Taipei (TW) - TSMC will be constructing an new 300 mm wafer factory in Taiwan, and convert one of its factories into a Research & Development center for future 32 nm, 22 nm and 15 nm chips. If you want to read how next-gen GPUs are going to look like, read on.
Even though several western companies claim that they are the largest chipmaker in the world, in the end that title goes to Taiwanese Semiconductor Manufacturing Company, if you look at the output. This manufacturing giant has manufactured semiconductors for all the major players in the industry, including AMD, Intel, Microsoft, Nintendo, Nvidia, Sony, Sun Microsystems and many more.
But with growing complexity of chips, TSMC is feeling the heat, and its current "full-node-half-node" approach (TSMC is known for its middle-steps, like a 55 nm production process) is not something that isn’t likely to work in the future.
In order to keep up with the growing demands from its customers, TSMC decided to invest more than $5 billion to convert its 300 mm Fab 12 into an R&D facility. This research and development facility will explore techniques and technologies that are required to switch to 32 nm, 22 nm and 15 nm processes. The company said the facility will be used as a test shop for each new manufacturing process, after which tools and machines would transfer to its Fab 14.
Sound familiar? Intel uses a similar approach: The company develops new production technologies in its D1D Fab in Oregon. However, the factory combines research (22 nm), development (32 nm) and production (45 nm) under one roof. This setup allows the company to simply shift setups next door within the same building.
TSMC’s should impact future GPUs from AMD and Nvidia, UltraSparc processors from Sun and of course, components for next-gen consoles - all manufactured in more advanced manufacturing process.