Take a look at this:

http://hwbot.org/result.do?resultId=873225 (note the 64kB L2)

 

I would understand if it was CPU-Z bug, but the performance product in PiFast sucks aswell

http://hwbot.org/listResults.do?cpuModelId=505&applicationId=6&filterUser=true&filterBlocked=true&limit=100

Could be a semi-dead chip, I've seen other AXP's losing half of their cache before.

Looks like Massman is right

 

Here's my XP1700+ Palomino result: 11x190 2-2-2-7: 48.406s (256kB L2)

http://hwbot.org/result.do?resultId=875152

 

Exactly the same result at 95MHz lower frequency but same memory timings/clockspeed. It must be the cache issue

I've recieved another prime examble of FUBAR'ed L2 cache:

http://hwbot.org/result.do?resultId=875223

 

Well, at least I got away with shear frequency

Hi,

 

I also had this in the past, when I tried to add L2 cache on a barton. It turned out that I removed some cache:

 

http://www.hwbot.org/result.do?resultId=567862

Hi,

 

I also had this in the past, when I tried to add L2 cache on a barton. It turned out that I removed some cache:

 

http://www.hwbot.org/result.do?resultId=567862

 

That one seems to be in the wrong category btw, if it's a Barton:p

No actually some of the thorton were barton cores, and I had one. After removing the L2 cache it became a duron....