Google parent company Alphabet introduced Trillium, a line of Artificial Intelligence data center chips. The company claims that the new version of the chip is almost five times faster than its previous one.

One of the few practical substitutes for Nvidia’s industry-dominating top-tier processors is Alphabet’s endeavor to develop specialized chips for AI data centers. The processors, in conjunction with the software that is directly linked to Google’s tensor processing units (TPUs), have enabled the corporation to capture a significant portion of the market.

About 80% of the market for AI data center chips is controlled by Nvidia, with Google’s TPUs accounting for the great majority of the remaining 20%. Instead of selling the chip itself, the company leases access through its cloud computing platform.

Google officials claim that the sixth-generation Trillium processor, which drives technologies that generate text and other media from huge models, will outperform the TPU v5e by a factor of 4.7 in terms of computational performance. The energy efficiency of the Trillium processor is 67% higher than that of the v5e.

By the end of 2024, cloud users will be able to buy the new processor. The Google engineers were able to increase the total bandwidth and high-bandwidth memory capacity in order to achieve even higher performance improvements. The enormous amounts of complex memory required for AI models have prevented performance improvements from happening much faster.

The chips are meant to be used in 256-chip pods, which the company claims might eventually be extended to hundreds of pods.

Read More : https://thecioconnect.com/