编程

huawei-cloud-ascend-op-mfu-calculator

试用

Calculate MFU (Machine FLOP Utilization) for operators like matmul/GEMM/FlashAttention on Ascend NPU, providing clear formulas and derivation process Use thi...

它能做什么

Calculate MFU (Machine FLOP Utilization) for operators like matmul/GEMM/FlashAttention on Ascend NPU, providing clear formulas and derivation process Use this skill when the user wants to: (1) calculate MFU for matrix operations, (2) analyze operator performance efficiency, (3) understand hardware utilization, (4) optimize operator implementation Trigger: user mentions "MFU", "machine flop utilization", "operator FLOPs", "matmul performance", "GEMM efficiency", "Ascend MFU", "算子MFU", "算力利用率", "矩阵乘效率", "GEMM性能", "FlashAttention性能"

技能文档

Huawei Cloud Ascend Operator MFU Calculator

Overview

This skill calculates MFU (Machine FLOP Utilization) for operators like matmul/GEMM/FlashAttention on Ascend NPU, providing clear formulas and derivation process.

Architecture: Input Validation → FLOPs Calculation → Achieved TFLOPs/s → MFU Calculation → Result Analysis

Related Skills:

  • huawei-cloud-ascend-profiler-db-explorer - Profiling data analysis for operator performance data

Prerequisites

  1. Python 3.8+ installed
  2. Basic understanding of FLOPs calculation concepts

Usage Scenarios

Typical Problem Scenarios:

  • Evaluating how well an operator utilizes Ascend NPU compute power
  • Comparing performance of different operator implementations
  • Identifying optimization opportunities for matrix operations

Typical User Utterances:

  • "Calculate MFU for my GEMM operator"
  • "What's the machine FLOP utilization for FlashAttention?"
  • "Analyze my matmul operator performance efficiency"

Workflow

  1. Input Collection: Gather operator parameters (matrix dimensions, data types, execution time)
  2. FLOPs Calculation: Compute theoretical FLOPs for the operation
  3. Achieved Performance: Calculate achieved TFLOPs/s from execution time
  4. MFU Calculation: Apply formula MFU = Achieved FLOPs / Peak FLOPs
  5. Result Analysis: Provide interpretation and optimization suggestions

MFU Calculation Formula

MFU = (Achieved FLOPs / Peak FLOPs) × 100%

Where:

  • Achieved FLOPs = Operation FLOPs / Execution Time
  • Peak FLOPs = Hardware-specific peak performance (e.g., Ascend 910B: 256 TFLOPs for FP16)

Reference Documents

DocumentDescription
Ascend 910B Series Technical SpecificationsOfficial Ascend 910B series product specifications
MFU Calculation MethodologyDetailed MFU calculation formulas and examples
FlashAttention Technical PaperOriginal FlashAttention research paper

Enhanced Features

Intelligent Bottleneck Diagnoser

  • AI-powered bottleneck diagnosis that analyzes profiling data to identify root causes automatically
  • Classifies bottlenecks into categories: memory-bound, compute-bound, communication-bound, or operator-fallback
  • Provides actionable optimization recommendations with priority ranking
  • Includes pattern matching for known performance anti-patterns

Parameter Confirmation

ParameterDescriptionRequired
operatorOperator type (matmul/flash_attention/gemm, etc.)Yes
flopsTheoretical FLOPs of the operatorYes
time_msOperator execution time (milliseconds)Yes
peak_tflopsHardware peak computing power (TFLOPS)Yes
deviceNPU device type (910B/910, etc.)No

相关技能

Develop and optimize custom operators using AscendC programming language. Analyze operator performance bottlenecks and conduct optimization validation. Based...

作者 huaweicloud-skills-team1 次安装

Collect operator-level performance data on Ascend NPU using msopprof tool. Supports both device mode and simulator mode, generates performance analysis repor...

作者 huaweicloud-skills-team1 次安装

用自然语言控制华为昇腾 NPU,本地或 SSH 远程执行 npu-smi 命令。

作者 huaweicloud-skills-team7 次安装

把自然语言性能问题转换为安全 SQL,查询 Ascend PyTorch Profiler / msprof 数据库中的算子、通信与调度耗时。

作者 huaweicloud-skills-team6 次安装

Migrate vision/detection/segmentation small models to Ascend NPU, covering the full workflow: model structure analysis, migration verification, performance p...

作者 huaweicloud-skills-team1 次安装

在华为云昇腾 910B DevServer 上按单机或双机(16 卡)拓扑部署并测试 LLM、VL、Embedding、Rerank 模型。

作者 huaweicloud-skills-team7 次安装

huaweicloud-skills-team 的更多技能

浏览全部技能

用自然语言控制华为昇腾 NPU,本地或 SSH 远程执行 npu-smi 命令。

作者 huaweicloud-skills-team7 次安装

在华为云昇腾 910B DevServer 上按单机或双机(16 卡)拓扑部署并测试 LLM、VL、Embedding、Rerank 模型。

作者 huaweicloud-skills-team7 次安装

面向华为云资源的只读查询能力,用于资源清点、核对与参数发现。

作者 huaweicloud-skills-team6 次安装

通过本地 Python SDK 只读查询华为云 IAM 资源(用户、用户组、策略、委托、AK/SK、MFA、安全设置)。

作者 huaweicloud-skills-team6 次安装