LynxBenchAI Results

AI hardware ranked by Global Score (GT), with per-category breakdown. Provisional results — not the final, official release.

Get Started →
# Device Accelerator / GPU GT Global Score Training Category Inference Category Compute Category Platform OS
1
NVIDIA B200
1264.11
1065.97
2203.38
585.14
Native Linux
2
NVIDIA B300 SXM6 AC
1218.26
1057.55
2085.33
551.76
Native Linux
3
NVIDIA H200
1013.59
798.77
1899.59
464.68
Native Linux
4
NVIDIA H100 80GB HBM3
913.53
700.38
1659.36
471.03
Native Linux
5
NVIDIA RTX PRO 6000 Blackwell Server Edition
702.50
583.16
1177.35
362.95
Native Linux
6
NVIDIA GeForce RTX 5090
585.52
511.71
969.64
279.54
Native Linux
7
NVIDIA GeForce RTX 4090
441.85
332.78
648.22
361.93
Native Linux
8
NVIDIA A100-SXM4-80GB
431.56
411.48
652.75
207.50
Native Linux
9
NVIDIA A100 80GB PCIe
396.50
383.74
578.03
199.17
Native Linux
10
NVIDIA L40S
390.19
314.88
518.49
339.34
Native Linux
11
NVIDIA GeForce RTX 3090
199.21
206.19
290.74
87.31
Native Linux
12
NVIDIA GeForce RTX 3080
157.47
162.69
237.65
64.77
Native Linux
13
NVIDIA GeForce RTX 3080 (WSL)
156.12
156.52
228.56
72.46
WSL
14
NVIDIA GeForce RTX 5060 Ti
142.76
132.41
234.64
61.42
Native Linux
15
NVIDIA RTX 2000 Ada Generation Laptop GPU
73.67
60.47
99.69
59.72
WSL
16
AMD Radeon RX 9060 XT
73.58
36.99
119.43
110.65
Native Linux
17
Intel Graphics (Intel Core Ultra 7 165U)
8.02
6.65
9.15
8.95
Native Linux
18
NVIDIA GeForce MX550
5.87
8.47
4.88
4.07
WSL
19
AMD Ryzen 5 9600X
4.90
8.63
2.42
6.48
Native Linux
20
Intel Graphics [0x46a6] (12th Gen Intel Core i7-1270P)
0.43
0.31
0.00
3.83
WSL

The Personal Edition submits your benchmark results to TechnoLynx's servers as part of participating in the public leaderboard. If that isn't acceptable for your use case, contact TechnoLynx about Press, Pro, or Enterprise licenses.

Get Started

LynxBenchAI End-User License Agreement · LynxBenchAI Privacy Policy

Provisional results — not the final, official release · GT = LynxBenchAI Global Score, higher is better. Category columns aggregate the test-case and precision results measured under a declared, bounded optimisation budget. Bars are scaled to each column's own leader, not a shared ceiling -- compare bars down a column, not across.