The Image Sensors 2009 conference wrapped up in San Diego recently.
Attendance was very low -- no surprise in these tough economic times. Despite that, there were several good presentations that offered insight into new directions at some companies worth watching -- like Texas Instruments.

http://semicondr.com/Blog/Entries/2009/10/9_RE-cycle.html

We get our first look inside the production version of Intel's 32nm
process thanks to Dick James and his colleagues at Chipworks who
posted some images of the very first teardown of a device from the
Westmere product family - specifically the Clarkdale desktop CPU.

After some brief comments on Intel 32nm, there is a preview of the
Image Sensors 2009 Conference set for next week in San Diego.

http://semicondr.com/Blog/Entries/2009/10/9_RE-cycle.html

Keeping technology cycles to exactly two years (and announcements to under two weeks) Intel announced their working 22 nm SRAM test chips at the Intel Developer Forum in San Francisco. In addition to the smallest SRAM cell size ever reported, several other key details were disclosed:

• SRAM cell size of 0.092 square microns
• 22 nm shuttle chips are more than just SRAM including mixed signal components like PLLs
• Continuing with 193 nm immersion, several layers of double patterning will be added to the 32 nm flow
• PMOS saturation currents continue to close the gap on NMOS

http://www.semicondr.com/Blog/Entries/2009/9/22_Intel_22nm_SRAM_Wafer_Shown_at_IDF.html

In my early days as a reverse engineer at Chipworks, one of my most interesting (and challenging) projects involved materials identification and recreation of the processing sequence for a blue emitting LED. That was way back in 1998, so the idea of broad spectrum solid-state lighting was really new and the materials to create them - GaN for my particular device of interest - were truly exotic.

To read the rest of this entry, please visit Don's blog at http://semicondr.com/Blog/Entries/2009/9/10_LED_Teardowns.html.