A link to the ISSCC keynote by Mark Horowitz, curator of cpudb. IMO worth a listening to for most anyone interested in CPU performance related issues. http://isscc.org/media/2014/plenary/Mark_Horowitz/NewStandar...

The plots shown in his presentation basically show off the underlying data in cpudb.

Interesting talk. ISSCC also offers an edx crash course

ISSCCx ISSCC Previews - Circuit and System Insights (https://courses.edx.org/courses/IEEEx/ISSCCx/3T2014/info)

Nope:

  Failed to connect to cpudb.stanford.edu port 80: Connection refused

Latest instance in the Wayback Machine. (Don't forget to support such resources.)

http://web.archive.org/web/20141013114803/http://cpudb.stanf...

----

Edit:

I don't know why this is at the top. I see now that couple of other people cited the archive in comments that are now 6 hours old:

https://news.ycombinator.com/item?id=8693001

https://news.ycombinator.com/item?id=8693004

(So send them the karma...)

The actual link to get the .tar.gz archive is http://cpudb.stanford.edu/cpudb.1416196069.tar.gz (on the download page, both links points to the .zip file). They are basically the same size anyway (~728kiB).

A few interesting charts showing the evolution of CPUs:

* Transistor size (http://cpudb.stanford.edu/visualize/technology_scaling): the process is still regularly scaling down

* Clock frequency (http://cpudb.stanford.edu/visualize/clock_frequency): processors do not tick significantly faster than 4GHz, a speed they reached in 2004/2005

* Performance (http://cpudb.stanford.edu/visualize/performance_by_freq_and_...): I am still unable to display this graph

The clock frequency graph shows a processor at 21.3GHz. It's the Z3480, whose clock actually ticks at 2.13GHz (http://ark.intel.com/products/70102/Intel-Atom-Processor-Z34...).

I love that it goes back to MOS 6502, Zilog Z80 etc. It also includes the WDC 65C02 which is still in production.

Interesting. Pity it does not feature anything DEC beyond Alpha.

Is it down? Any mirrors?

Wayback Machine: http://web.archive.org/web/20141013114803/http://cpudb.stanf...

It seems so. Here is the mirror on archive.org

https://web.archive.org/web/20141013114803/http://cpudb.stan...

I've seen this happening several times after submission to HN.

It would be slashdot effect if slashdot was still relevant.

On Spain is called "Meneame" effect