Skip to main content
Log in

Mojo module

info

Contains information about GPU architectures and their capabilities.

This module provides detailed specifications for various GPU models including NVIDIA and AMD GPUs. It includes information about compute capabilities, memory specifications, thread organization, and performance characteristics.

Aliases

  • A10 = Info(__init__[__mlir_type.!kgen.string]("A10"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("ampere"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](8.5999999999999996), __init__[__mlir_type.!kgen.string]("sm_86"), 72, 32, 1536, 32, 64, 2048, 32, 102400, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 16, 128, 4, 1024):
  • A100 = Info(__init__[__mlir_type.!kgen.string]("A100"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("ampere"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](8), __init__[__mlir_type.!kgen.string]("sm_80"), 108, 32, 2048, 32, 64, 2048, 32, 167936, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 32, 128, 4, 1024):
  • B100 = Info(__init__[__mlir_type.!kgen.string]("B100"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("blackwell"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](10), __init__[__mlir_type.!kgen.string]("sm_100a"), 132, 32, -1, 32, 64, 1536, 32, 59392, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 32, 128, 4, 1024):
  • B200 = Info(__init__[__mlir_type.!kgen.string]("B100"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("blackwell"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](10), __init__[__mlir_type.!kgen.string]("sm_100a"), 132, 32, -1, 32, 64, 1536, 32, 59392, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 32, 128, 4, 1024):
  • DEFAULT_GPU = from_name[::StringSlice[::Bool():
  • DEFAULT_GPU_ARCH = _accelerator_arch():
  • DEFAULT_GPU_TARGET = from_name[::StringSlice[::Bool().target():
  • H100 = Info(__init__[__mlir_type.!kgen.string]("H100"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("hopper"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](9), __init__[__mlir_type.!kgen.string]("sm_90a"), 132, 32, 2048, 32, 64, 2048, 32, 233472, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 32, 128, 4, 1024):
  • L4 = Info(__init__[__mlir_type.!kgen.string]("L4"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("ada"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](8.9000000000000004), __init__[__mlir_type.!kgen.string]("sm_89"), 58, 32, 1536, 32, 64, 2048, 32, 102400, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 24, 128, 4, 1024):
  • MI300X = Info(__init__[__mlir_type.!kgen.string]("MI300X"), Vendor(__init__[__mlir_type.!pop.int_literal](1)), __init__[__mlir_type.!kgen.string]("hip"), __init__[__mlir_type.!kgen.string]("gfx942"), __init__[__mlir_type.!kgen.string]("amdhsa-code-object-version=5"), __init__[__mlir_type.!pop.float_literal](9.4000000000000003), __init__[__mlir_type.!kgen.string]("CDNA3"), 304, 64, 2048, 64, 32, 2048, 2, 65536, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 2, 128, 4, 1024):
  • NoGPU = Info(__init__[__mlir_type.!kgen.string]("NoGPU"), Vendor(__init__[__mlir_type.!pop.int_literal](0)), __init__[__mlir_type.!kgen.string]("none"), __init__[__mlir_type.!kgen.string]("no_gpu"), __init__[__mlir_type.!kgen.string](""), __init__[__mlir_type.!pop.int_literal](0), __init__[__mlir_type.!kgen.string](""), 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, __init__[__mlir_type.!kgen.string]("none"), 0, 0, 0, 0, 0, 0):
  • OrinNano = Info(__init__[__mlir_type.!kgen.string]("Orin Nano"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("ampere"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](8.6999999999999993), __init__[__mlir_type.!kgen.string]("sm_87"), 8, 32, 1536, 32, 64, 2048, 32, 167936, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 16, 128, 4, 1024):
  • RTX5090 = Info(__init__[__mlir_type.!kgen.string]("RTX5090"), Vendor(__init__[__mlir_type.!pop.int_literal](2)), __init__[__mlir_type.!kgen.string]("cuda"), __init__[__mlir_type.!kgen.string]("blackwell"), __init__[__mlir_type.!kgen.string]("nvptx-short-ptr=true"), __init__[__mlir_type.!pop.float_literal](12), __init__[__mlir_type.!kgen.string]("sm_120a"), 170, 32, -1, 32, 64, 1536, 32, 59392, 65536, 256, __init__[__mlir_type.!kgen.string]("warp"), 255, 65536, 32, 128, 4, 1024):

Structs

  • Info: Comprehensive information about a GPU architecture.
  • Vendor: Represents GPU vendors.

Functions

  • is_cpu: Checks if the target is a CPU (compile-time version).
  • is_gpu: Checks if the target is a GPU (compile-time version).
  • is_valid_target: Checks if the target is valid (compile-time version).