With the more and more embedded special cores for artificial intelligence (AI) in processors, it is essential to quantify and compare their efficiency. This has been well understood by Primate Labs which has recently released Geekbench AI, a tool to test the AI performance of the processors used in desktops and laptops.
Understanding AI-Focused Processors
Some of the new processors also have Neural Processing Units (NPUs) which are used to handle Artificial Intelligence tasks. These cores are sometimes referred to in terms of TOPS (Tera Operations Per Second), a measurement which can make performance look the same across different chips. Still, the plain TOPS scores do not reflect the entire picture, and this is when the Geekbench AI steps in.
The New in Geekbench AI
Geekbench AI is the rebranded version of the earlier software by Primate Labs known as Geekbench ML. While “Machine Learning” (ML) might be a more precise term, “AI” resonates better with the general public and reflects the tool’s purpose: Expanding on the sub-focus area of AI performance measurement.
The tool is cross-platform, that means it can compare the efficiency of AI on different operating systems and with different hardware. This flexibility shows the performance of the AI algorithms in a given system and thus assist users in decision making.
A look at how the Geekbench AI benchmarks and evaluates the performance of a device
Geekbench AI breaks down test results into three categories: The details of CPU, GPU and NPU performance. This kind of segmentation enables understanding of the specific role of each element in the overall performance of AI systems. Surprisingly, NPUs are still an ASIC specialized in energy efficient AI, while GPUs, especially those incorporated into processors, have already become quite powerful in AI thanks to their SIMD capabilities.
Also, Geekbench AI is not only a measure of a device’s computing prowess. It assesses performance through three key metrics:It assesses performance through three key metrics:
Total Precision: Global performance of AI.
Average Precision: Reliability in the accomplishment of the work.
Quantified Scores: Quantifying the outcome of benchmarking process.
These metrics give a better understanding of how well AI is doing as compared to traditional processor benchmarks that differentiate between single core and multi-core performance.
A tool that will grow with AI
With progression in the field of AI, so is the case with Geekbench AI. At the moment, the ranking is based on the mobile devices, including smartphones and tablets, but desktop processors are to be added soon. This evolution can assist users to be in a position to cope up with the ever dynamic area of AI computing.
For the purpose of comparing the performance of various processors, the official Geekbench AI ranking also offers a continuously updating list of the results, which includes a number of devices, with more being added in the future.
In the contemporary world where AI is reaching deeper into computing, applications like Geekbench AI are indispensable for keeping abreast of the developments and select the right devices.