Overclock.net › Forums › Components › Memory › 3 sticks from dual channel kit on triple channel slots?
New Posts  All Forums:Forum Nav:

3 sticks from dual channel kit on triple channel slots?

post #1 of 11
Thread Starter 
Hey, I've been using this for many years now and never ran into any issues until I tried overclocking.

When I built this pc in 2010, I had this 2x2gb kit (g.skill ripjaws, 1600mhz, 9-9-9-24 1.5v) and bought a triple channel motherboard, so I looked up online for another stick to complete the triple channel. The stick was identical to the other two, even came from another 2x2gb kit, and I didn't know much back then (not really sure what's the deal even now) so I went for it.

Few months later I bought another kit, the same g.skill ripjaws, same specs, except this time a proper 3x4gb kit, so since then I've been using these 18gb in 3x4gb(1kit 3sticks) + 3x2gb(1kit 2sticks + half kit of 2sticks) configuration.

As I mentioned, I had never ran into any issues (until now maybe) and always overlooked cpu-z, but now I noticed that it's reporting my 2gb dimms as 1333mhz(or 1066, not sure) when they're actually 1600mhz. (The 4gb dimms are showing up correctly as 1600mhz).

So what's going on? I never really believed ram kits were actually paired or connected in any form, and never actually found consistent information when I bought them (I did look up, perhaps not hard enough) so does anybody know what exactly is happening here?
post #2 of 11
There are over 20 different timing setting for memory. Manufacturers only give you four of them, sometimes more, but never all of them.

Manufactuirers frequently change the ICs (the black squares) they use. The specs (4 or more) may be the same, but then again some of the "unpublished" specs may be different. CPU-Z doesn't tell the whole story.. You can use RamMon to read most of those specs for your memory sticks.

Usually you can do one or both of the following to fix issues:
Manually set the memory speed to that of the slowest stick.
Manually set all the timings to the timings of the slowest stick.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
post #3 of 11
Thread Starter 
okay, thanks for the clarification.

I had two different understandings of ram:

1-ram shows on cpu-z as half speed (my 1600mhz ram sticks show as 800mhz) because it's called DDR (double data rate) and thus each stick operates at 800mhz doubled.

2-ram shows as half speed because it's split between the two sticks of a pair (wich never made much sense for me because I've been on triple channel for ages)


wich one is correct? the first seems really more plausible, but the second one would explain why I'd get errors while overclocking and wrong speed readings on cpu-z when I'm using a dual channel kit plus a single stick in a triple channel. at least seemed more plausible than a difference in the 20th hidden timing setting or a slight binning difference. but for the second to make sense, 3 sticks kits should show as 533,3mhz if it were divided by 3, so the first one is still more likely to be correct. wich is confusing for me, and I never found a definitive answer.
post #4 of 11
Number one is correct.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
post #5 of 11
Thread Starter 
Quote:
Originally Posted by billbartuska View Post

Number one is correct.

great, thanks.

so, in fact, it's not recommended, but I'm not very likely to have errors due to having 2 sticks from a kit and a single stick from an identical kit, am I correct? or should I avoid mixing at all costs?
post #6 of 11
Quote:
Originally Posted by fael097 View Post

Quote:
Originally Posted by billbartuska View Post

Number one is correct.

great, thanks.

so, in fact, it's not recommended, but I'm not very likely to have errors due to having 2 sticks from a kit and a single stick from an identical kit, am I correct? or should I avoid mixing at all costs?
Mixing memory is a crap shoot, sometimes it works, sometimes no. If no, sometimes you can tweak the memory in your bios and get it to all work together, but speeds and timings will be those of the slowest stick. There are memory timings that the memory manufacturers don't publish, so even buying the exact same mamory from the same manufacturer won't guarantee a match as memory manufacturers frequently change the ICs (the small black squares) on their memory sticks. The "published" specs may be the same, but all the specs may not.

You can use RamMon to look at "unpublished" memory timings on your sticks.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
post #7 of 11
Thread Starter 
right, I downloaded rammon, just didn't test yet cause I took off the mixed sticks to tweak my overclock, but I'll test my ram soon enough.

thanks again
post #8 of 11
Thread Starter 
Ok so I just got my ram testes and all the timings and specs that RamMon could possibly list are identical on the 3 sticks, even though one of them was bought separately. Guess that's a good thing huh?

So for this question let's just consider these 3 sticks were bought as one 3x2gb kit:

Having one 3x4gb (actual)kit and one 3x2gb kit, I should be fine as long as I set everything according to the lowest timings and speeds, correct?
post #9 of 11
Do th3e sticks in the 3 x kit match the sticks in the 2 x set?

But, heck! Slap them is and see what happens at defaults.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
post #10 of 11
As long as the memory can do the same speed and timings, and doesn't need a difference in voltage to get there, mixing is fine.
Computer
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1400 3.8Ghz Gigabyte AB350M-HD3 MSI Geforce GTX 1060 3GB Gaming X G.Skill 16GB(2x8GB) 3200Mhz DDR4 
Hard DriveOSMonitorPower
SSDs and HDDs Windows 10 x64 Pro Acer G7 G227HQLbi @75Hz EVGA 430w 
Case
Deepcool Frame 
  hide details  
Reply
Computer
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1400 3.8Ghz Gigabyte AB350M-HD3 MSI Geforce GTX 1060 3GB Gaming X G.Skill 16GB(2x8GB) 3200Mhz DDR4 
Hard DriveOSMonitorPower
SSDs and HDDs Windows 10 x64 Pro Acer G7 G227HQLbi @75Hz EVGA 430w 
Case
Deepcool Frame 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Memory
Overclock.net › Forums › Components › Memory › 3 sticks from dual channel kit on triple channel slots?