(
+        "fp16_fast_gemv cannot be run when n % block_dim.x != 0");
+  } else if (k % block_dim.x != 0) {
+    throw std::runtime_error(
+        "fp16_fast_gemv cannot run when k % block_dim.x != 0");
+  } else if ((k / block_dim.x) % 8 != 0) {
+    throw std::runtime_error(
+        "fp16_fast_gemv cannot run when num_per_thread % 8 != 0");
+  }
+
+  dim3 grid_dim(1, n / block_dim.y);
+  unsigned int num_per_thread = k / block_dim.x;
+
+  auto stream = at::cuda::getCurrentCUDAStream();
+
+  auto out_sizes = X.sizes().vec();
+  out_sizes.front() = m;
+  out_sizes.back() = n;
+  auto Y = at::empty(out_sizes, X.options().dtype(at::kHalf));
+
+  gemv_fp16<<<grid_dim, block_dim, 0, stream>>>(
+      (half*)W.data_ptr(), // mat
+      (half*)X.data_ptr(), // vec
+      (half*)Y.data_ptr(), // res
+      k,
+      num_per_thread);
+
+  C10_CUDA_KERNEL_LAUNCH_CHECK();
+
+  return Y;
+}
+
+} // namespace fbgemm_gpu