• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Ram timings

F

Filter

[H]F Junkie
Joined
Dec 30, 2001
Messages
9,531
how important are ram speed and timings to F@H

Does 2t or 1t make a huge difference etc.

i tried searching a little and really didn't find any concrete information i was looking for.
 
Generally it seems like speed > timings with f@h but yeah there's not much info one way or the other...other than anecdotal hypothesizing at best.

Something I posted in a different thread here in the DC forums...
zero2dash said:
Speed.
CAS doesn't mean much.

Good link I found through google. http://foldingforum.org/viewtopic.php?f=38&t=14427#p141738
Wrish said:
No, just go for lowest real latency. Ex:

"1600 MHz 8-8-8-24": CL8 / 1.600 GHz = 5 ns latency

"1333 MHz 7-7-7-21": CL7 / 1.333 GHz = 5.25 ns latency

"2000 MHz 8-8-8-24": CL8 / 2.000 GHz = 4 ns latency

"1600 MHz 6-7-7-18": CL6 / 1.600 GHz = 3.75 ns latency

In this case, the last choice calculates out to the least latency, although the imperfect scaling of some of the timings probably means the 2000 Mhz RAM is just as fast.
Click to expand...
Click to expand...

(I have no idea how he calculated the ns latency figures though.) :confused:
I will say I do get lower TPF by a few minutes and higher PPD by at least a few K with my 2000 9-9-9-24 than I did my 1333@1600 9-9-9-24. Comparing my findings vs. others with 1600 with lower timings (I believe capreppy has XMS3 1600 7-8-7-20something), my 2000 9-9-9-24 has better performance.

Again though, info is pretty much anecdotal and PG has not announced anything officially on which is better so YMMV. ;) IMHO, go for speed over latency. Why buy 8 or 7 CAS 1600 when 2000 with 9 CAS costs the same and gives you more leg room on OC? ;)

re: 1T v 2T, unless you're running with all dimm slots populated you should be able to run pretty much anything 1T. When I had my 920's and x58's and 1 of them was 6x2gb, I ran 2T. On my SR-2, I'm still running 6x2gb but it's only the red ram slots and technically 3x2gb per cpu so I can run 1T.
 
well i have my memory at 1764 at 8 8 8 24 at 2t

was more curious about the 2t vs 1t thing if i should lower my memory speed in favor of 1t.
 
2t vs 1t have a little impact but that depend on what plateform. What matter the most is a combination of faster bandwidth and lower latencies so aim for both at the same time.
 
zero2dash said:
(I have no idea how he calculated the ns latency figures though.) :confused:
Click to expand...

Using this example:

"1600 MHz 6-7-7-18": CL6 / 1.600 GHz = 3.75 ns latency

1.600GHz or 1600MHz is 1,600,000,000 cycles per second. Divide that by 1 and you get 0.000 000 000 625 seconds per cycle, or 0.625 nanoseconds per cycle. CL6 means you need six cycles, so 6 x 0.625 = 3.75 nanoseconds of latency.
 
You must log in or register to reply here.
Back
Top