![]() |
|
|
#1 |
|
Aug 2002
223 Posts |
Greetings,
I have a TM5600 CPU (an RLX Server Blade) that I installed the client on. The LiveSiever reports that it doesn't know the CPU type, and doesn't know about the cache. Some info: http://www.transmeta.com/technology/specifications/tm5600.html (My specific CPU) Developer Info: http://www.transmeta.com/developers/crusoe_docs.html Would this be something that can be added to the client? Also, what kind of hit am I taking with an unknown CPU type? Thanks so much! -Paulie |
|
|
|
|
#2 | |
|
Bamboozled!
"πΊππ·π·π"
May 2003
Down not across
250018 Posts |
Quote:
The hit should be quite small. The code in question tries to work out how big the L1 and L2 caches are and then arrange the sieving in appropriately sized chunks so that cache hits are maximised. The default is conservatively sized and so if the caches are bigger than the default you will take a performance hit. Observation has shown that it's generally quite small, on the order of a few percent. Paul |
|
|
|
|
|
#3 |
|
Aug 2002
223 Posts |
Great, thanks so much. I have a few of these in a lab environment that I could use, if I'm getting 98% thats super. At 667Mhz, we're not talking screamers here by any stretch. :)
On a side note, is seems to be an issue running the NFSgui via a Terminal session. Since these blades are headless, when I run the gui, it works, shows updates to the progress bar, but the benchmarks never show, and the status line at the bottom says "it appears the client is stopped", even though it's running and updating. I run the CLI on a Mac at home, never used the GUI before. :) |
|
|
|
|
#4 |
|
Bamboozled!
"πΊππ·π·π"
May 2003
Down not across
250018 Posts |
I have tried coding the information given on the TM web pages for cache sizes, etc, but don't have any hardware to test it on. Prospective testers, please mail me.
Paul |
|
|
|
|
#5 |
|
Aug 2002
223 Posts |
Sent PM to ya Xilman.
|
|
|
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Factor found that should have been found by P-1 | tha | Data | 65 | 2020-08-05 21:11 |
| found one! what do I do now? | jord161 | PrimeNet | 10 | 2018-06-22 21:57 |
| New PRP found | ltd | Prime Sierpinski Project | 12 | 2008-07-31 17:29 |
| Found something here | em99010pepe | ElevenSmooth | 8 | 2006-01-02 10:12 |
| M42 found? | Unregistered | Data | 1 | 2005-02-19 18:20 |