ghost-in-the-machine/sample-report.json at main · mmazco/ghost-in-the-machine · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
{
  "report_name": "Llama-7B Inference Run",
  "timestamp": "2026-01-28T15:30:00Z",
  "device_info": {
    "chip": "Wormhole",
    "temp": 58,
    "power": 95
  },
  "l1_usage": {
    "used_bytes": 838860,
    "total_bytes": 1048576
  },
  "op_timeline": [
    {
      "op_id": 1,
      "name": "embedding",
      "cores": [[0,0], [0,1], [0,2], [1,0], [1,1], [1,2]],
      "duration_us": 120
    },
    {
      "op_id": 2,
      "name": "matmul",
      "cores": [[0,0], [1,1], [2,2], [3,3], [4,4], [5,5], [6,6]],
      "duration_us": 280
    },
    {
      "op_id": 3,
      "name": "layernorm",
      "cores": [[4,4], [4,5], [5,4], [5,5]],
      "duration_us": 65
    },
    {
      "op_id": 4,
      "name": "attention",
      "cores": [[0,0], [2,2], [4,4], [6,6], [8,8], [10,10]],
      "duration_us": 350
    },
    {
      "op_id": 5,
      "name": "softmax",
      "cores": [[3,3], [4,3], [5,3], [3,4], [4,4]],
      "duration_us": 45
    },
    {
      "op_id": 6,
      "name": "dropout",
      "cores": [[7,7], [8,7], [7,8]],
      "duration_us": 30
    },
    {
      "op_id": 7,
      "name": "linear",
      "cores": [[9,0], [9,1], [9,2], [10,0], [10,1], [10,2], [11,0], [11,1]],
      "duration_us": 190
    },
    {
      "op_id": 8,
      "name": "gelu",
      "cores": [[6,8], [6,9], [7,8], [7,9]],
      "duration_us": 55
    },
    {
      "op_id": 9,
      "name": "residual",
      "cores": [[2,6], [3,6], [2,7], [3,7]],
      "duration_us": 40
    },
    {
      "op_id": 10,
      "name": "output",
      "cores": [[5,10], [6,10], [5,11], [6,11], [7,10], [7,11]],
      "duration_us": 85
    }
  ],
  "perf_summary": {
    "total_ops": 10,
    "avg_utilization": 0.82,
    "total_duration_us": 1260
  }
}