Overclock.net › Forums › Components › Memory › Samsung Miracle Memory Club
New Posts  All Forums:Forum Nav:

Samsung Miracle Memory Club - Page 24

post #231 of 1055
Quote:
Originally Posted by ivanlabrie View Post

4 sticks with your weak imc will hinder the oc...if I were you I'd settle with 1866mhz cl8-10-8-27-2t with 1.7v. That should easily work.

Ok. I thought SB's can easily do 2133MHz because of their robust IMCs?

Replacing these RAM kits with the Samsung OEM modules will not increase the overclocking capability of my system because I'm limited by my CPU's IMC, right?

While testing for RAM stability, is it fine to leave my CPU at Offset Mode set to Auto and at stock clocks? This is the optimized default settings.
post #232 of 1055
Quote:
Originally Posted by kevindd992002 View Post

Ok. I thought SB's can easily do 2133MHz because of their robust IMCs?

Replacing these RAM kits with the Samsung OEM modules will not increase the overclocking capability of my system because I'm limited by my CPU's IMC, right?

While testing for RAM stability, is it fine to leave my CPU at Offset Mode set to Auto and at stock clocks? This is the optimized default settings.

Yeah, defaultss are fine....compare bandwidth with maxxmem after 5 passes with IBT max memory and a fuull superpi32m run..
you won't see an improvement with Samsung cause they only shine with higher than 2133, they can't tighten timings like your ram.
post #233 of 1055
Quote:
Originally Posted by ivanlabrie View Post

Yeah, defaultss are fine....compare bandwidth with maxxmem after 5 passes with IBT max memory and a fuull superpi32m run..
you won't see an improvement with Samsung cause they only shine with higher than 2133, they can't tighten timings like your ram.

Ok. I left my CPU settings at default and I thought I was already stable. I tried reaching 2000% coverage successfully with 8-11-8-27-2t 2133 but failed at around 2100%. Is HCI Memtest really a reliable program for testing RAM stability? What are the RAM stability testing software famously used here at OCN? Superpi32m and IBT max memory only?
post #234 of 1055
Quote:
Originally Posted by kevindd992002 View Post

Ok. I left my CPU settings at default and I thought I was already stable. I tried reaching 2000% coverage successfully with 8-11-8-27-2t 2133 but failed at around 2100%. Is HCI Memtest really a reliable program for testing RAM stability? What are the RAM stability testing software famously used here at OCN? Superpi32m and IBT max memory only?

Those three are the best, that and prime95 blend with 90% ram but it takes much longer.
For quicker testing run Superpi 32m and then 5 passes of IBT with max memory, that should do. And again, you won't be able to get 2133mhz on that cpu, try 1866mhz with those timings, trust me it'll be as fast or faster than the Samsungs at 2133mhz.
post #235 of 1055
Quote:
Originally Posted by kevindd992002 View Post

Ok. I left my CPU settings at default and I thought I was already stable. I tried reaching 2000% coverage successfully with 8-11-8-27-2t 2133 but failed at around 2100%. Is HCI Memtest really a reliable program for testing RAM stability? What are the RAM stability testing software famously used here at OCN? Superpi32m and IBT max memory only?
For RAM testing I'd use a combination of HCI memtest and Prime95 custom blend with ~90% memory use. HCI's boot test is better than the windows based one, but isn't free (and slow as hell... takes around 23.5 hours to reach 500% coverage on my system).
ITB and Superpi/Hyperpi are good for fast, obvious fault finding, but aren't as good for true stability testing as the other two. When trying new settings, I usually do one Hyperpi run, 10 passes with max memory on ITB, and then I move on to HCI and P95.
Beestje 2.0
(20 items)
 
