Support H100 for other CUDA extensions

csrc/ft_attention/setup.py

 # Adapted from https://github.com/NVIDIA/apex/blob/master/setup.py
-import torch
-from torch.utils.cpp_extension import BuildExtension, CppExtension, CUDAExtension, CUDA_HOME
-from setuptools import setup, find_packages
-import subprocess
-
 import sys
 import warnings
 import os
+from packaging.version import parse, Version
+
+from setuptools import setup, find_packages
+import subprocess
+
+import torch
+from torch.utils.cpp_extension import BuildExtension, CppExtension, CUDAExtension, CUDA_HOME
+
 
 # ninja build does not work unless include_dirs are abs path
 this_dir = os.path.dirname(os.path.abspath(__file__))
@@ -16,22 +19,19 @@ def get_cuda_bare_metal_version(cuda_dir):
     raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
     output = raw_output.split()
     release_idx = output.index("release") + 1
-    release = output[release_idx].split(".")
-    bare_metal_major = release[0]
-    bare_metal_minor = release[1][0]
+    bare_metal_version = parse(output[release_idx].split(",")[0])
 
-    return raw_output, bare_metal_major, bare_metal_minor
+    return raw_output, bare_metal_version
 
 
 def check_cuda_torch_binary_vs_bare_metal(cuda_dir):
-    raw_output, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(cuda_dir)
-    torch_binary_major = torch.version.cuda.split(".")[0]
-    torch_binary_minor = torch.version.cuda.split(".")[1]
+    raw_output, bare_metal_version = get_cuda_bare_metal_version(cuda_dir)
+    torch_binary_version = parse(torch.version.cuda)
 
     print("\nCompiling cuda extensions with")
     print(raw_output + "from " + cuda_dir + "/bin\n")
 
-    if (bare_metal_major != torch_binary_major) or (bare_metal_minor != torch_binary_minor):
+    if (bare_metal_version != torch_binary_version):
         raise RuntimeError(
             "Cuda extensions are being compiled with a version of Cuda that does "
             "not match the version used to compile Pytorch binaries.  "
@@ -53,8 +53,8 @@ def raise_if_cuda_home_none(global_option: str) -> None:
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-    if int(bare_metal_major) >= 11 and int(bare_metal_minor) >= 2:
+    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    if bare_metal_version >= Version("11.2"):
         return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args
 
@@ -72,15 +72,18 @@ if not torch.cuda.is_available():
         "If you wish to cross-compile for a single specific architecture,\n"
         'export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.\n',
     )
-    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None:
-        _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-        if int(bare_metal_major) == 11:
-            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
-            if int(bare_metal_minor) > 0:
+    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None and CUDA_HOME is not None:
+        _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+        if bare_metal_version >= Version("11.8"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6;9.0"
+        elif bare_metal_version >= Version("11.1"):
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6"
+        elif bare_metal_version == Version("11.0"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
         else:
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5"
 
+
 print("\n\ntorch.__version__  = {}\n\n".format(torch.__version__))
 TORCH_MAJOR = int(torch.__version__.split(".")[0])
 TORCH_MINOR = int(torch.__version__.split(".")[1])
@@ -98,10 +101,16 @@ if os.path.exists(os.path.join(torch_dir, "include", "ATen", "CUDAGeneratorImpl.
 raise_if_cuda_home_none("--ft_attention")
 # Check, if CUDA11 is installed for compute capability 8.0
 cc_flag = []
-# cc_flag.append("-gencode")
-# cc_flag.append("arch=compute_70,code=sm_70")
+_, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+if bare_metal_version < Version("11.0"):
+    raise RuntimeError("ft_attention is only supported on CUDA 11 and above")
+cc_flag.append("-gencode")
+cc_flag.append("arch=compute_70,code=sm_70")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_80,code=sm_80")
+if bare_metal_version >= Version("11.8"):
+    cc_flag.append("-gencode")
+    cc_flag.append("arch=compute_90,code=sm_90")
 
 ext_modules.append(
     CUDAExtension(

csrc/fused_dense_lib/setup.py

 import os
 import subprocess
+from packaging.version import parse, Version
 
 import torch
 from setuptools import setup
@@ -10,16 +11,14 @@ def get_cuda_bare_metal_version(cuda_dir):
     raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
     output = raw_output.split()
     release_idx = output.index("release") + 1
-    release = output[release_idx].split(".")
-    bare_metal_major = release[0]
-    bare_metal_minor = release[1][0]
+    bare_metal_version = parse(output[release_idx].split(",")[0])
 
