Overclock.net › Forums › Components › Memory › Overclocked ram & Why 1:1?
New Posts  All Forums:Forum Nav:

Overclocked ram & Why 1:1? - Page 7

post #61 of 88
Alright Choggs do me ONE favor and test your own hypothesis according to this.

7
Quote:
3224 MB/s ÷ 400MHz = 8.06 MB/s per 1MHz with 1:1 Ratio
3774 MB/s ÷ 600MHz = 6.29 MB/s per 1MHz with 2:3 Ratio
4047 MB/s ÷ 800MHz = 5.06 MB/s per 1MHz with 1:2 Ratio
6*500=3000 @ 1:1 DDR1000
8.06MB/s/MHz * 1000MHz = 8060MB/s

9*333=3000 @ 3:2 DDR999
6.29MB/s/MHz * 999 = 6284MB/s

You should get near those numbers if what you're saying is true. Or, at least, in a 1.28/1 ratio to one another. If it's GREATER than 1.28/1 better at 1:1 then you can blame it on the extra FSB and your theory MIGHT be/is likely true. If it is equal to 1.28/1 better on the 1:1 then you're likely right. If it is less then 1.28/1 better then you're 100% wrong as the extra FSB only serves to HELP the 1:1 look faster than the other and still falls short.

The ONLY way for you to REALLY test your idea is on an unlinked 680i chipset, and even then, that does away the very divider you're trying to test... which really means that there is NO WAY to test your hypothesis and you're now just philosophizing about ram speed, which is about as useless as a $3 bill.
Raptor Raper
(13 items)
 
  
CPUMotherboardGraphicsRAM
Something with transistors FOXCONN Boodrage -> eVGA classified e760 GTX 480 / HD6850@1090/1200 3x2 gig G Skill Trident DDR2000 
Hard DriveOptical DriveOSMonitor
2x128GB Crucial C300 Sumsung SATA DVD-RW 64 bit Windows 7 Samsung 245BW 
KeyboardPowerCaseMouse
Microsoft curve Tagan ITZ 800W DD Torture Rack Diamondback 
  hide details  
Reply
Raptor Raper
(13 items)
 
  
CPUMotherboardGraphicsRAM
Something with transistors FOXCONN Boodrage -> eVGA classified e760 GTX 480 / HD6850@1090/1200 3x2 gig G Skill Trident DDR2000 
Hard DriveOptical DriveOSMonitor
2x128GB Crucial C300 Sumsung SATA DVD-RW 64 bit Windows 7 Samsung 245BW 
KeyboardPowerCaseMouse
Microsoft curve Tagan ITZ 800W DD Torture Rack Diamondback 
  hide details  
Reply
post #62 of 88
Quote:
Originally Posted by Choggs396 View Post
Of course the FSB is going to be different, it doesn't matter what the FSB is, as long as the CPU Multiplier is set to make the ACTUAL CPU SPEED THE SAME.
you have officially stepped outside the realm of practlcal use. wouldn't the concern for most people be fast overall system performance? in the real world, the FSB is the workhorse of any PC. let's not get so caught up in this discussion of theoretical math that we forget why we overclock in the first place.

i think the fact that nobody can think up an experiment that would prove or disprove what Choggs is saying means this entire debate is completely irrelevant. just make everything go as fast as you can. period.
The C'roe
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6320 @ 2.8 Asus P5W DH Deluxe (2) Radeon X1900XTX CF 2GB G.Skill HK 4-4-3-5 
Hard DriveOptical DriveOSMonitor
640 - 2x320 Samsung RAID-0 Dual DVDs XP MCE Acer 22" + Dell 19" LCDs - 2960x1024 rez :D 
KeyboardPowerCaseMouse
Logitech MX Zalman 600W RT Modified Antec 900 Logitech MX 
Mouse Pad
Giant 
  hide details  
Reply
The C'roe
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6320 @ 2.8 Asus P5W DH Deluxe (2) Radeon X1900XTX CF 2GB G.Skill HK 4-4-3-5 
Hard DriveOptical DriveOSMonitor
640 - 2x320 Samsung RAID-0 Dual DVDs XP MCE Acer 22" + Dell 19" LCDs - 2960x1024 rez :D 
KeyboardPowerCaseMouse
Logitech MX Zalman 600W RT Modified Antec 900 Logitech MX 
Mouse Pad
Giant 
  hide details  
Reply
post #63 of 88
Yeah I see that Ihatethedukes has a point, I could run my ram at ddr2 1000 4-4-4-12 for you at both 1:1 or on a divider depending on what I did with the multi but there is no way for me to test the same ram speeds but with a different multi and still have the same ghz on the cpu. Otherwise I would run a bunch o fbenches for you.

I can say that running my rig at 4.0ghz (445x9) with ram at 1:1 (445mhz)4-4-4-12 timings is not as effective as running a 4:5 divider (556mhz)with 5-5-5-13 timings. I get better benchmark scores including 3dmark scores ecspecially the 01/03, aquamark,super pi, all everst benches that relate to ram and sisoft.

I think this is a little off the point of what you are trying to prove though and relates back more to what pauldovi was trying to say to begin with.