Beestje 1.5
(16 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 Sabertooth X79 ASUS GeForce GTX 680 ASUS GeForce GTX 680 
RAMHard DriveHard DriveHard Drive
Samsung low profile x4 OCZ Vertex 3 Crucial M4 Samsung HD103SJ 
Optical DriveCoolingCoolingOS
Lite-On BD drive Megahalems rev. B Cooler Master Excalibur Windows 7 home premium x64 
MonitorKeyboardPowerCase
ASUS VG278H Logitech G710+ Corsair TX750 Silverstone Raven 2 Evolution 
MouseMouse PadAudioAudio
Corsair M60 Razer Kabuto Echo Audiofire 2 Sennheiser HD515 
CPUMotherboardGraphicsRAM
i7 920 Asus P6T Club3d HD7870 royalKing Corsair Dominator 3x4GB 
Hard DriveHard DriveOptical DriveCooling
Crucial V4 128GB Seagate Barracuda LP 2TB LG GH22LS Scythe Mugen 2b 
CoolingCoolingMonitorKeyboard
CM Excalibur x2 CM R4 x3 Samsung BX2450 Microsoft Sidewinder X6 
PowerCaseMouseMouse Pad
Seasonic M12 II 520 Antec 600 Microsoft Intellimouse Razer Kabuto 
  hide details  
Reply
Beestje 2.0
(20 items)
 
Beestje 1.5
(16 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 Sabertooth X79 ASUS GeForce GTX 680 ASUS GeForce GTX 680 
RAMHard DriveHard DriveHard Drive
Samsung low profile x4 OCZ Vertex 3 Crucial M4 Samsung HD103SJ 
Optical DriveCoolingCoolingOS
Lite-On BD drive Megahalems rev. B Cooler Master Excalibur Windows 7 home premium x64 
MonitorKeyboardPowerCase
ASUS VG278H Logitech G710+ Corsair TX750 Silverstone Raven 2 Evolution 
MouseMouse PadAudioAudio
Corsair M60 Razer Kabuto Echo Audiofire 2 Sennheiser HD515 
CPUMotherboardGraphicsRAM
i7 920 Asus P6T Club3d HD7870 royalKing Corsair Dominator 3x4GB 
Hard DriveHard DriveOptical DriveCooling
Crucial V4 128GB Seagate Barracuda LP 2TB LG GH22LS Scythe Mugen 2b 
CoolingCoolingMonitorKeyboard
CM Excalibur x2 CM R4 x3 Samsung BX2450 Microsoft Sidewinder X6 
PowerCaseMouseMouse Pad
Seasonic M12 II 520 Antec 600 Microsoft Intellimouse Razer Kabuto 
  hide details  
Reply
post #236 of 1055
Quote:
Originally Posted by ivanlabrie View Post

Those three are the best, that and prime95 blend with 90% ram but it takes much longer.
For quicker testing run Superpi 32m and then 5 passes of IBT with max memory, that should do. And again, you won't be able to get 2133mhz on that cpu, try 1866mhz with those timings, trust me it'll be as fast or faster than the Samsungs at 2133mhz.

Ok. I'll try those. Do I need to run Superpi 32m for one full pass and then 5 passes of IBT with max memory?

Also, how long should I be running HCI Memtest for long stability testing? How many % coverage is the target?
Quote:
Originally Posted by Vonnis View Post

For RAM testing I'd use a combination of HCI memtest and Prime95 custom blend with ~90% memory use. HCI's boot test is better than the windows based one, but isn't free (and slow as hell... takes around 23.5 hours to reach 500% coverage on my system).
ITB and Superpi/Hyperpi are good for fast, obvious fault finding, but aren't as good for true stability testing as the other two. When trying new settings, I usually do one Hyperpi run, 10 passes with max memory on ITB, and then I move on to HCI and P95.

Why is the HCI boot test better than the windows based one? Simply because it can test all RAM? Yeah, I can imagine that it will take significantly a lot longer since the DOS-based program isn't optimized for multi-core CPUs.

What is the difference between Superpi and Hyperpi?

