Skip to content

Latest commit

 

History

History
61 lines (43 loc) · 1.78 KB

Apple-M1.md

File metadata and controls

61 lines (43 loc) · 1.78 KB

Examples

  • A14
  • M1

References

  1. Discover Metal enhancements for A14 Bionic
  2. Mesa driver details
  3. Dissecting the Apple M1 GPU: 1, 2, 3
  4. M1 Benchmarks
  5. M1 reverse engineering
  6. iGPU Cache Setups Compared, Including M1
  7. Reverse engineering the Apple G13 GPU architecture

Notes

  • FP16:
    • fp32 has same rate as fp16. [4]
    • that there is a penalty (of exactly one cycle) for switching between FP32 and FP16 operation. ref
    • FP32 ALU rate is half of FP16 rate on A14 (and earlier chips). That has not changed on A14. F32 ALU rate relative to F16 increased on M1.

Features

  • Ray tracing (software).

Specs

  • Local memory (L1): [6]

    • size: 32KB
    • latency: 43ns
    • bandwidth: 671 GB/s

  • L2 Cache: [6]

    • size: 1MB
    • latency: 76.3ns
    • bandwidth: 384 GB/s

  • System level cache (L3): [6]

    • size: 8MB
    • latency: 266ns
    • bandwidth: 134 GB/s

  • RAM: [6]

    • latency: 311ns
    • bandwidth: 50.4 GB/s

  • CPU to GPU bandwidth: 17 GB/s [6]

  • GPU to CPU bandwidth: 17.5 GB/s [6]

  • M1 8-core GPU:

    • max clock: 1278 MHz
    • 16 EU per core
    • 8 fp32/fp16 ALU per EU
    • 2 SFU per EU [4]
    • 1024 total ALUs (FMA/cy)
    • theoretical fp32/fp16 performance: 1278 MHz * 1024 ALUs = 1308*10^9^ FMA/s = 2.6 TFLOPS