CNFET Rivals Silicon Performance

Baldrson writes "Applied Physics Letters is carrying a paper on a CNFET (carbon nanotube field-effect transistors) advance that now rivals silicon performance for both n and p type devices. There is also a New York Times article in which it is reported that "it would be two to three more years before I.B.M. was ready to work on prototypes of future nanotube chips and as many as 10 years before they would be commercially available". This is may be what's at the end of the road for CMOS."

Comments

The end of the road for CMOS?

[inkfox]

This could be the end of the road for processor architecture as we know it.

We're getting to the point where we switch faster and faster, but it still takes time for signals to propagate! Something's going to have to change in the fundamental design of the processor. We're talking exceptionally deep pipelines here, with changes in our data source meaning a huuuuuuuge penalty while dozens of stages are dumped in favor of a new code/data stream. If we're to take fullest advantage of these new architectures, we're talking about structuring computing around SIMD-style programmed tunnels which execute identical operations on fat streams of data, not depending on the results for execution control. This would be akin to writing programs the way you write shaders on modern graphics hardware.

Whether these changes can happen without fundamentally restructuring the way we program remains to be seen. But we're fast getting to the point where something's got to give if we're to take fullest advantage of these new technologies.

Or maybe this just means cycle tuning and assembly are coming back in style. :)

Re:The end of the road for CMOS?

[jmv]

The end of the road for CMOS?

From what I understand, it's not really the end of CMOS. Their transistor seems to be MOSFET (Metal-Oxide-Semiconductor Field Effect Transistor) on carbon nanotubes instead of silicon. That would mean you could still do CMOS (Complementary MOS) with it (maybe called CN-CMOS).