Cinebench 2024, Geekbench AI and UL Procyon:
After the 3DMark series benchmarks, let’s tackle a series of tests concerning tasks for productive use. Even if this is not the primary function we’ll be attributing to the graphics card, we couldn’t do without offering you a small panel of tests, especially concerning AI, which is becoming more and more present.
Cinebench 2024 GPU:
This is a benchmark we won’t be presenting again, so much so that it’s used for our CPU tests, but here, with version 2024, it’s our graphics card we’re going to put to the test. Please note that the version included in BenchMate will not work for this CPU test, so you’ll need to download the full version from the Maxon website. As you can see, Intel graphics cards are still not compatible with this benchmark.
We decided to integrate the Cinebench 2024 benchmark to test our GPUs, but unfortunately, Intel cards are not currently supported. These graphics will be expanded in the coming weeks with the addition of other RTX 4000 and AMD 7000 cards.
UL Procyon AI Image Generation:
UL Benchmark not only offers its 3DMark benchmark suite, but also UL Procyon, to meet the demand for Artificial Intelligence evaluation. This benchmark was chosen for its ease of implementation. The benchmark will generate 4 x 4 images and provide us with a score as well as a result in the form of the time, in seconds, required to generate an image. The software supports several AI inference engines, depending on the GPU used.
As we’re dealing here with entry-level models, we’ll be using the benchmark in Stable Diffusion 1.5 (FP16) mode. It is intended for mid-range GPUs. The test will create 16 images at a resolution of 512 x 512, in batches of 4. There’s a test called Stable Diffusion XL (FP16), where 16 images will be created, at 1024 x 1024, in batches of 1.
As you can see, we’ve indicated both results: one is expressed as a score and the other as the time, in seconds, required to generate an image. The more reliable the latter, the faster the image generation. The B580’s improvement on the previous generation is striking. It does better than the RTX 4060, but is not on a par with the RTX 4060 Ti.
On the “XL” version, even if the scores seem fairly identical, it’s the rendering in seconds per frame that needs to be particularly observed. We can see that it takes longer to generate an image, which is logical, but here it’s our ARC B580 that fares best.
The software has also just added a new test, the AI Text Generation Benchmark, which we’ll be adding to these tests shortly.
Geekbench AI 1.1.0:
As you may have gathered, Geekbench AI will measure AI performance in machine learning, deep learning and AI-specific workloads. The application can be run on the CPU, GPU or NPU, provided the latter is compatible. As is already the case for other applications, Geekbench AI will provide three scores: ” Full precision “, ” Half precision ” and ” Quantized “.
Since a new version has just been released, we’ve had to repeat a series of new tests, so we’re updating our tables and will be publishing them shortly.
