Core Ultra 9 285K And Ultra 5 245K Review: Intel Arrow Lake Tested

Intel Core Ultra 200s Arrow Lake 15
Running AI workloads natively on local hardware, or at the "edge" instead of in the "cloud," is just now emerging into the mainstream on PCs. As a result, reliable, easily repeatable benchmarks for these workloads aren't prolific just yet, though UL has already built a few into its Procyon benchmark suite.

UL Procyon AI Machine Vision Benchmark

Let's take a look at how these processors do with the UL Procyon AI Computer Vision benchmark, running solely on the CPU cores using Integer precision with all of the competitive processors, but also including the Core Ultra 200S' NPU.

procyon ai intel arrow lake core ultra 200s benchmarks

Procyon AI offers different frameworks and precision levels on any given hardware. The AMD Ryzen processors don't have any sort of NPU and were configured to use DirectML. The Core Ultra 200S series have the same NPU used in Intel's Meteor Lake mobile processors and support OneAPI. When using OneAPI, however, CPU performance was all over the map with the Core Ultra 200S series processors, fluctuating between 150-ish to over 500 points. We're not sure what was happening there just yet, but will continue investigating. When leveraging the NPU, however, performance was consistently high and clearly dominates all of the other desktop processors tested.

LAME XP Audio Encoding

In our custom LAME XP MP3 encoding test, we convert 16, large uncompressed WAV files to the MP3 format, which is a common use case for many end users, to provide portability and storage of their digital audio content. The LAME engine is an open-source MP3 audio encoder that is used widely in a multitude of third party applications. For this test, we created 16-copies of our own 223MB WAV file (an 11 minute Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME XP tool. Processing times are recorded below, listed in minutes:seconds. Shorter times equate to better performance.

lame intel arrow lake core ultra 200s benchmarks

The Core Ultra 5 245K and Ultra 9 285K performed well in our custom LAME audio encoding test. The Ultra 5's support for fewer threads resulted in a score slightly behind the 14th Gen Core i5, but the Ultra 9 led the pack, besting the Ryzen 9 9950X despite the latter's support for 32 threads.

Blackmagic RAW Video Encoding Speed

The Blackmagic RAW Speed Test is a CPU and GPU benchmarking tool that tests full-resolution Blackmagic RAW video decode performance. The tool can be used to evaluate the performance at various resolutions and bitrates on the CPU or using OpenCL on a GPU. We're reporting two results here, both using 8K resolution, but at differing bitrates and compression levels.

blackmagic 1 intel arrow lake core ultra 200s benchmarks


blackmagic 2 intel arrow lake core ultra 200s benchmarks

The Core Ultra 5 245K and Ultra 9 285K also performed well in the Blackmagic Speed test. Once again we see the Ultra 9 taking the pole position, besting every other platform we tested. The Ultra 5 also does well, but ultimately lands ever so slightly behind the 14th Gen Core i5-14600K.

x265 Video Encoding

The multi-threaded HWBOT x265 Benchmark is based on the open source x265 video encoder that leverages modern CPU instructions to accelerate video encoding. We tested the CPUs represented here with the 64-bit encoder using the default 1080p and 4K workloads.

x265 intel arrow lake core ultra 200s benchmarks

You'll notice there are mostly AMD systems in the chart above. Unfortunately, the MSI motherboard we used in our 14th Gen Intel test rig doesn't perform as expected when HPET (the High Precision Event Timer) is enabled, and HPET is required to run this test. HPET works properly on our Socket AM5 ASUS mobo, though, hence all of the Ryzen results above. Interestingly, the test ran properly on the Ultra 5 245K, but when we switched over the Ultra 9 285K -- without changing anything else in the test system -- this benchmark would take exceedingly long to launch and would fail. Since we had Ultra 5 245K data, you can see it hanging with the Ryzen 5 9600X with the more taxing 4K workload, but trailing the test of the systems at 1080p.

Cinebench 2024 Rendering Benchmark

Cinebench rendering performance test based on Cinema 4D from Maxon. Cinema 4D is a 3D rendering and animation tool suite used by animation houses and producers like Sony Animation and many others. It's very demanding of system processor resources and can utilize any number of threads, which makes it an excellent gauge of computational throughput. This is a multi-threaded, multi-processor aware benchmark that renders and animates 3D scenes and tracks the length of the entire process. The rate at which each test system was able to render the entire scene is represented in the graphs below.

cinebench intel arrow lake core ultra 200s benchmarks

Cinebench 2024 was a strong point for the Core Ultra 5 245K and Ultra 9 285K. Here, the Core Ultra 200S processors put up some of the best single-threaded results we have seen to date, with the Ultra 9 285K putting up THE best ST number. And the Ultra 9 leads the pack overall as well, with the highest multi-threaded result. The Ultra 5 245K clearly outruns the 14th Gen Core i5 as well.

POV-Ray CPU Ray Tracing Benchmark

POV-Ray, or the Persistence of Vision Ray-Tracer, is an open source tool for creating realistically lit images. We tested with POV-Ray's standard 'one-CPU' and 'all-CPU' benchmarking tools on all of our test machines and recorded the scores reported for each. Results are measured in pixels-per-second throughput; higher scores equate to better performance.

 
povray intel arrow lake core ultra 200s benchmarks

Our results with POV-Ray mirror what we saw with Cinebench. Here the Core Ultra 5 245K and Ultra 9 285K once again put up the best single-threaded results, with excellent multi-threaded scores. The Ultra 9 285K leads the pack and the Ultra 5 245K clearly outruns the 14th Gen Core i5.

Blender Rendering Benchmarks

Blender is a free and open source 3D creation suite that can handle everything from modeling, rigging, and animation to simulation, rendering, compositing and motion tracking. It has a purpose-built benchmarking tool that will track the time it takes to complete rendering a particular model (or models). We used the CPU-focused benchmark with all three models currently available...

blender intel arrow lake core ultra 200s benchmarks

These heavy rendering workloads run really well on the Core Ultra 200S series. Both the Core Ultra 5 245K and Ultra 9 285K perform well in Blender, though the Ryzen 9 9950X takes the lead overall. Still, the Core Ultra 5 245K and Ultra 9 285K outperform their previous-gen counterparts by relatively large margins and put up highly competitive numbers with all three models.

Y-Cruncher Multi-Threaded Pi Calculator

Y-Cruncher is a multi-core-capable tool that calculates the value of pi to a specified number of digits. In this case, we ran the tool on all threads and had the application calculate the value of pi to 1 billion digits. The values below are the time required to perform the calculation expressed in seconds. As a result, lower values indicate better performance.

ycruncher intel arrow lake core ultra 200s benchmarks

The Core Ultra 5 245K and Ultra 9 285K once again clearly outpace their previous-gen counterparts, but the Zen 5-based Ryzen processors, especially those with higher core / thread counts, take the lead in Y-Cruncher.

Related content