Leaderboard Β· All comparisons Β· LynxBenchAI 26Q3 Index

Intel Graphics (Intel Core Ultra 7 165U) vs NVIDIA GeForce MX550 for AI

Device pages: Intel Graphics (Intel Core Ultra 7 165U) Β· NVIDIA GeForce MX550

On the LynxBenchAI 26Q3 Index, the Intel Graphics (Intel Core Ultra 7 165U) scores 1.19x the NVIDIA GeForce MX550 overall (Training 0.78x, Inference 1.54x, Compute 1.96x). It leads on Inference and Compute and trails on Training. The two runs used different platforms (native Linux and WSL).

Scores on the LynxBenchAI 26Q3 Index

Ratio = the Intel Graphics (Intel Core Ultra 7 165U)'s score divided by the MX550's
ScoreRatioIntel Graphics (Intel Core Ultra 7 165U)MX550
Overall (GT)1.19x7.596.35
Training0.78x6.618.52
Inference1.54x8.895.78
Compute1.96x8.224.19
  • Runs on the Intel Graphics (Intel Core Ultra 7 165U) only: FP8 inference.

When to Choose Which

  • Choose the MX550 for training (1.29x the Intel Graphics (Intel Core Ultra 7 165U)).
  • Choose the Intel Graphics (Intel Core Ultra 7 165U) for inference (1.54x the MX550).
  • Choose the Intel Graphics (Intel Core Ultra 7 165U) for raw compute (1.96x the MX550).

From the LynxBenchAI 26Q3 Index category scores, which are not raw throughput; a lead under 10% counts as level.

Per Model and Precision

LynxBenchAI score per model and precision; ratio = Intel Graphics (Intel Core Ultra 7 165U) / MX550
ModelPrecisionRatioIntel Graphics (Intel Core Ultra 7 165U)MX550
compute
Dense MatMulBF161.16x2.462.13
Dense MatMulFP163.19x2.450.77
Dense MatMulFP321.05x4.574.36
Dense MatMulFP640.58x0.490.85
Sparse MatMulFP325.55x41.327.44
Sparse MatMulFP640.47x4.228.96
inference
DistilBERTFP162.92x14.925.12
DistilBERTFP8n/a0.01not run
DistilBERTINT81.16x10.729.21
DistilGPT2FP162.93x15.945.45
DistilGPT2FP8n/a0.01not run
DistilGPT2INT81.19x9.888.33
Flan-T5 SmallFP163.45x20.015.81
Flan-T5 SmallFP8n/anot runnot run
Flan-T5 SmallINT80.75x13.8618.58
GraphGPS (Peptides)FP166.82x7.851.15
GraphGPS (Peptides)FP8n/a0.22not run
GraphGPS (Peptides)INT816.03x5.210.33
MobileNetV3FP160.51x21.0241.47
MobileNetV3INT80.31x6.8221.89
training
DistilBERTBF161.00x6.266.24
DistilBERTFP321.22x12.179.95
DistilGPT2BF160.97x4.845.00
DistilGPT2FP321.16x10.489.04
Flan-T5 SmallBF160.96x5.756.02
Flan-T5 SmallFP320.95x10.7111.27
GraphGPS (Peptides)BF160.60x1.833.06
GraphGPS (Peptides)FP320.65x4.807.45
MobileNetV3BF160.81x5.556.89
MobileNetV3FP320.27x7.1326.84

Related Comparisons

Each figure is the first device's overall score divided by the second's.

Measure your own machine: pip install lynxbench-ai Β· Install guide β†’

How to Read These Results

A LynxBenchAI score is sustained throughput on a fixed catalogue of PyTorch models, measured after a warm-up at a batch size raised until throughput stops improving, so it is what the machine holds under load, not its best instant. Training, Inference and Compute are three separate scores; GT combines them (Training and Inference weigh 0.4 each, Compute 0.2).

Every run uses the stock, publicly released PyTorch stack, so a result is broadly indicative of the out-of-the-box PyTorch experience on that hardware, and so of custom PyTorch models too. It does not run your own model, and it is not a vendor-tuned peak like an MLPerf submission.

Scores compare only within one Index: the 26Q3 Index fixes the models, precisions, correctness thresholds and scoring, and a later Index may change them. Each test is one timed window, the scores have no 0-100 scale, and a precision a device cannot run scores zero rather than being skipped.

LynxBenchAI is built by TechnoLynx, which the European Commission engaged as an independent expert advisor on AI performance and benchmarking for AI Gigafactories under the EuroHPC Joint Undertaking (case study).

Questions

Is the Intel Graphics (Intel Core Ultra 7 165U) faster than the NVIDIA GeForce MX550 for AI?
On the LynxBenchAI 26Q3 Index, the Intel Graphics (Intel Core Ultra 7 165U) scores 1.19x the NVIDIA GeForce MX550 overall (Training 0.78x, Inference 1.54x, Compute 1.96x). It leads on Inference and Compute and trails on Training. The two runs used different platforms (native Linux and WSL).
Should I choose the Intel Graphics (Intel Core Ultra 7 165U) or the NVIDIA GeForce MX550?
Choose the MX550 for training (1.29x the Intel Graphics (Intel Core Ultra 7 165U)). Choose the Intel Graphics (Intel Core Ultra 7 165U) for inference (1.54x the MX550). Choose the Intel Graphics (Intel Core Ultra 7 165U) for raw compute (1.96x the MX550). These follow the LynxBenchAI 26Q3 Index category scores, which are not raw throughput.
Where is the difference between the Intel Graphics (Intel Core Ultra 7 165U) and the NVIDIA GeForce MX550 largest?
On the LynxBenchAI 26Q3 Index, the ratio is highest on GraphGPS (Peptides) at INT8 (Inference), where the Intel Graphics (Intel Core Ultra 7 165U) scores 16.03x the NVIDIA GeForce MX550, and lowest on MobileNetV3 at FP32 (Training), at 0.27x.
Which precisions run on only one of the Intel Graphics (Intel Core Ultra 7 165U) and the NVIDIA GeForce MX550?
On the LynxBenchAI 26Q3 Index, only the Intel Graphics (Intel Core Ultra 7 165U) runs FP8 inference; a precision a device cannot execute scores zero.
What does this result tell me about my own PyTorch model?
Every run uses the stock, publicly released PyTorch stack, so a result is broadly indicative of the out-of-the-box PyTorch experience on that hardware, and so of custom PyTorch models too. It does not run your own model, and it is not a vendor-tuned peak like an MLPerf submission.

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