Same question as above, how long should I be running HCI Memtest for long stability testing? How many % coverage is the target?
Edited by kevindd992002 - 1/19/13 at 7:15pm
post #237 of 1055
Quote:
Originally Posted by kevindd992002 View Post

Ok. I'll try those. Do I need to run Superpi 32m for one full pass and then 5 passes of IBT with max memory?

Also, how long should I be running HCI Memtest for long stability testing? How many % coverage is the target?
Why is the HCI boot test better than the windows based one? Simply because it can test all RAM? Yeah, I can imagine that it will take significantly a lot longer since the DOS-based program isn't optimized for multi-core CPUs.

What is the difference between Superpi and Hyperpi?

Same question as above, how long should I be running HCI Memtest for long stability testing? How many % coverage is the target?

1000% in Windows should be fine, 5 passes of IBT, one superpi 32m and hyperpi for some extra reassurance. But that only if you need rock solid stability for something like Folding at home, otherwise IBT max mem 5 passes and superpi 32m are plenty. Couple that with 18hs of prime95 blend with 90% of your ram once you finish testing your cpu oc with stock ram.
post #238 of 1055
Quote:
Originally Posted by ivanlabrie View Post

1000% in Windows should be fine, 5 passes of IBT, one superpi 32m and hyperpi for some extra reassurance. But that only if you need rock solid stability for something like Folding at home, otherwise IBT max mem 5 passes and superpi 32m are plenty. Couple that with 18hs of prime95 blend with 90% of your ram once you finish testing your cpu oc with stock ram.

But I can do 1000% coverage with 2133MHz with no problems, does that mean I'm already stable? And yes I need rock solid stability.

For superpi 32m, do you use version 1.8WP? Select 32m and then just press calculate and wait for it to finish?

You mean when testing my CPU OC, I should first revert to stock ram settings? I thought it is better to OC the RAM first and then use those settings when OC'ing the CPU?
post #239 of 1055
Quote:
Originally Posted by kevindd992002 View Post

Why is the HCI boot test better than the windows based one? Simply because it can test all RAM? Yeah, I can imagine that it will take significantly a lot longer since the DOS-based program isn't optimized for multi-core CPUs.

What is the difference between Superpi and Hyperpi?

Same question as above, how long should I be running HCI Memtest for long stability testing? How many % coverage is the target?
The boot test is better because it seems to test in a different way. It found an error before even reaching 1% of the full memory pass, after the windows based test reached 1000% coverage with no issues on the exact same settings. Since it's such a long test, I generally go for just 500% coverage here. In my experience that's a lot more solid that 1000% coverage in windows.
Hyperpi is a multi-threaded front-end for Superpi. You specify how many threads you want to run, and it'll spawn that number of Superpi instances.

If you need rock-solid stability, you have to use several tests for extended periods of time. For a time I used HCI boot test only, because I could run Prime95 for days without errors yet fail within minutes in HCI. Tjj226 Angel had the exact opposite experience. I've gone back to using both because of that, just to be sure.

Oh, and when you run P95, make sure it finishes all FFT lengths. Passing all of those is more important than running it for 24 hours but only covering 75% of the FFT list in my experience.
Beestje 2.0
(20 items)
 
Beestje 1.5
(16 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 Sabertooth X79 ASUS GeForce GTX 680 ASUS GeForce GTX 680 
RAMHard DriveHard DriveHard Drive
Samsung low profile x4 OCZ Vertex 3 Crucial M4 Samsung HD103SJ 
Optical DriveCoolingCoolingOS
Lite-On BD drive Megahalems rev. B Cooler Master Excalibur Windows 7 home premium x64 
MonitorKeyboardPowerCase
ASUS VG278H Logitech G710+ Corsair TX750 Silverstone Raven 2 Evolution 
MouseMouse PadAudioAudio
Corsair M60 Razer Kabuto Echo Audiofire 2 Sennheiser HD515 
CPUMotherboardGraphicsRAM
i7 920 Asus P6T Club3d HD7870 royalKing Corsair Dominator 3x4GB 
Hard DriveHard DriveOptical DriveCooling
Crucial V4 128GB Seagate Barracuda LP 2TB LG GH22LS Scythe Mugen 2b 
CoolingCoolingMonitorKeyboard
CM Excalibur x2 CM R4 x3 Samsung BX2450 Microsoft Sidewinder X6 
PowerCaseMouseMouse Pad
Seasonic M12 II 520 Antec 600 Microsoft Intellimouse Razer Kabuto 
  hide details  
Reply
Beestje 2.0
(20 items)
 
