Artificial intelligence: Qualcomm, Nvidia fight for the top spot in AI chip efficiency tests
Qualcomm Inc’s artificial intelligence chips beat Nvidia Corp’s in two of three measures of power efficiency in new test data released Wednesday, while the Taiwanese startup won both in one category.
Nvidia dominates the training market for AI models with huge amounts of data. But once these AI models are trained, they are used more widely in so-called “inference,” performing tasks such as generating text responses to prompts and deciding whether an image contains a cat.
Analysts believe the market for inference chips in data centers will grow rapidly as companies incorporate artificial intelligence technologies into their products, but companies such as Alphabet Inc’s Google are already exploring how to cover the extra costs.
One of the biggest costs is electricity, and Qualcomm has used its history designing chips for battery-powered devices like smartphones to create the Cloud AI 100 chip, which aims to save power.
In test data released Wednesday by MLCommons, a benchmarking benchmark widely used in the artificial intelligence industry, Qualcomm’s AI 100 beat Nvidia’s flagship H100 chip in image classification based on how many data center server queries each chip can handle. watts.
Qualcomm’s chips achieved 197.6 server queries per watt compared to Nvidia’s 108.4 queries per watt. Neuchips, a startup founded by Taiwanese chip researcher Youn-Long Lin, came out on top with 227 queries per watt.
Qualcomm also beat Nvidia in object detection with a score of 3.2 queries per watt compared to Nvidia’s 2.4 queries per watt. Object recognition can be used in applications such as analyzing retail materials to see where shoppers go most often.
However, Nvidia came out on top in both absolute performance and power efficiency in the natural language processing test, the most commonly used artificial intelligence technology in systems like chatbots. Nvidia achieved 10.8 samples per watt, while Neuchips came in second with 8.9 samples per watt and Qualcomm was third with 7.5 samples per watt.
Read all the Latest Tech News here.
