I was wondering why GPUs and CPUs are built on different chip sizes? 32mm vs. 28mm and now 22mm (as examples). Wouldn't it be easier and more cost effective to use the same chip sizes and tech to make GPUs and CPUs? I was also wondering if multiple cores and hyper threading might be useful to GPU development. If anyone could help me understand why these two technologies which seem similar are different in production etc. it would be much appreciated.
Reducing any chip size provides more yield per wafer plus reduces the thermal envelope. The reason GPUs and CPUs are different is because they are made by different companies with differing technological I.Q.s. You can be sure the GPU makers are doing their best to catch up.
Oh so GPU manufacturers are behind CPU manufacturers? So why doesn't Intel make GPUs like AMD? And why do nvidia and AMD only use tsmc and not a company that say intel uses?
Intel have their own fabs I believe and are almost certainly way ahead of everyone else hen it comes to fab technology.