Replace the "G" in GPU

I got into tech because I like video games. My first computer related job was selling machines at Computer City (RIP) so I could get discounts and build the ultimate gaming PC on 1998. Back then, 3d rendering was the next big thing and the newly introduced Voodoo Graphics cards were the must-have item!

Time passed, GPU capabilities grew and video games became more and more impressive, approaching photo realism in massive open worlds, but something new happened. People started to realize that they could use GPU parallel processing for a lot more than graphics. The parallel processing capabilities used for matrix transformations and vector calculations in your video games meant that these chips were perfectly suited for any task that required the same mathematical operations running simultaneously as many times and as quickly as possible.

I first noticed it with Crypto Miners. Suddenly the best cards were always sold out as crypto miners were buying them as soon as they came off the truck to string together in massive crypto mining farms. GPU prices skyrocketed on the demand. Scientists learned to use them to crunch data faster than was ever possible before, and Generative AIs depend on GPU parallel processing to sift through massive collected unstructured training data and spit out useful responses.

Meanwhile, in the gaming world, PC Game graphics are inherently tied to Console System generations. The rush to upgrade has been slowed by the interconnected markets of PlaysStation, Xbox, and Steam. In most cases, a 5 year old GPU will still play current releases. The "G" utility in GPU has become something of an "also"

The true value of the technology is in Parallel Algebraic Processing, so I propose we rename the product. "Algebraic Processing Unit" would be cool, but the term "APU" was already taken by AMD when they glued CPUs and GPUs together. "Parallel processing unit" could work but "PPU" is used in a lot of other adjacent fields like software and electricity. We also run into issues with "Mathematical Processing Unit" as MPU is taken.

So, where do we go? What's the proper name for these miracles of parallel processing? Leave your recommendations in the comments.