-    return raw_output, bare_metal_major, bare_metal_minor
+    return raw_output, bare_metal_version
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-    if int(bare_metal_major) >= 11 and int(bare_metal_minor) >= 2:
+    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    if bare_metal_version >= Version("11.2"):
         return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args
 

csrc/layer_norm/setup.py

 # Adapted from https://github.com/NVIDIA/apex/blob/master/setup.py
+import sys
+import warnings
+import os
+from packaging.version import parse, Version
+
 import torch
 from torch.utils.cpp_extension import BuildExtension, CppExtension, CUDAExtension, CUDA_HOME
 from setuptools import setup, find_packages
 import subprocess
 
-import sys
-import warnings
-import os
-
 # ninja build does not work unless include_dirs are abs path
 this_dir = os.path.dirname(os.path.abspath(__file__))
 
@@ -16,22 +17,19 @@ def get_cuda_bare_metal_version(cuda_dir):
     raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
     output = raw_output.split()
     release_idx = output.index("release") + 1
-    release = output[release_idx].split(".")
-    bare_metal_major = release[0]
-    bare_metal_minor = release[1][0]
+    bare_metal_version = parse(output[release_idx].split(",")[0])
 
-    return raw_output, bare_metal_major, bare_metal_minor
+    return raw_output, bare_metal_version
 
 
 def check_cuda_torch_binary_vs_bare_metal(cuda_dir):
-    raw_output, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(cuda_dir)
-    torch_binary_major = torch.version.cuda.split(".")[0]
-    torch_binary_minor = torch.version.cuda.split(".")[1]
+    raw_output, bare_metal_version = get_cuda_bare_metal_version(cuda_dir)
+    torch_binary_version = parse(torch.version.cuda)
 
     print("\nCompiling cuda extensions with")
     print(raw_output + "from " + cuda_dir + "/bin\n")
 
-    if (bare_metal_major != torch_binary_major) or (bare_metal_minor != torch_binary_minor):
+    if (bare_metal_version != torch_binary_version):
         raise RuntimeError(
             "Cuda extensions are being compiled with a version of Cuda that does "
             "not match the version used to compile Pytorch binaries.  "
@@ -53,8 +51,8 @@ def raise_if_cuda_home_none(global_option: str) -> None:
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-    if int(bare_metal_major) >= 11 and int(bare_metal_minor) >= 2:
+    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    if bare_metal_version >= Version("11.2"):
         return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args
 
@@ -72,15 +70,18 @@ if not torch.cuda.is_available():
         "If you wish to cross-compile for a single specific architecture,\n"
         'export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.\n',
     )
-    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None:
-        _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-        if int(bare_metal_major) == 11:
-            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
-            if int(bare_metal_minor) > 0:
+    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None and CUDA_HOME is not None:
+        _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+        if bare_metal_version >= Version("11.8"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6;9.0"
+        elif bare_metal_version >= Version("11.1"):
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6"
+        elif bare_metal_version == Version("11.0"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
         else:
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5"
 
+
 print("\n\ntorch.__version__  = {}\n\n".format(torch.__version__))
 TORCH_MAJOR = int(torch.__version__.split(".")[0])
 TORCH_MINOR = int(torch.__version__.split(".")[1])
@@ -98,10 +99,16 @@ if os.path.exists(os.path.join(torch_dir, "include", "ATen", "CUDAGeneratorImpl.
 raise_if_cuda_home_none("--fast_layer_norm")
 # Check, if CUDA11 is installed for compute capability 8.0
 cc_flag = []
+_, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+if bare_metal_version < Version("11.0"):
+    raise RuntimeError("dropout_layer_norm is only supported on CUDA 11 and above")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_70,code=sm_70")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_80,code=sm_80")
+if bare_metal_version >= Version("11.8"):
+    cc_flag.append("-gencode")
+    cc_flag.append("arch=compute_90,code=sm_90")
 
 ext_modules.append(
     CUDAExtension(

csrc/rotary/setup.py

 # Adapted from https://github.com/NVIDIA/apex/blob/master/setup.py
+import sys
+import warnings
+import os
+from packaging.version import parse, Version
+
 import torch
 from torch.utils.cpp_extension import BuildExtension, CppExtension, CUDAExtension, CUDA_HOME
 from setuptools import setup, find_packages
 import subprocess
 
-import sys
-import warnings
-import os
-
 # ninja build does not work unless include_dirs are abs path
 this_dir = os.path.dirname(os.path.abspath(__file__))
 