Now this wasnt true with AMD processers like socket 939. With those guys if I ran my ram at basically the same as fsb (which is a little different with ddr I guess) with the tightest timings then it performed as good and usually better then running my ram at faster speeds then the fsb. But that is a different situation.
xeon
(13 items)
 
  
CPUMotherboardGraphicsRAM
W3520 x58 ud5 9800gtx+ tridents 
Hard DriveOSMonitorPower
wd black x2 W7 64 acer 23 ocz 850hx 
Case
haf932 
  hide details  
Reply
xeon
(13 items)
 
  
CPUMotherboardGraphicsRAM
W3520 x58 ud5 9800gtx+ tridents 
Hard DriveOSMonitorPower
wd black x2 W7 64 acer 23 ocz 850hx 
Case
haf932 
  hide details  
Reply
post #64 of 88
Quote:
Originally Posted by MrBungle View Post
just make everything go as fast as you can. period.
wow, i must be brilliant. can someone please sticky that and call it "Mr. Bungle's Overclocking Guide?"
The C'roe
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6320 @ 2.8 Asus P5W DH Deluxe (2) Radeon X1900XTX CF 2GB G.Skill HK 4-4-3-5 
Hard DriveOptical DriveOSMonitor
640 - 2x320 Samsung RAID-0 Dual DVDs XP MCE Acer 22" + Dell 19" LCDs - 2960x1024 rez :D 
KeyboardPowerCaseMouse
Logitech MX Zalman 600W RT Modified Antec 900 Logitech MX 
Mouse Pad
Giant 
  hide details  
Reply
The C'roe
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6320 @ 2.8 Asus P5W DH Deluxe (2) Radeon X1900XTX CF 2GB G.Skill HK 4-4-3-5 
Hard DriveOptical DriveOSMonitor
640 - 2x320 Samsung RAID-0 Dual DVDs XP MCE Acer 22" + Dell 19" LCDs - 2960x1024 rez :D 
KeyboardPowerCaseMouse
Logitech MX Zalman 600W RT Modified Antec 900 Logitech MX 
Mouse Pad
Giant 
  hide details  
Reply
post #65 of 88
Quote:
Originally Posted by Nerogk View Post
Okay I read the guide but still am unsure about how I was able to overclock the memory that well. After reading the guide, I can use better terminology:

This is DDR2-800 memory that I am running at DDR2-1000 by relaxing timings from 4-5-4-11 to 5-5-5-12 and leaving the voltage at 2.1v. Also, the divider is obviously not 1:1, everyone says 1:1 is best..but all that would do would be lower my memory bandwith....why would I want to do that...
It's only good if you're running a low multi with a high FSB.
AMD :D
(13 items)
 
  
CPUMotherboardGraphicsRAM
AMD 1055T Asus M4A89TD PRO/USB3 XFX 6950 Flashed 6970 BIOS Patriot Extreme Performance Sector 5 G Series 8GB 
Hard DriveCoolingOSMonitor
Patriot Pyro 60gb, OCZ Agility 3, 2x 7200.11 50... CoolerMaster Hyper 212 Windows 7 Ultimate 64bit Acer 230HL 
KeyboardPowerCaseMouse
Razer BlackWidow Delta 750w Antec 600 v2 Logitech G400 
Mouse Pad
SteelSeries 4HD 
  hide details  
Reply
AMD :D
(13 items)
 
  
CPUMotherboardGraphicsRAM
AMD 1055T Asus M4A89TD PRO/USB3 XFX 6950 Flashed 6970 BIOS Patriot Extreme Performance Sector 5 G Series 8GB 
Hard DriveCoolingOSMonitor
Patriot Pyro 60gb, OCZ Agility 3, 2x 7200.11 50... CoolerMaster Hyper 212 Windows 7 Ultimate 64bit Acer 230HL 
KeyboardPowerCaseMouse
Razer BlackWidow Delta 750w Antec 600 v2 Logitech G400 
Mouse Pad
SteelSeries 4HD 
  hide details  
Reply
post #66 of 88
Jeez, i go out for a few drinks and the place goes nuts. In my experience so far there is no difference between my 1/1.66 divider and my 1/1.33 divider as far as mb/sec goes. I dont know if that has anything to do with the fact that i cant get my processor to clock any higher or not. As i cannot adjust my cpu multiplier i am kinda limited in the experimentation that i can do on my particular setup.
The Beast Deux
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-5820k 4.4Ghz GA-X99-Gaming 5 Zotac AMP! GTX-470 G.Skill DDR4-2133 4X4Gb 
Hard DriveOptical DriveCoolingOS
Samsung 950 Pro M.2 256Gb Sony Optiarc DVD-R+ Corsair H-100i Windows 10 Pro 
MonitorKeyboardPowerCase
View Sonic 22.5" LED Microsoft Wireless Desktop 2000 Antec True Power New 750 CoolerMaster HAF-932 
MouseAudio
Microsoft Wireless Desktop 2000 Creative Sound Core 3D 
  hide details  
Reply
The Beast Deux
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-5820k 4.4Ghz GA-X99-Gaming 5 Zotac AMP! GTX-470 G.Skill DDR4-2133 4X4Gb 
Hard DriveOptical DriveCoolingOS
Samsung 950 Pro M.2 256Gb Sony Optiarc DVD-R+ Corsair H-100i Windows 10 Pro 
MonitorKeyboardPowerCase
View Sonic 22.5" LED Microsoft Wireless Desktop 2000 Antec True Power New 750 CoolerMaster HAF-932 
MouseAudio
Microsoft Wireless Desktop 2000 Creative Sound Core 3D 
  hide details  
Reply
post #67 of 88
Quote:
Originally Posted by Choggs396 View Post
Thanks for the information. But those numbers actually prove my point. Why?

