chew*

Becasue at same exact speeds on same exact board 32m is 10 secs slower than deneb......

I did, I used thuban

Btw bruce has 2x512m as well. Your latency is faster....ram speed is not the total equasion for latency....cpu speed is. At your speed latency looks like it Bugged by 2ns which we have seen.

well look at massman myself and bruce. All our scores are consistent. Unelss of course we just don't know what we are doing

I think you may be victim of bugged latency >>>> do a compare here http://hwbot.org/result.do?resultId=1017603&title=Highest score of Massman

True but then again a mod actually benching this bench would be better at determining whats bugged and whats not. However it doesn't take much common sense to compare your own scores to others to figure out what is legit and whats not. Quite a few legit baseline results out there now.

bugged latency why would you keep posting bugged latency scores......your latency should be around 40+ if not more at your speeds.

bugged latency

bugged latency

No pressure

hmmm it looks right to me, most of these newer chips/boards have trouble with it (1/1). Based on the ddr 1 results it looks right on par. These newer cpu's just seem to be less effecient.

preety crazy latency with that 3800 chip bruce

Looking good, I forsee a new 1st place on ln2.

hmm weird I did report it jmke, bug with submission reporting?

yep bugged latency, not even close to a realistic score.

Nice run dumo :up:

I reported it 2 days prior.

considering thuban = 2000 c6 unganged 32m possible, I would not call 1600 c6 spectacular, not taking anything from your results, just saying the board could use some improvements

Where to start....... 9 sec on phenom II 550 in spi 1m? Heh ok. Where to start. First 512K was run, second the time is a saved run. Third not even close to realistic. Why is this score even still up. http://hwbot.org/community/submission/1016621_grgan_superpi_phenom_ii_x2_550_be_9sec_0ms

Oh lol, yah i've been setting timings for both since forever Even in asus Also the board I told you about via PM is like that as well.

Massman, I really don't like this bench. Not really indicative of real world bandwidth. I'm getting a rather large boost from less cache on venice versus sandiego, so maybe bh-5 is faster

Nice to see you finally join us

Was it tied into MRL?

Nice run, Just found me a sandy chip today, give tccd one more go then time to see if these IMC's can handle bh-5.

Try MRL values. They may be really slack. Also try booting at 234, loosening MRL and then using OC knob see if it goes higher. It's just a guess but that should just boost reads not write and latency, always suspected they run some internal latency that is not visible slack to get higher clocks.