@@ -16,22 +17,19 @@ def get_cuda_bare_metal_version(cuda_dir):
     raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
     output = raw_output.split()
     release_idx = output.index("release") + 1
-    release = output[release_idx].split(".")
-    bare_metal_major = release[0]
-    bare_metal_minor = release[1][0]
+    bare_metal_version = parse(output[release_idx].split(",")[0])
 
-    return raw_output, bare_metal_major, bare_metal_minor
+    return raw_output, bare_metal_version
 
 
 def check_cuda_torch_binary_vs_bare_metal(cuda_dir):
-    raw_output, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(cuda_dir)
-    torch_binary_major = torch.version.cuda.split(".")[0]
-    torch_binary_minor = torch.version.cuda.split(".")[1]
+    raw_output, bare_metal_version = get_cuda_bare_metal_version(cuda_dir)
+    torch_binary_version = parse(torch.version.cuda)
 
     print("\nCompiling cuda extensions with")
     print(raw_output + "from " + cuda_dir + "/bin\n")
 
-    if (bare_metal_major != torch_binary_major) or (bare_metal_minor != torch_binary_minor):
+    if (bare_metal_version != torch_binary_version):
         raise RuntimeError(
             "Cuda extensions are being compiled with a version of Cuda that does "
             "not match the version used to compile Pytorch binaries.  "
@@ -53,8 +51,8 @@ def raise_if_cuda_home_none(global_option: str) -> None:
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-    if int(bare_metal_major) >= 11 and int(bare_metal_minor) >= 2:
+    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    if bare_metal_version >= Version("11.2"):
         return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args
 
@@ -72,15 +70,18 @@ if not torch.cuda.is_available():
         "If you wish to cross-compile for a single specific architecture,\n"
         'export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.\n',
     )
-    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None:
-        _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-        if int(bare_metal_major) == 11:
-            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
-            if int(bare_metal_minor) > 0:
+    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None and CUDA_HOME is not None:
+        _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+        if bare_metal_version >= Version("11.8"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6;9.0"
+        elif bare_metal_version >= Version("11.1"):
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6"
+        elif bare_metal_version == Version("11.0"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
         else:
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5"
 
+
 print("\n\ntorch.__version__  = {}\n\n".format(torch.__version__))
 TORCH_MAJOR = int(torch.__version__.split(".")[0])
 TORCH_MINOR = int(torch.__version__.split(".")[1])
@@ -91,10 +92,16 @@ ext_modules = []
 raise_if_cuda_home_none("rotary_emb")
 # Check, if CUDA11 is installed for compute capability 8.0
 cc_flag = []
+_, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+if bare_metal_version < Version("11.0"):
+    raise RuntimeError("rotary_emb is only supported on CUDA 11 and above")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_70,code=sm_70")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_80,code=sm_80")
+if bare_metal_version >= Version("11.8"):
+    cc_flag.append("-gencode")
+    cc_flag.append("arch=compute_90,code=sm_90")
 
 ext_modules.append(
     CUDAExtension(

csrc/xentropy/setup.py

 # Adapted from https://github.com/NVIDIA/apex/blob/master/setup.py
+import sys
+import warnings
+import os
+from packaging.version import parse, Version
+
 import torch
 from torch.utils.cpp_extension import BuildExtension, CppExtension, CUDAExtension, CUDA_HOME
 from setuptools import setup, find_packages
 import subprocess
 
-import sys
-import warnings
-import os
-
 # ninja build does not work unless include_dirs are abs path
 this_dir = os.path.dirname(os.path.abspath(__file__))
 
@@ -16,22 +17,19 @@ def get_cuda_bare_metal_version(cuda_dir):
     raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
     output = raw_output.split()
     release_idx = output.index("release") + 1
-    release = output[release_idx].split(".")
-    bare_metal_major = release[0]
-    bare_metal_minor = release[1][0]
+    bare_metal_version = parse(output[release_idx].split(",")[0])
 
-    return raw_output, bare_metal_major, bare_metal_minor
+    return raw_output, bare_metal_version
 
 
 def check_cuda_torch_binary_vs_bare_metal(cuda_dir):
-    raw_output, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(cuda_dir)
-    torch_binary_major = torch.version.cuda.split(".")[0]
-    torch_binary_minor = torch.version.cuda.split(".")[1]
+    raw_output, bare_metal_version = get_cuda_bare_metal_version(cuda_dir)
+    torch_binary_version = parse(torch.version.cuda)
 
     print("\nCompiling cuda extensions with")
     print(raw_output + "from " + cuda_dir + "/bin\n")
 
