Funny you should mention that. I now have the name of the manager running the testing. I can drill down through the organization and find the project manager.
Seems IBM tested theirs for awhile on RC-5.
I'm not making any promises... but let me see what I can do. No harm in asking.
Let's see here... this machine was supposed to have 2,728 Alpha processors. If calculated at 667 Mhz and doing the same rate as a celeron (I have no idea about Alpha performance)
That should yield 1819576 Mhz or in other terms:
This machine will crack 5.09 Gk/sec or do 1,639,838 blocks per day.
I wonder how many days we can wrangle?
Ah, I think the Alpha's are more efficient than the PPro relatives... in fact I think they're either the PowerPC 603 or 604 from the speed calculator...
that would yield 6GKey/s or a total of 1,932,666WU's/day
/me crosses fingers and hopes you can get it for a day or two Dan
unfortunatly I have no stats to back this up, but I recall the alpha blew at rc5. I might be completly wrong but even if they cracked like 486's the darn thing would still put out keys!
Looking at Dnet's speed page (http://www.distributed.net/speed/), it appears that an Alpha at 667Mhz puts about about 1.34Mkeys/s (100Kkeys/s faster than a celery 450).
With 2728 processors, that works out to 3.66Gkeys/s....
Definitely would be a nice addition to the team!!!!!!!!!!!
If I remember correctly, IBM did test the world's fastest supercomputer on RC5 for like 6 days or so. In those 6 days it was faster than any other TEAM....pretty amazing....
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.