AMD Bolstering Radeon RX 6000 Series Performance with “Infinity Cache”?

The FPS Review may receive a commission if you purchase something after clicking a link in this article.

Image: AMD

momomo_us has spotted an interesting trademark filed by AMD last month that alludes to a new technology dubbed “Infinity Cache.” The document suggests that Navi 2X GPUs (i.e., Radeon RX 6000 Series) could be outfitted with a special memory cache for increased performance.

This is cool news for those of you who have bought into the popular theory that red team’s flagship RDNA 2 GPU (Radeon RX 6900 XT?) will only sport a 256-bit memory bus, which critics believe could hamper performance. (In comparison, NVIDIA’s GeForce RTX 3080 flaunts a 320-bit memory interface width.)

A subsection confirms that the trademark is related to “graphics processors; video graphics processors; graphics processing unit (GPU); GPU cores; graphics cards; video cards; video display cards; video capture cards; accelerated data processors; accelerated video processors,” and more.

It’s too early to tell how Infinity Cache works or what connection it has to Zen’s Infinity Fabric technology (AMD’s high-performance CPU interconnect architecture), if any, but this could be the secret sauce that puts the Navi 21 variants in line with green team’s flagship GPU.

“I was told that on the highest-end SKUs, we’re looking at up to 128 MB of Infinity Cache,” reported RedGamingTech. “It basically aims to reduce the data duplication across caches.”

Tsing Mui
News poster at The FPS Review.

Recent News