AMD Ryzen AI 300 Series Dominates Intel Core Ultra 7 Lunar Lake Performance For Linux Developers & Creators

Written by Michael Larabel in Processors on 4 October 2024 at 01:30 PM EDT. Page 9 of 9. 37 Comments.
srsRAN Project benchmark with settings of Test: PUSCH Processor Benchmark, Throughput Total. Ryzen AI 9 365 - ASUS Zenbook S 16 was the fastest.
srsRAN Project benchmark with settings of Test: PDSCH Processor Benchmark, Throughput Total. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
Liquid-DSP benchmark with settings of Threads: 16, Buffer Length: 256, Filter Length: 512. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
Liquid-DSP benchmark with settings of Threads: 16, Buffer Length: 256, Filter Length: 512. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.

Going into this when spending $1400 USD on a Lunar Lake laptop and excited to see how the Core Ultra 7 200V series would perform under Linux, I was much more optimistic even with acknowledging just the 8-core offerings. But with the Core Ultra 7 256V even falling behind the AMD Strix Point laptops in performance-per-Watt was disappointing. At least it's a step forward for Intel generationally with power efficiency compared to their prior laptop processors.

TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
OpenVINO benchmark with settings of Model: Face Detection FP16-INT8, Device: CPU. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
OpenVINO benchmark with settings of Model: Person Detection FP16, Device: CPU. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
OpenVINO benchmark with settings of Model: Weld Porosity Detection FP16-INT8, Device: CPU. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
OpenVINO benchmark with settings of Model: Handwritten English Recognition FP16-INT8, Device: CPU. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
ONNX Runtime benchmark with settings of Model: yolov4, Device: CPU, Executor: Standard. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
ONNX Runtime benchmark with settings of Model: fcn-resnet101-11, Device: CPU, Executor: Standard. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.
ONNX Runtime benchmark with settings of Model: super-resolution-10, Device: CPU, Executor: Parallel. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.

The AMD Ryzen AI 300 series laptops were also performing better in CPU-based AI workloads thanks to the Zen 5 CPUs offering AVX-512.

CPU Power Consumption Monitor benchmark with settings of Phoronix Test Suite System Monitoring.

The CPU power is a big improvement generationally compared to prior Intel Core CPUs, especially for peak CPU power consumption.

Geometric Mean Of All Test Results benchmark with settings of Result Composite, Intel Core Ultra 7 256V Lunar Lake Linux Benchmarks. Ryzen AI 9 HX 370 - ASUS Zenbook S 16 was the fastest.

On a geo mean basis across all of the benchmarks, the Core Ultra 7 256V performance in 300+ benchmarks put the Lunar Lake performance similar to that of the Core i7 1280P Alder Lake laptop. The AMD Ryzen AI 300 series laptops were around 1.6x the performance of the Core Ultra 7 Lunar Lake in these benchmarks.

If you are predominantly just using a web browser without much multi-tasking or just running some simple Python scripts and other single-threaded programs without much performance sensitive work concurrently, the Intel Core Ultra 200V series comes out nice with its good performance on the four P cores and a big step-up for power efficiency compared to Meteor Lake and prior generations. But for those running a lot of creator workloads, code compilation, and other workloads that are typically multi-threaded, the Core Ultra 7 256V within the ASUS Zenbook S 14 wasn't impressive at all. With just four P cores and four low-power E cores without Hyper Threading, the multi-threaded performance fell short not just against AMD Zen 5 but even prior Intel Core CPUs as well. The power efficiency of Lunar Lake remained in good shape and a big improvement compared to prior Intel Core CPUs, but often was still falling behind the AMD Ryzen AI 300 "Strix Point" laptops in performance-per-Watt. In only a subset of the multi-threaded tests was the Core Ultra 7 256V coming out ahead of the AMD Ryzen AI 9 365 / Ryzen AI 9 HX 370 for the best power efficiency.

It's great to see Intel making significant gains in power efficiency but at least for Linux multi-threaded workloads or those running a lot of apps concurrently, it's hard to see much value. Especially with this ASUS Zenbook S 14 with Core Ultra 7 256V is of similar price to the AMD Ryzen AI 9 365 within the ASUS Zenbook S 16. The Xe2 graphics performance issues are also disappointing. Stay tuned to Phoronix to see what of these early Lunar Lake Linux woes are addressed in the near-term and how the Core Ultra 7 200V series is able to evolve over the longer term. For now at least for the vast majority of Linux users the Ryzen AI 300 series is much more compelling over Lunar Lake.

If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.

Related Articles
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.