by Anton Shilov
12/29/2008 | 01:59 PM
Nvidia Corp., a leading supplier of graphics adapters, is projected to completely transit its graphics processing units (GPUs) family to Taiwan Semiconductor Manufacturing Co.’s 40nm fabrication process by late Q3 2009. The move will be one of the most aggressive actions of Nvidia in the recent years.
According to Nvidia’s presumable roadmap for 2009, which was published by VR-Zone web-site, the company will introduce 5 graphics processing units made using 40nm process technology next year and will also bring-in core-logic set for Intel central processing units (CPUs) produced using 40nm fabrication tech. The new family of products will completely substitute the existing lineup, according to the news-story.
The new chips are code-named GT212, GT300, GT214, GT216, GT218 and iGT209. It is rumoured that the Nvidia GT300 will support DirectX 11 application programming interface (API). Other details regarding the new GPUs and IGP are unknown.
Earlier this year it transpired that TSMC provides lower capacities for thinner process technologies as fewer clients actually adopt them. Nvidia has already complained about inability of TSMC to produce enough chips using high-end process technologies. Considering the ongoing global economic slump, the semiconductor company may become even more conservative about the 40nm production capacities and its largest clients – AMD/ATI and Nvidia – will have to fight not only on the market of standalone discrete GPUs for market share, but also at TSMC for manufacturing capacities.
Nvidia did not comment on the news-story.