Intel Rendering Toolkit & OpenVINO AI GPU Performance On Intel Panther Lake's Xe3 B390

Written by Michael Larabel in Graphics Cards on 3 March 2026 at 12:30 PM EST. Page 2 of 4. 2 Comments.

👁 OpenVINO benchmark with settings of Model: Face Detection FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

First up was looking at the OpenVINO AI toolkit performance on the Intel graphics. OpenVINO does also support Intel's NPU but for this testing was looking just at the iGPU performance with the NPU model support being rather limited and not too reliable in my testing. We see immediately the great gains made by Lunar Lake and then with Panther Lake delivering much greater throughput too.

👁 OpenVINO benchmark with settings of Model: Face Detection FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

Besides much higher throughput with the Intel Arc B390 Xe3 graphics, the latency was also much lower for OpenVINO with this Core Ultra X7 358H laptop.

👁 OpenVINO benchmark with settings of Model: Age Gender Recognition Retail 0013 FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Age Gender Recognition Retail 0013 FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Detection FP16, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Detection FP16, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Weld Porosity Detection FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Weld Porosity Detection FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

Across the variety of OpenVINO benchmarks run on the Intel iGPUs, there was very nice uplift out of the Arc B390 graphics with Panther Lake. On a performance-per-Watt basis, Panther Lake was typically slightly ahead of Lunar Lake.

👁 OpenVINO benchmark with settings of Model: Vehicle Detection FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Vehicle Detection FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Vehicle Bike Detection FP16, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Vehicle Bike Detection FP16, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Machine Translation EN To DE FP16, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Machine Translation EN To DE FP16, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Face Detection Retail FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Face Detection Retail FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Road Segmentation ADAS FP16-INT8, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Road Segmentation ADAS FP16-INT8, Device: GPU, Benchmark Mode: Latency. Core Ultra X7 358H Panther Lake was the fastest.

👁 OpenVINO benchmark with settings of Model: Noise Suppression Poconet-Like FP16, Device: GPU, Benchmark Mode: Throughput. Core Ultra X7 358H Panther Lake was the fastest.

If you are doing AI development around OpenVINO and currently using Meteor Lake or older, upgrading to Panther Lake would be very worthwhile for the Arc B390 graphics.

2 Comments - Next Page

Page: 1 2 3 4 Next Page

URL: https://www.phoronix.com/review/intel-b390-openvino-render/2

⇱ Intel Rendering Toolkit & OpenVINO AI GPU Performance On Intel Panther Lake's Xe3 B390 - Phoronix

Intel Rendering Toolkit & OpenVINO AI GPU Performance On Intel Panther Lake's Xe3 B390