Intel Lunar Lake vs. AMD Strix Point Platform Profile Performance Comparison

Written by Michael Larabel in Computers on 16 October 2024 at 11:30 AM EDT. Page 5 of 6. 8 Comments.
GROMACS benchmark with settings of Implementation: MPI CPU, Input: water_GMX50_bare. Ryzen AI 9 365: balance_performance was the fastest.
GROMACS benchmark with settings of Implementation: MPI CPU, Input: water_GMX50_bare. Ryzen AI 9 365: balance_performance was the fastest.
GROMACS benchmark with settings of Implementation: MPI CPU, Input: water_GMX50_bare. Ryzen AI 9 365: balance_performance was the fastest.

For those frequently doing demanding work while on battery power, hopefully the "power" profile numbers provide some nice insight.

NAMD benchmark with settings of Input: ATPase with 327,506 Atoms. Ryzen AI 9 365: balance_performance was the fastest.
NAMD benchmark with settings of Input: ATPase with 327,506 Atoms. Ryzen AI 9 365: balance_performance was the fastest.
NAMD benchmark with settings of Input: ATPase with 327,506 Atoms. Ryzen AI 9 365: balance_performance was the fastest.

For the more grueling workloads, the Ryzen AI 9 365 with 10 cores / 20 threads was much more performant than the Core Ultra 7 256V at 8 cores.

Apache Cassandra benchmark with settings of Test: Writes. Ryzen AI 9 365: balance_performance was the fastest.
Apache Cassandra benchmark with settings of Test: Writes. Ryzen AI 9 365: balance_performance was the fastest.
Apache Cassandra benchmark with settings of Test: Writes. Ryzen AI 9 365: balance_performance was the fastest.

The Core Ultra 7 256V has a lot of interesting edge potential in small devices for edge compute performance at ~5 Watts or less in the power profile mode.

TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 365: balance_performance was the fastest.
TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 365: balance_performance was the fastest.
TensorFlow benchmark with settings of Device: CPU, Batch Size: 64, Model: ResNet-50. Ryzen AI 9 365: balance_performance was the fastest.

The AMD Ryzen AI 300 series with AVX-512 support and more cores/threads continued yielding better performance and performance-per-Watt for TensorFlow and other AI workloads.

Related Articles
System76 Thelio Astra Reviewed: High-End ARM64 Developer Desktop
Intel Core Ultra 7 256V Lunar Lake With ASUS Zenbook Performing Better After New Linux Patch
Intel Announces Core Ultra 200S Arrow Lake CPUs
NVIDIA RTX 6000 Ada Generation vs. Radeon PRO Performance On Ubuntu Linux 24.04 LTS
TUXEDO Sirius 16 Gen 2 Linux Laptop Pairs The AMD Ryzen 7 8845HS With Radeon RX 7600M XT
Framework Laptop 13 With Intel Core Ultra Series, 2.8K 120Hz Display