Earlier this week, Intel teased the performance of its upcoming dedicated GPU. Now it has come to light that AMD’s near and dear and Taiwan’s own TSMC will be responsible for Intel’s foray into high-end graphics, the Intel Alchemist GPU.
Intel was once dead set on using its proprietary process nodes, but that’s not the case anymore. The latest offering by Intel that competes with Nvidia and AMD in their own ball game, the Intel Arc series, is going to use the services of Taiwan’s chipset foundry, just like AMD.
TSMC, or Taiwan Semiconductor Manufacturing Company, is the most valuable semiconductor company on the planet right now, and for a good reason. The world has been going through a chipset shortage which has affected almost every technological industry, phones, cars and computers alike. For Intel to get a hold of TSMC’s services right now could mean big things about the shortage and chipset availabilities or lack thereof.
The Intel Alchemist is supposed to be built upon TSMC’s 6nm Finfet (N6) process. We should note that N6 is on its way to becoming the most popular version of that architecture, which is currently the N7. The N7 is used on AMD’s entire series of Ryzen 5000 CPUs.This only leaves Nvidia using a different chipset manufacturer, which relies on Samsung for all of its needs.
The 6nm N6 isn’t the latest of TSMC’s offerings, which would be the 5nm N5. N6 is quite simple and offers EUV lithographic technology apart from being quite similar to the N7, so it should ensure that Intel remains competitive with Nvidia and AMD in the high-end graphics market.
The Intel Alchemist will be the first in a series of graphics cards that Intel releases, to be followed by Battlemage, Celestial and Druid to follow the initial Intel Alchemist launch. The series will be based on Intel’s new Xe-HPG microarchitecture and will be released in Q1, 2022.