by Anton Shilov
08/11/2011 | 11:26 PM
Nvidia Corp. said it had learnt the lessons with 40nm process technology and would not repeat its mistakes with the 28nm fabrication process. The company spent time on learning the peculiarities of the 28nm manufacturing technology from Taiwan Semiconductor Manufacturing Company and already has working 28nm chips.
"We are far better prepared for 28nm than we were for 40nm. Because we took it so much more seriously. We were successful on so many different nodes for so long that we all collectively, as an industry, forgot how hard it is. So, one of the things that we did this time around was to set up an entire organization that is dedicated to advanced nodes. We have had many, many tests chips run on 28nm, we have working silicon," said Jen-Hsun Huang, chief executive officer of Nvidia.
TSMC's 40nm manufacturing process was plagued with low yields conditioned by problems with the technology itself, manufacturing issues and design errors. Nvidia started to develop its 40nm lineup without knowing about the potential issues and when it ran into a set of unexpected problems it had to delay launch of its key-products based on Fermi architecture. With 28nm process technology, the company seems to be better prepared.
"[Our experience with 28nm] is looking really good, it is looking much better than our experience with 40nm. It is just a comprehensive, across-the-board engagement between TSMC and ourselves making sure that we are ready for production ramp when the time comes. So I feel really good about 28nm," said Jen-Hsun Huang.
Nvidia said earlier this month that it plans to test-drive its next-generation Kepler graphics processing unit (GPU) this year and introduce the new chips commercially in 2012.
Kepler is Nvidia's next-generation graphics processor architecture that is projected to bring considerable performance improvements and will likely make the GPU more flexible in terms of programmability, which will speed up development of applications that take advantage of GPGPU (general purpose processing on GPU) technologies. Some of the technologies that Nvidia promised to introduce in Kepler and Maxwell (the architecture that will succeed Kepler) include virtual memory space (which will allow CPUs and GPUs to use the "unified" virtual memory), pre-emption, enhance the ability of GPU to autonomously process the data without the help of CPU and so on.