Soulkeeper
Diamond Member
I built my uncle a comp a few weeks back
asus a7v333
athlonxp 1700+
256mb pc2700 ocz mem
since i built it i got to play with it a bit
here are the results of a few hours of playing with and benching with diff mem settings/timings
sandra
I left the fsb at 133 and the proc at default clock
mem voltages at the default because the asus board wouldn't let me change that (which is my only gripe against the asus a7v333) the original a7v has mem voltage jumpers but not this newer one
mem frequency cas/ras/cas-to-ras/cmd/ap Interleave turbo/normal sandra results
4:5 = 166 1:1 = 133
166 2.5-3-3 1cmd 6ap auto normal 2020/1912
166 2.5-2-2 1cmd 6ap auto normal 2040/1933
166 2.5-2-2 1cmd 6ap 4 bank normal 2040/1934
166 2.5-2-2 1cmd 6ap 2 bank normal 2022/1912
166 2.5-2-2 1cmd 6ap disabled normal 1991/1873
166 2.5-2-2 1cmd 5ap 4 bank normal 2042/1934
166 2.5-2-2 2cmd 5ap 4 bank normal 2033/1911
166 2-2-2 2cmd 6ap 4 bank normal NO BOOT (FROZE)
166 2-2-2 2cmd 6ap 2 bank normal LOCK UP (BSOD)
166 2-3-3 1cmd 5ap 4 bank normal 2032/1929
166 2-3-3 1cmd 5ap 4 bank turbo 2072/1970
166 2.5-2-2 1cmd 5ap 4 bank turbo 2076/1980
133 2.5-2-2 1cmd 5ap 4 bank normal 2022/1914
133 2-2-2 1cmd 5ap 4 bank turbo 2029/1930
the Turbo setting does more than change what it says it changes or it would not have had an effect on the scores yet it did
i also stress tested the stable ones with proc and mem burn-in stuff and had no stability at these above timings that didn't fail
5ap is faster than 6ap 1cmd is faster than 2cmd 4 bank is faster than 2 bank
I couldn't get 166mhz to run at fast timing so eventhough the sandra scores for the 166 at slower timings is higher than that of the 133 at fast timings I think the better timings are more important (since the comp would be running seti this would also help)
the memory claimed to be able to run at 166mhz 2-2-2 1cmd timings but wouldn't hold that at default voltage
if i could increase the voltage there is no doubt in my mind that it would hold the fastest timings
I couldn't help but wonder if the heat spreader on the memory was actually doing more harm than good
i didn't want to take my uncle's memory apart tho
and didn't mess with the fsb timings cause i didn't want to oc his system
don't get ocz mem
i am thinking that maybe the mem will have a burn in time
maybe after running at 133 2-2-2 for awhile it will take 166 2-2-2 on a later date
i didn't have the time to really run any gaming benchmarks I did run madonion on two different settings but forgot to write the scores down
i didn't put the sandra cpu bench results because their differences weren't really too profound
conclusion: the faster timings are far more valuable then higher clock frequencies and all
asus a7v333
athlonxp 1700+
256mb pc2700 ocz mem
since i built it i got to play with it a bit
here are the results of a few hours of playing with and benching with diff mem settings/timings
sandra
I left the fsb at 133 and the proc at default clock
mem voltages at the default because the asus board wouldn't let me change that (which is my only gripe against the asus a7v333) the original a7v has mem voltage jumpers but not this newer one
mem frequency cas/ras/cas-to-ras/cmd/ap Interleave turbo/normal sandra results
4:5 = 166 1:1 = 133
166 2.5-3-3 1cmd 6ap auto normal 2020/1912
166 2.5-2-2 1cmd 6ap auto normal 2040/1933
166 2.5-2-2 1cmd 6ap 4 bank normal 2040/1934
166 2.5-2-2 1cmd 6ap 2 bank normal 2022/1912
166 2.5-2-2 1cmd 6ap disabled normal 1991/1873
166 2.5-2-2 1cmd 5ap 4 bank normal 2042/1934
166 2.5-2-2 2cmd 5ap 4 bank normal 2033/1911
166 2-2-2 2cmd 6ap 4 bank normal NO BOOT (FROZE)
166 2-2-2 2cmd 6ap 2 bank normal LOCK UP (BSOD)
166 2-3-3 1cmd 5ap 4 bank normal 2032/1929
166 2-3-3 1cmd 5ap 4 bank turbo 2072/1970
166 2.5-2-2 1cmd 5ap 4 bank turbo 2076/1980
133 2.5-2-2 1cmd 5ap 4 bank normal 2022/1914
133 2-2-2 1cmd 5ap 4 bank turbo 2029/1930
the Turbo setting does more than change what it says it changes or it would not have had an effect on the scores yet it did
i also stress tested the stable ones with proc and mem burn-in stuff and had no stability at these above timings that didn't fail
5ap is faster than 6ap 1cmd is faster than 2cmd 4 bank is faster than 2 bank
I couldn't get 166mhz to run at fast timing so eventhough the sandra scores for the 166 at slower timings is higher than that of the 133 at fast timings I think the better timings are more important (since the comp would be running seti this would also help)
the memory claimed to be able to run at 166mhz 2-2-2 1cmd timings but wouldn't hold that at default voltage
if i could increase the voltage there is no doubt in my mind that it would hold the fastest timings
I couldn't help but wonder if the heat spreader on the memory was actually doing more harm than good
i didn't want to take my uncle's memory apart tho
and didn't mess with the fsb timings cause i didn't want to oc his system
don't get ocz mem
i am thinking that maybe the mem will have a burn in time
maybe after running at 133 2-2-2 for awhile it will take 166 2-2-2 on a later date
i didn't have the time to really run any gaming benchmarks I did run madonion on two different settings but forgot to write the scores down
i didn't put the sandra cpu bench results because their differences weren't really too profound
conclusion: the faster timings are far more valuable then higher clock frequencies and all