Beestje 1.5
(16 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 Sabertooth X79 ASUS GeForce GTX 680 ASUS GeForce GTX 680 
RAMHard DriveHard DriveHard Drive
Samsung low profile x4 OCZ Vertex 3 Crucial M4 Samsung HD103SJ 
Optical DriveCoolingCoolingOS
Lite-On BD drive Megahalems rev. B Cooler Master Excalibur Windows 7 home premium x64 
MonitorKeyboardPowerCase
ASUS VG278H Logitech G710+ Corsair TX750 Silverstone Raven 2 Evolution 
MouseMouse PadAudioAudio
Corsair M60 Razer Kabuto Echo Audiofire 2 Sennheiser HD515 
CPUMotherboardGraphicsRAM
i7 920 Asus P6T Club3d HD7870 royalKing Corsair Dominator 3x4GB 
Hard DriveHard DriveOptical DriveCooling
Crucial V4 128GB Seagate Barracuda LP 2TB LG GH22LS Scythe Mugen 2b 
CoolingCoolingMonitorKeyboard
CM Excalibur x2 CM R4 x3 Samsung BX2450 Microsoft Sidewinder X6 
PowerCaseMouseMouse Pad
Seasonic M12 II 520 Antec 600 Microsoft Intellimouse Razer Kabuto 
  hide details  
Reply
post #240 of 1055
Quote:
Originally Posted by Vonnis View Post

The boot test is better because it seems to test in a different way. It found an error before even reaching 1% of the full memory pass, after the windows based test reached 1000% coverage with no issues on the exact same settings. Since it's such a long test, I generally go for just 500% coverage here. In my experience that's a lot more solid that 1000% coverage in windows.
Hyperpi is a multi-threaded front-end for Superpi. You specify how many threads you want to run, and it'll spawn that number of Superpi instances.

If you need rock-solid stability, you have to use several tests for extended periods of time. For a time I used HCI boot test only, because I could run Prime95 for days without errors yet fail within minutes in HCI. Tjj226 Angel had the exact opposite experience. I've gone back to using both because of that, just to be sure.

Oh, and when you run P95, make sure it finishes all FFT lengths. Passing all of those is more important than running it for 24 hours but only covering 75% of the FFT list in my experience.

I only test my folding oc profile that much, and yeah, I'd reccomend running all those tests for extra reassurance. Prime95 blend with 90% ram is quite thorough but not enough for cpu testing.
Lately I have grown fond of Cinebench for a quick determination of the chip's capabilities, then IBT max mem for 20 passes, Prime95 large fft's till it completes all lenghts used and Prime95 blend with 90% ram once I have determined my preferred ram oc for 24/7 use. But that's only a good idea if you wanna fold or a maniac of stability tongue.gif

Another good approach for not so exhaustive testing is running cinebench, checking whea errors in the Windows Event Logger and then running whatever application or games you plan to use for extended time periods. You'll then see if it crashes or throws some whea errors in the event logger while doing so and decide wether that oc is a keeper or not.
If it passed cinebench it will most likely work good for gaming and regular use or at least need a tiny vcore bump. thumb.gif (remember to test the ram oc separately before doing this, cause cinebench doesn't stress the ram much if at all)
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Memory
Overclock.net › Forums › Components › Memory › Samsung Miracle Memory Club