I said:


Now, you tested those bandwidths at different speeds. But if you do the math, the MB/s compared to every single MHz shows I am right. You have to run a control (the MHz) - to Bandwidth to show fair comparisons, like this... My proof:

3224 MB/s ÷ 400MHz = 8.06 MB/s per 1MHz with 1:1 Ratio
3774 MB/s ÷ 600MHz = 6.29 MB/s per 1MHz with 2:3 Ratio
4047 MB/s ÷ 800MHz = 5.06 MB/s per 1MHz with 1:2 Ratio

So if you can read the math correctly, then the bold clearly shows a decrease in real bandwidth at the same speed using un-synced ratios.

what does mb/s per mhz have to do with speed? that is more of efficiency. that's like sayin a 4mhz chip that does 20ops per hz is better then one that runs at 300hz cause it only does 10 per second. I still choose the later there cause overall speed is better. and with what you have shown here in this post and said in the last post you would rather have less performance but have more bandwidth per mhz... even though it's slower. paul if at 800mhz your only pulling 4047mb/s something's not right as my centon 533 overclocked to 667 did better then that.

and eitherway both test prove paul's point that running your memory on a divider to achieve the fastest memory speed is best. you have shown that he has shown that. why slow the memory down just so you can have the 1:1. and for some like me with my p5b I'm fsb limited NOT ram limited.
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
post #68 of 88
[QUOTE=Choggs396;2348790]
Quote:
You contradicted yourself because that would mean there are TWO VARIABLES. One being the MEMORY SPEED and one being the RATIO. From the quote above I WAS TRYING TO DISCERN HOW THE RATIO, NOT MEMORY SPEED IN MEGAHERTZ, makes performance better.
the ratio by itself won't make the ram faster. but the effect of the ratio making the ram run faster will. you get more bandwidth doing so and if timings are kept at the same number they are then tighter at the higher speed, copy speed is quicker, anything related to the ram at that point is quicker with the higher mhz.
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
post #69 of 88
Quote:
Of course the FSB is going to be different, it doesn't matter what the FSB is, as long as the CPU Multiplier is set to make the ACTUAL CPU SPEED THE SAME.

very much so does make a difference as to FSB bus speed as the fsb has to carry all that information through to the cpu. that's like saying it doesn't matter how fast the cpu is I have teh ultra leet 8800 ultra for my p3
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
compy jr
(13 items)
 
  
CPUMotherboardGraphicsHard Drive
xeon 3350 at 3.6ghz Asus P5k-dlx HIS 7970 @ 1150/1900 Lots of them.  
OSMonitorKeyboardPower
vista ultimate x64/xp sp3 compaq 21" junk dell keyboard corsair tx850m 
CaseMouseMouse Pad
upgraded metal box You don't need a mouse Desk 
  hide details  
Reply
post #70 of 88
Haven't gone through 8 pages but 1:1 is great but if you can only get a nice OC on say 4:5 it's fine.

What I would do is try to get DDR2 1066 or such and run the FSB @ 533 with a multi of 6 and get 3.1Ghz OC..

DDR2 800
Run FSB at 400 with a 8 multi and get 3.2Ghz clock 1:1..

I don't know the exact multi of the e6600 but I think it is 8 or 7..
Workoholic REborn
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 2600K P8Z68-V PRO NVIDIA GeForce GTX 1080 Ti 12GB 1x4GB+1x8GB 
Hard DriveHard DriveOptical DriveCooling
Intel 520 Series WD Black ASUS DVD+RW Sunbeam Twister 120 
OSMonitorMonitorKeyboard
Windows 7 Ultimate LG OLED65C7P Oculus Rift Logitech K400 
PowerCaseMouse
Corsair 620 Modular Lian Li V1020B G9x 
  hide details  
Reply
Workoholic REborn
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 2600K P8Z68-V PRO NVIDIA GeForce GTX 1080 Ti 12GB 1x4GB+1x8GB 
Hard DriveHard DriveOptical DriveCooling
Intel 520 Series WD Black ASUS DVD+RW Sunbeam Twister 120 
OSMonitorMonitorKeyboard
Windows 7 Ultimate LG OLED65C7P Oculus Rift Logitech K400 
PowerCaseMouse
Corsair 620 Modular Lian Li V1020B G9x 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Memory
Overclock.net › Forums › Components › Memory › Overclocked ram & Why 1:1?