deepseek-ai/DeepGEMM
File Complexity*
The distribution of files by complexity: the sum of the McCabe indexes of their units.
File Complexity Overall
  • There are 62 files with units, with 10,108 lines of code.
    • 0 very complex files (McCabe index sum 251+; 0 lines of code)
    • 1 complex files (McCabe index sum 126-250; 618 lines of code)
    • 4 medium complex files (McCabe index sum 51-125; 1,725 lines of code)
    • 9 slightly complex files (McCabe index sum 26-50; 1,762 lines of code)
    • 48 simple files (McCabe index sum 1-25; 6,003 lines of code)
  • 66 main files have no units (e.g. languages without unit analysis, or files without functions) and are not classified.
system0% | 6% | 17% | 17% | 59%
Legend:
251+
126-250
51-125
26-50
1-25


explore: grouped by folders | grouped by complexity
File Complexity per Extension
251+
126-250
51-125
26-50
1-25
py0% | 31% | 18% | 12% | 37%
hpp0% | 0% | 16% | 18% | 64%
File Complexity per Logical Decomposition
primary
251+
126-250
51-125
26-50
1-25
scripts0% | 63% | 36% | 0% | 0%
csrc0% | 0% | 16% | 18% | 64%
deep_gemm0% | 0% | 0% | 24% | 75%
Most Complex Files (Top 50)

The McCabe index of a file grows with its size: per 100 lines tells a large but plain file from a dense one, and most complex unit whether the complexity sits in one unit or is spread over many.

File# lines# unitsMcCabe index
(sum)		per 100
lines		most complex
unit
generate_pyi.py
in scripts
 618 19 203 33 40
runtime_utils.hpp
in csrc/jit_kernels/impls
 352 19 83 24 8
gemm.hpp
in csrc/apis
 806 29 77 10 9
360 11 66 18 11
sm100.hpp
in csrc/jit_kernels/heuristics
 207 8 60 29 35
sm90.hpp
in csrc/jit_kernels/heuristics
 166 6 42 25 24
einsum.hpp
in csrc/apis
 241 8 40 17 9
math.py
in deep_gemm/utils
 136 15 37 27 6
smxx_layout.hpp
in csrc/jit_kernels/impls
 255 7 32 13 11
attention.hpp
in csrc/apis
 389 10 31 8 10
bench.py
in deep_gemm/testing
 102 6 31 30 23
layout.hpp
in csrc/utils
 138 12 30 22 8
mega_gate.hpp
in csrc/jit_kernels/heuristics
 146 7 29 20 15
mega_moe.hpp
in csrc/jit_kernels/heuristics
 189 7 28 15 11
127 17 25 20 3
epilogue_class.hpp
in csrc/jit_kernels/impls
 89 20 23 26 4
__init__.py
in deep_gemm/mega
 179 8 23 13 7
smxx_cublaslt.hpp
in csrc/jit_kernels/impls
 181 8 21 12 5
sm100_sparse_mqa_logits.hpp
in csrc/jit_kernels/impls
 183 4 21 11 13
layout.hpp
in csrc/apis
 108 4 20 19 9
sm100_mqa_logits.hpp
in csrc/jit_kernels/impls
 314 12 19 6 4
a_fused_m_grouped_gemm.py
in deep_gemm/legacy
 73 3 19 26 13
a_fused_k_grouped_gemm.py
in deep_gemm/legacy
 73 2 18 25 13
b_fused_k_grouped_gemm.py
in deep_gemm/legacy
 72 2 18 25 13
config.hpp
in csrc/jit_kernels/heuristics
 155 9 17 11 3
m_grouped_gemm.py
in deep_gemm/legacy
 68 3 17 25 10
mlopart.py
in deep_gemm/utils
 97 10 17 18 4
mega_mhc.hpp
in csrc/apis
 208 3 16 8 13
sm100_fp8_fp4_gemm_1d1d.hpp
in csrc/jit_kernels/impls
 501 8 15 3 4
mega_moe.hpp
in csrc/apis
 330 5 14 4 6
runtime.hpp
in csrc/runtime
 69 8 14 20 3
sm100_mega_mhc.hpp
in csrc/jit_kernels/impls
 129 1 12 9 12
sm100_fp8_fp4_mega_moe.hpp
in csrc/jit_kernels/impls
 201 1 12 6 12
locality_domain.hpp
in csrc/runtime
 87 7 11 13 3
runtime.hpp
in csrc/jit_kernels/heuristics
 49 10 11 22 2
sm100_bf16_gemm.hpp
in csrc/jit_kernels/impls
 396 7 11 3 4
sm90_bf16_gemm.hpp
in csrc/jit_kernels/impls
 391 7 11 3 4
dist.py
in deep_gemm/utils
 53 3 11 21 6
numeric.py
in deep_gemm/testing
 37 3 10 27 4
48 2 8 17 7
sm100_bf16_mega_moe.hpp
in csrc/jit_kernels/impls
 136 1 8 6 8
sm90_fp8_gemm_1d2d.hpp
in csrc/jit_kernels/impls
 312 5 8 3 4
utils.hpp
in csrc/jit_kernels/heuristics
 23 2 7 30 4
sm100_bf16_mega_gate.hpp
in csrc/jit_kernels/impls
 75 1 7 9 7
sm90_fp8_mqa_logits.hpp
in csrc/jit_kernels/impls
 205 3 7 3 4
utils.py
in deep_gemm/testing
 32 3 6 19 3
mlopart.hpp
in csrc/runtime
 29 5 5 17 1
common.hpp
in csrc/jit_kernels/heuristics
 45 1 5 11 5
sm100_tf32_hc_prenorm_gemm.hpp
in csrc/jit_kernels/impls
 95 1 5 5 5
sm90_fp8_gemm_1d1d.hpp
in csrc/jit_kernels/impls
 200 3 5 3 3
About This Analysis
  • File complexity is the sum of the McCabe indexes (conditional complexity) of the units (methods, functions...) in a file.
  • Files are classified in five categories based on that sum: 1-25 (simple files), 26-50 (slightly complex files), 51-125 (medium complex files), 126-250 (complex files), 251+ (very complex files). The default bands are five times the unit complexity bands, as a file is a group of units.
  • Only main code files with units are classified. Files without units (languages without unit analysis, or files without functions) have no complexity to measure and are left out rather than counted as simple.
  • A very complex file holds many decisions, which often means it has too many responsibilities: a candidate to split, and a place where changes need more testing.
Learn more...