Overclock.net › Forums › Components › Memory › Correct tRC timings
New Posts  All Forums:Forum Nav:

Correct tRC timings - Page 2

post #11 of 18
Quote:
Originally Posted by damric View Post

Don't mind AMD4ME, he is the antithesis of OCN, the pursuit of performance.

Damric-

My purpose for being on OCN is to help people learn by presenting accurate technical information and dispelling mis-information. I am aware from your previous comments that you do not take kindly to technical realities that refute your beliefs. Making the conversation personal is totally inappropriate and counter-productive to the discussion.

Your very own Far Cry 2 test data confirmed exactly what Tom's, AnandTech and X-Bit Labs independently confirmed and stated - there is no significant gains in system performance from increased frequencies above ~1333 MHz. and even less from changes in latencies in REAL world applications. No where have I suggested that people not tune their PC as you imply.

What I have explained numerous times is that tweaking the DDR3 RAM settings does NOT provide the system gains that DDR and DDR2 RAM produced because DDR3 is not saturated 100% of the time due to it's greater bandwidth and as such is not a system bottleneck at or above ~1333 MHz. for typical desktop applications.

Explaining to people that DDR3 RAM @ ~1333 MHz. is not a system bootleneck helps them understand why they see no changes in real applications with DDR3 RAM changes above ~1333 MHz. Your 50% RAM overclock from 1333 MHz. to 2000 MHz. on a FX CPU made this reality perfectly clear and I thank you for posting your FC2 test data so people could see for themselves that there is no significant gains in real applications.

BTW, you should refrain from inappropriate posts such as your flame bait above, which is clearly NOT on topic nor constructive to understanding the realities of DDR3 RAM OC'ing. You've made numerous inappropriate comments directed towards me in several threads because you are unwilling to accept the realities associated with DDR3 RAM even from your own FC2 test data. Insulting people does not change the test data nor reality and is inappropriate forum behavior.
Edited by AMD4ME - 4/17/12 at 10:21am
post #12 of 18
Quote:
Originally Posted by AMD4ME View Post

Damric-
My purpose for being on OCN is to help people learn by presenting accurate technical information and dispelling mis-information. I am aware from your previous comments that you do not take kindly to technical realities that refute your beliefs. Making the conversation personal is totally inappropriate and counter-productive to the discussion.
Your very own Far Cry 2 test data confirmed exactly what Tom's, AnandTech and X-Bit Labs independently confirmed and stated - there is no significant gains in system performance from increased frequencies above ~1333 MHz. and even less from changes in latencies in REAL world applications. No where have I suggested that people not tune their PC as you imply.
What I have explained numerous times is that tweaking the DDR3 RAM settings does NOT provide the system gains that DDR and DDR2 RAM produced because DDR3 is not saturated 100% of the time due to it's greater bandwidth and as such is not a system bottleneck at or above ~1333 MHz. for typical desktop applications.
Explaining to people that DDR3 RAM @ ~1333 MHz. is not a system bootleneck helps them understand why they see no changes in real applications with DDR3 RAM changes above ~1333 MHz. Your 50% RAM overclock from 1333 MHz. to 2000 MHz. on a FX CPU made this reality perfectly clear and I thank you for posting your FC2 test data so people could see for themselves that there is no significant gains in real applications.
BTW, you should refrain from inappropriate posts such as your flame bait above, which is clearly NOT on topic nor constructive to understanding the realities of DDR3 RAM OC'ing. You've made numerous inappropriate comments directed towards me in several threads because you are unwilling to accept the realities associated with DDR3 RAM even from your own FC2 test data. Insulting people does not change the test data nor reality and is inappropriate forum behavior.

Personally i found that my system performed best when i had my 1600mhz rated ram running at 1200 with cas 5 timings.

But some people are all about the mhz! biggrin.gif

In the end, whatever makes each person is happy is fine
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
post #13 of 18
Quote:
Originally Posted by bringonblink View Post

Personally i found that my system performed best when i had my 1600mhz rated ram running at 1200 with cas 5 timings.
But some people are all about the mhz! biggrin.gif
In the end, whatever makes each person is happy is fine

Some folks also Under-clock the RAM frequency so they can OC the CPU higher - which results in better performance too. If 1200 MHz. and CAS 5 works best for you, then that is what you should use. Be advised however that synthetic benches do not always 100% represent real application performance.
post #14 of 18
Quote:
Originally Posted by AMD4ME View Post

Some folks also Under-clock the RAM frequency so they can OC the CPU higher - which results in better performance too. If 1200 MHz. and CAS 5 works best for you, then that is what you should use. Be advised however that synthetic benches do not always 100% represent real application performance.