-    if (bare_metal_major != torch_binary_major) or (bare_metal_minor != torch_binary_minor):
+    if (bare_metal_version != torch_binary_version):
         raise RuntimeError(
             "Cuda extensions are being compiled with a version of Cuda that does "
             "not match the version used to compile Pytorch binaries.  "
@@ -53,8 +51,8 @@ def raise_if_cuda_home_none(global_option: str) -> None:
 
 
 def append_nvcc_threads(nvcc_extra_args):
-    _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-    if int(bare_metal_major) >= 11 and int(bare_metal_minor) >= 2:
+    _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+    if bare_metal_version >= Version("11.2"):
         return nvcc_extra_args + ["--threads", "4"]
     return nvcc_extra_args
 
@@ -72,15 +70,18 @@ if not torch.cuda.is_available():
         "If you wish to cross-compile for a single specific architecture,\n"
         'export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.\n',
     )
-    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None:
-        _, bare_metal_major, bare_metal_minor = get_cuda_bare_metal_version(CUDA_HOME)
-        if int(bare_metal_major) == 11:
-            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
-            if int(bare_metal_minor) > 0:
+    if os.environ.get("TORCH_CUDA_ARCH_LIST", None) is None and CUDA_HOME is not None:
+        _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+        if bare_metal_version >= Version("11.8"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6;9.0"
+        elif bare_metal_version >= Version("11.1"):
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0;8.6"
+        elif bare_metal_version == Version("11.0"):
+            os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5;8.0"
         else:
             os.environ["TORCH_CUDA_ARCH_LIST"] = "6.0;6.1;6.2;7.0;7.5"
 
+
 print("\n\ntorch.__version__  = {}\n\n".format(torch.__version__))
 TORCH_MAJOR = int(torch.__version__.split(".")[0])
 TORCH_MINOR = int(torch.__version__.split(".")[1])
@@ -98,10 +99,16 @@ if os.path.exists(os.path.join(torch_dir, "include", "ATen", "CUDAGeneratorImpl.
 raise_if_cuda_home_none("--xentropy")
 # Check, if CUDA11 is installed for compute capability 8.0
 cc_flag = []
+_, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
+if bare_metal_version < Version("11.0"):
+    raise RuntimeError("xentropy is only supported on CUDA 11 and above")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_70,code=sm_70")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_80,code=sm_80")
+if bare_metal_version >= Version("11.8"):
+    cc_flag.append("-gencode")
+    cc_flag.append("arch=compute_90,code=sm_90")
 
 ext_modules.append(
     CUDAExtension(

flash_attn/ops/fused_dense.py

@@ -421,6 +421,8 @@ class FusedMLP(nn.Module):
             'auto': heuristic will be picked automatically:
                 For CUDA >= 11.8, we set heuristic=0 for both fp16 and bf16 for best perf.
                 For CUDA <= 11.7, we set heuristic=1 for fp16 and heuristic=-1 for bf16.
+                For H100, we set heuristic=-1 for both fp16 and bf16 as the fused cuBlasLt implementation
+                is slower than the unfused version.
         return_residual: whether to return the input x along with the output. This is for
             performance reason: for post-norm architecture, returning the input allows us
             to fuse the backward of nn.Linear with the residual connection.
@@ -442,6 +444,9 @@ class FusedMLP(nn.Module):
         dtype = x.dtype if not torch.is_autocast_enabled() else torch.get_autocast_gpu_dtype()
         if self.heuristic == 'auto':
             if self.activation == 'gelu_approx':
+                if torch.cuda.get_device_capability('cuda') == (9, 0):
+                    heuristic = -1
+                else:
                     cuda_ver = tuple(map(int, torch.version.cuda.split('.')))
                     heuristic = 0 if cuda_ver >= (11, 8) else (1 if dtype == torch.float16 else -1)
             else:

setup.py

@@ -108,7 +108,7 @@ raise_if_cuda_home_none("flash_attn")
 cc_flag = []
 _, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
 if bare_metal_version < Version("11.0"):
-    raise RuntimeError("FlashAttention is only supported on CUDA 11")
+    raise RuntimeError("FlashAttention is only supported on CUDA 11 and above")
 cc_flag.append("-gencode")
 cc_flag.append("arch=compute_75,code=sm_75")
 cc_flag.append("-gencode")