Message boards : Number crunching : E2180 vs. E4500, is a 2MB L2 Cache profitable ?
Author | Message |
---|---|
![]() ![]() Send message Joined: 29 Nov 05 Posts: 4 Credit: 7,332,084 RAC: 2,041 |
How big is the insufficient output with only 1MB Cache ? My ambition is, to get a high Faktor "WU's/consumption". I know such influence: http://www.matbe.com/divers/zoom/impact-cache-performances-core-2-duo/000000059546.png/ but in dont found this comprised with BOINC. The intention is a low power PC whith a appreciable output. At the moment my PC (with a E4400 with 2950 Mhz and vcore 1,24 V) need approximate 105 W. perhaps can someone help me with informations referring to this ? ![]() |
Tom Philippart![]() Send message Joined: 29 May 06 Posts: 183 Credit: 834,667 RAC: 0 |
I can't provide exact numbers, but a bigger L2 cache does have a big influence on performance in protein folding. On folding@home it nearly doubles the performance to have more than 2MB of L2 Cache. ![]() |
![]() Send message Joined: 3 Nov 05 Posts: 1833 Credit: 120,171,440 RAC: 18,177 ![]() |
i have an E6420 @3.2GHz (2MB per core here): https://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=481396 Taking a page of results from yours gives an average of 17.88 credits/hour (from: sum(results)/sum(time)*3600) Mine gets 22.31 c/h but is running faster, so: 22.31 / 3.2 x 2.95 = 20.57 credits/hour. I don't know how representative those samples are, and if you're running much software on yours it might reduce the score due to cache-thrashing, but it looks like the extra 1MB is worth around 13% from these numbers. HTH Danny |
![]() ![]() Send message Joined: 16 Jul 07 Posts: 132 Credit: 98,025 RAC: 0 |
|
Message boards :
Number crunching :
E2180 vs. E4500, is a 2MB L2 Cache profitable ?
©2025 University of Washington
https://www.bakerlab.org