Yeah i know that, these were game benches, but even though it was only a few FPS faster, i was more interested in the lower voltage needed for the ram thumb.gif
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
post #15 of 18
Quote:
Originally Posted by bringonblink View Post

Yeah i know that, these were game benches, but even though it was only a few FPS faster, i was more interested in the lower voltage needed for the ram thumb.gif

Lower voltage is cool. DDR4 RAM will be 1.2v by default. It's not a big change but if you're looking to lower power consumption, every little bit helps.
post #16 of 18
Stich to even number.2 4 6 etc
post #17 of 18
Quote:
Originally Posted by drbaltazar View Post

Stich to even number.2 4 6 etc

Eh? I've never heard of this. Any source?
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II X4 965 @ 4GHZ Asus Crosshair V Formula NB@ 2.8GHZ VTX 5850 @ 800/1000 Corsair Vengeance LP 8GB 1600 9 9 9 27 1T 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 256GB 000F WD Green 3TB WD30EZRX LG Prolimetech Megahalems + San Ace 
OSMonitorPowerCase
Windows 7 64bit Dell U2312HM Antec CP-850 Antec 1200 
MouseAudio
Razer Salmosa Asus Xonar Xense + PC350 
  hide details  
Reply
post #18 of 18
Ok, so here's what I've found with memory timings:

CL+tRCD+tRP = tRAS; then tRP + tRAS = tRC.

Please note that tRCD and tRP can be +/- up to 2. then simply follow the formula through. I have found this to be the most stable. I do fairly extensive testing, especially when it comes to memory, because of the potential of OS corruption.

Using this formula I was able to get my x3 720BE @ 3.6 to run my memory @ 1600, 7-7-7-21-28-1T timings with my NB @ 2650.
I was fairly impressed with the performance. The most noticeable gains were of course in synthetic benchmarks, however, video applications also benefited from it as well.

For instance, my 3Dmark06 score went up almost 1000 points by simply overclocking the memory from 1333 8-8-8-24-32-1T to 1600 7-7-7-21-28. Although, that was, in part, due to also overclocking the CPU/NB.

It just takes a lot of time to get everything, regarding the memory, just right.
It took me three days of solid tweaking and testing to get it there. However, I'm really particular about my memory overclocks.

My stress testing regimen at the time:
4 full passes of memtest.
24 hours of Pime95 blend or custom test to use the maximum amount of memory.
then, at least, one day of folding at home.

I've found folding at home to be very sensitive when it comes to any CPU or memory errors.

I, personally, haven't noticed any OS corruption due to overclocking the tRC. However, I always overclock the other timings right with it. I believe there needs to be a ratio between all of the available timings and you don't simply adjust one without adjusting the others.
Black Box
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 3770k Asus P8Z77-Vpro Evga 780 Classified Crucial Ballistix 
Hard DriveHard DriveOptical DriveCooling
A-Data 128GB SSD Samsung Spinpoint F3 1TB Pioneer BDR-207DBK Corsair H80 
OSMonitorKeyboardPower
Win 7 pro 64 Hannspree 25" 1080p LCD 2ms Razer Lycosa PC Power & Cooling 760w 
CaseMouseMouse PadAudio
Corsair 550D MX 518 Allsop Creative XFi Xtreme music 
  hide details  
Reply
Black Box
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 3770k Asus P8Z77-Vpro Evga 780 Classified Crucial Ballistix 
Hard DriveHard DriveOptical DriveCooling
A-Data 128GB SSD Samsung Spinpoint F3 1TB Pioneer BDR-207DBK Corsair H80 
OSMonitorKeyboardPower
Win 7 pro 64 Hannspree 25" 1080p LCD 2ms Razer Lycosa PC Power & Cooling 760w 
CaseMouseMouse PadAudio
Corsair 550D MX 518 Allsop Creative XFi Xtreme music 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Memory
Overclock.net › Forums › Components › Memory › Correct tRC timings