Go Back   Hardware Canucks > HARDWARE > RAM

    
Reply
 
LinkBack Thread Tools Display Modes
  #1 (permalink)  
Old September 15, 2008, 08:35 PM
Rookie
 
Join Date: Jul 2008
Posts: 34
Default Ram settings

I have been reading about this for a while and am a bit confused by it. I have an E8400 processor, P5QL-PRO motherboard and G.Skill F2-6400Cl4D 2X2Gb. The ram is 800mhz and it is running at 400mhz bus speed. I dont intend on OC my E8400 cuz I have no need to and it runs at 333mhz bus speed giving me a 5:6 fsb to ram ratio. Is this ok? I have heard a 1:1 ratio is best, and since I dont intend on OC my processor I would have to underclock my ram to 667mhz(or can I keep it at 800?) and run it at 333 bus speed to match the processor bus speed right? I could then theoretically run the memory with tighter timings too I assume. So in the end would this increase speed and performance in any way or is my current settings just fine?
Thanks for the advice, I am a bit confused on this issue.
Reply With Quote
  #2 (permalink)  
Old September 15, 2008, 09:53 PM
SugarJ's Avatar
Moderator
F@H
 
Join Date: Jan 2008
Location: Langley, BC
Posts: 6,141

My System Specs

Default

Use Everest, there's a free edition. Benchmark your memory as is. Make your changes, and benchmark again. You'll find out which is faster. I've run my mem both ways, and haven't been able to notice a difference in daily use.

From what I understand, the idea behind the 1:1 ratio is that the memory is synchronized with the FSB, and there's never a lost cycle. If you run 5:6, your memory may technically be running at a higher rate, but losing efficiency by waiting a cycle every once in a while. If you can run at DDR2-667 at say 4-4-4-10 vs. DDR2-800 5-5-5-15, it may actually be quicker in real life applications. Will your memory get that low? I don't know.
Reply With Quote
  #3 (permalink)  
Old September 15, 2008, 11:10 PM
MVP
 
Join Date: Dec 2007
Location: Edmonton
Posts: 268
Default

You've heard wrong.

1:1 is pretty much the slowest possible ratio other than on nForce chipsets, where 1T can be put to use.

I'd suggest benching for yourself; you'll discover the myth of 1:1 greatness is just that.
__________________
Gigabyte X58A-UD3R | i7 930 | Corsair H50
G.Skill RipJaws 4x2 GB DDR3 | HIS Radeon HD 5870
3x Intel X25-M 80 GB RAID-0; OCZ Agility 120 GB | Samsung SH-S243D
Corsair HX1000 | Dell 3007WFP & Samsung 204T | 7 Ultimate x64
Reply With Quote
  #4 (permalink)  
Old September 16, 2008, 05:02 PM
Hall Of Fame
 
Join Date: Jul 2008
Location: Canada
Posts: 1,221

My System Specs

Default

Also this whole debate about RAM timing is getting irrelevant as processors and gfx cards get faster - There is only so much frame rate our eyes can see - Most people out there will not notice a difference visually between running 4-4-4-12 and 5-5-5-18 or even 667 vs. 800 - people are running multi cores and top notch graphics cards and the frame rate to begin with is already high - so unless you are a hardcore gamer and want to squeeze every little frame rate possible, even a 5-10FPS increase, then you are better off buying good ram with low timings - but for most people, unless you benchmark you won't notice, and the better your initial frame rates and gfx/cpu, the less you will notice really - I've done the test myself with recent games - whether I had mine at 667 5-5-5-12 or 800 4-4-4-12, in both cases they were smooth as silk, and extremely playable, because my frame rate was high to begin with - of course if you have lower end gear, you might want that extra 5~15fps more :D So my advice to you - try it for yourself and benchmark it for yourself - results are subjective and everyone can have a different opinion on this!
Reply With Quote
  #5 (permalink)  
Old September 16, 2008, 05:14 PM
SugarJ's Avatar
Moderator
F@H
 
Join Date: Jan 2008
Location: Langley, BC
Posts: 6,141

My System Specs

Default

Quote:
Originally Posted by SugarJ View Post
I've run my mem both ways, and haven't been able to notice a difference in daily use.
Quote:
Originally Posted by DarKStar View Post
- but for most people, unless you benchmark you won't notice, and the better your initial frame rates and gfx/cpu, the less you will notice really - So my advice to you - try it for yourself and benchmark it for yourself - results are subjective and everyone can have a different opinion on this!
Isn't that what I said? Maybe a little less eloquently on my part?
Reply With Quote
  #6 (permalink)  
Old September 16, 2008, 06:36 PM
Hall Of Fame
 
Join Date: Jul 2008
Location: Canada
Posts: 1,221

My System Specs

Default

lol I was going through the replies very quickly unfortunately didn't have time to read all replies at the time I had to go :)
Reply With Quote
Reply


Thread Tools
Display Modes