Overclock.net banner

1 - 20 of 23 Posts

·
Registered
Joined
·
1,806 Posts
Discussion Starter #1
Well , i was wondering wich is the best?
I think it would be the one that is faster , but you guys are the expertees....
And also , Is 800mhz the faster RAM speed that exists?
Thanks
 

·
Registered
Joined
·
514 Posts
I think if your board, ram, and processor allow it you can run 1066? however I have been away from overclocking awhile?
 

·
Premium Member
Joined
·
5,809 Posts
Intel likes higher speed, while AMD utilizes tighter timings better.

And yes 800 is the fastest supported ram. Though good overclocking ram can come in 800+...
 

·
Premium Member
Joined
·
4,423 Posts
Thats wrong Corsair makes a ddr1066 RAM. Also I just read an article in MaxPC that said that the 6400 C3 vs the 8500 C5 was a tie. the 6400 had tighter timings and of course the 8500 has a higher clock. IMO higher FSB would seem better with fairly tight or at least as tight as they can go is the best
 

·
Registered
Joined
·
1,806 Posts
Discussion Starter #6
It's not for my sistem , i just wanna know....
 

·
Banned
Joined
·
16,364 Posts
ddr2 800 is NOT the fastest supported RAM, ATM 1066 is. All new Asus boards are certified to use kingston hyperx 1066 RAM. Most 800 mhz RAM can oc close to 1000mhz, but the latencies of the RAM can increase the efficiency of the RAM's clock speed. Ultimately-> fastest speed with the lowest latency will always be the fastest for any system, Intel or Amd.
 

·
I remember when......
Joined
·
3,812 Posts
There is no reason to run your RAM faster than your FSB. So the answer is, the fastest is the tightest timings you can achieve at 1:1 ratio.

Running 1066 RAM on a 300FSB system with 6-6-6-15 timings will not be better than running 600 speed RAM with 4-4-4-12 timings because the system will not be able to use anything more than 600 of the 1066 anyway.
 

·
Banned
Joined
·
16,364 Posts
Quote:
There is no reason to run your RAM faster than your FSB. So the answer is, the fastest is the tightest timings you can achieve at 1:1 ratio.
If thats true then why do bench mark scores and such continue to go up when you push your RAM farther past your FSB? I run a 1:1, yet can easily get a faster Pi time or benchmark test if I push my RAM to 900mhz or even farther. If it didnt do ANY good, then it wouldnt get ANY better....but it DOES........soooooooooo........?
 

·
Premium Member
Joined
·
8,616 Posts
I'll tell you soon. My XMS2s were running at 667 eff. at 4-4-4-12. Supposedly these one I ordered run at 900 eff. 4-4-4-8! I'll let you guys know. You reckon with enough voltage I can get em down to CL3?
 

·
Premium Member
Joined
·
5,141 Posts
Here you go ...

Slower
CPU @ 4501MHz FSB 409.2 Ratio 1:1 Mem @ 409.2 C4-4-4-5

Faster
CPU @ 4476MHz FSB 447.6 Ratio 2:3 Mem @ 671.4 C4-4-4-5

Bandwidth plays a large part in performance be it AMD or Intel and while tighter timings may help there will be a cross-over point where C3 is out run by C4 on the Intel systems as is C2 is to C2.5 on the AMD... Memory Bandwidth.

The only way to find out what works best is to try it on each ones own particular rig as the components will interact differently even if identical rigs are setup.

Find the Max of the CPU, Mem and MB and then work the options.

Quote:


Originally Posted by Chozart

Interesting.. would like to see some screenies.

 

·
Banned
Joined
·
16,364 Posts
Soooo back to what I was saying.....I think even though ddr2 does 2 commands per clock, its probably not enough to feed the 12'odd clock/cycle cpu. As I said before, ....if running your RAM faster than your FSB, was pointless...then why does it produce better test scores and faster results?

Thumper->

Quote:


There is no reason to run your RAM faster than your FSB. So the answer is, the fastest is the tightest timings you can achieve at 1:1 ratio.

Running 1066 RAM on a 300FSB system with 6-6-6-15 timings will not be better than running 600 speed RAM with 4-4-4-12 timings because the system will not be able to use anything more than 600 of the 1066 anyway.

 

·
Registered
Joined
·
807 Posts
cl3p20, i think you may be confused. Running your RAM faster than your FSB does NOT increase performance. RAM recieves data as fast as the cpu feeds it so RAM running faster than FSB would not help. Can we see some screenshots of your ram running faster than your fsb and yeilding better benchmarks?
 

·
Banned
Joined
·
16,364 Posts
Yes I would be glad to. I will post them later 2nite. Look at the above super Pi times as poted by others. Its not hard to see. Leave your cpu sped alone and increase your RAM speed...then tell me if your Pi time does NOT improve. I will post as many screenies as I can, w/ different tests to show this effect. I encourage others to do the same as this is a highly debated topic. Thankx
 

·
Premium Member
Joined
·
5,809 Posts
Quote:


Originally Posted by ncsa

Here you go ...

Slower
CPU @ 4501MHz FSB 409.2 Ratio 1:1 Mem @ 409.2 C4-4-4-5

Faster
CPU @ 4476MHz FSB 447.6 Ratio 2:3 Mem @ 671.4 C4-4-4-5

Bandwidth plays a large part in performance be it AMD or Intel and while tighter timings may help there will be a cross-over point where C3 is out run by C4 on the Intel systems as is C2 is to C2.5 on the AMD... Memory Bandwidth.

The only way to find out what works best is to try it on each ones own particular rig as the components will interact differently even if identical rigs are setup.

Find the Max of the CPU, Mem and MB and then work the options.

LOL i see no difference. The times are almost the same, just about .1sec difference.which i think just alone the inconsistancy of windows will cause that...
 
1 - 20 of 23 Posts
Top