• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

1T command rate with DDR2

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Dooms101

Member
Joined
Dec 28, 2009
Location
under a heatsink
Uhm, so I recently read somewhere about how running ram at a 1T command rate (versus 2T stock) helps increase memory performance %6 - %18 depending on whats being tested. So... does anyone know what tricks will help DDR2 run at 1T? I haven't actually attempted to try 1T yet, I thought I'd ask first. Would I have to decrease the frequency? Up the voltage? Loosen the timings? Or is it as simple as toggling to 1T in the BIOS?

My ram: Product Page

Specs:
DDR2-1066mhz (5-6-6-18-24-2T) @ 2.1v
DDR2-800mhz (5-6-6-18-24-2T) @ 1.8v
DDR2-667mhz (4-5-5-15-20-2T) @ 1.8v
DDR2-533mhz (3-4-4-12-16-2T) @ 1.8v

My current max OC's:
DDR2-1120mhz (6-6-6-20-26-2T) @ 2.25v (210mhz bus / 3:8)
DDR2-1040mhz (5-5-5-15-24-2T) @ 2.29v (260mhz bus / 1:2)
DDR2-925mhz (4-5-5-15-20-2T) @ 2.25v (225mhz bus / 1:2)

And my DIMM's stay in the 25c - 30c temps at full speed and 2.3v
 
Consider: a CAS operation at five clocks with 2T takes seven clocks to do.
A CAS operation at five clocks with 1T takes six clocks to do.
The difference in time for the operation is 14.3%.
 
I should clarify that I meant 1-2% faster overall performance, not memory performance. That's the hit I measured when I had to run DDR memory at 2T instead of 1T to prevent errors.

Will a 14.3% faster command rate result in 14.3% overall memory speed when commands make up only a small proportion of the memory accesses?
 
Well... I actually got it running fine at all stock settings using 1T CR, and it ran about 5% faster (talking about memory performance,) and 1-2% faster system performance. Obviously I OC'ing was alot harder. My best performing OC with 2T was:
1015mhz (1:2 divider) - (5-5-5-15-24-2T) @ 2.23v

and using 1T:
880mhz (1:2 divider) - (5-6-6-18-26-1T) @ 2.29v

The 2T OC was about 5% better (or 10% then stock) then the 1T OC, with about 4% increase in system performance. Also... SiSoft SANDRA reported the channel efficiency of the 2T to be about 78% and 1T to be about 63% when OC'd... I am guessing thats what makes the difference. Oh well, it sounded cool at least
 
Back