Overclock.net › Forums › Intel › Intel CPUs › Friend argues Prime 95 is useless. I need others opinions to ease it.
New Posts  All Forums:Forum Nav:

Friend argues Prime 95 is useless. I need others opinions to ease it. - Page 4

post #31 of 43
Quote:
Originally Posted by ZealotKi11er View Post

Keep in mind buy using better cooling then stock you are increasing stability.

Not to mention the fact that not all chips are created equal, and Intel tend to leave quite a bit of head room.
Serigatum
(13 items)
 
  
CPUMotherboardGraphicsRAM
4770k @ 4.7Ghz Asus P87 Deluxe EVGA GTX 780 Corsair Vengeance 2400Mhz 
Hard DriveHard DriveCoolingOS
Raid0: 2x512GB Samsung 840 Pro 2GB WD Black NZXT Kraken x60 Windows 8 
MonitorKeyboardPowerCase
Dell U2412M Razer Blackwidow Ultimate Corsair AX850 NZXT Phantom 820 
Mouse
Razer Ouroboros 
  hide details  
Reply
Serigatum
(13 items)
 
  
CPUMotherboardGraphicsRAM
4770k @ 4.7Ghz Asus P87 Deluxe EVGA GTX 780 Corsair Vengeance 2400Mhz 
Hard DriveHard DriveCoolingOS
Raid0: 2x512GB Samsung 840 Pro 2GB WD Black NZXT Kraken x60 Windows 8 
MonitorKeyboardPowerCase
Dell U2412M Razer Blackwidow Ultimate Corsair AX850 NZXT Phantom 820 
Mouse
Razer Ouroboros 
  hide details  
Reply
post #32 of 43
Quote:
Originally Posted by NihilOC View Post

So by your decree all people should either be happy with stock speeds, or they shouldn't care about instability? rolleyes.gif

Bit silly when you can get perfectly good stable overclocks.
I'd rather have better performance even if it crashed once a decade. Caring about hours and hours of p95 stability is about like keeping a car at WOT on a dyno for hours on end. Gamers don't need it, pro benchmarkers never have it, HPC people don't want to risk their applications.

Mind you, I'm not talking about just setting a OC and being done with it, stability testing is important. But if you fail Prime in the 30th hour, I don't see that as cause to drop down by 100 MHz.
post #33 of 43
I didn't bother responding yesterday, but coincidentally right after you had posted this NCIX put up a video showing what they do to test their computers before sending them out to the customers, and they run prime95 & furmark at the same time for awhile before shipping them out thumb.gif



personally I have never run p95 for more than 30 minutes, but I do always run it for 30 mins.
 
Philip J. Fry
(6 items)
 
Turanga Leela
(1 item)
 
CPUMotherboardGraphicsRAM
i5 4690K 4.2ghz 1.15v Asus Z97I-Plus EVGA GTX1070 ACX 3.0 HyperX 16gb 1866mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 Evo 250GB Western Digital Caviar Black1TB NZXT Kraken X41 Windows 7 Ultimate 
MonitorMonitorMonitorPower
ASUS VN247H-P ASUS VN247H-P ASUS VN247H-P SPI Magna Gold 750w 
CaseMouseMouse PadAudio
Corsair Obsidian 250D Zowie EC1-A Steelseries QCK XXL Fostex TH-X00 Mahogany 
AudioAudioAudio
Fiio E10 Olympus Modmic 4.0 Beyerdynamic DT770 80ohm Removable Cable Mod 
CPUGraphicsRAMHard Drive
Intel Core i5 M 480 Nvidia GeForce 310M 4gb Samsung ram 640gb Samsung HM641JI 
OSOther
Windows 7 Home Premium Samsung QX410 Laptop 
Keyboard
Which one? :P 
  hide details  
Reply
 
Philip J. Fry
(6 items)
 
Turanga Leela
(1 item)
 
CPUMotherboardGraphicsRAM
i5 4690K 4.2ghz 1.15v Asus Z97I-Plus EVGA GTX1070 ACX 3.0 HyperX 16gb 1866mhz 
Hard DriveHard DriveCoolingOS
Samsung 850 Evo 250GB Western Digital Caviar Black1TB NZXT Kraken X41 Windows 7 Ultimate 
MonitorMonitorMonitorPower
ASUS VN247H-P ASUS VN247H-P ASUS VN247H-P SPI Magna Gold 750w 
CaseMouseMouse PadAudio
Corsair Obsidian 250D Zowie EC1-A Steelseries QCK XXL Fostex TH-X00 Mahogany 
AudioAudioAudio
Fiio E10 Olympus Modmic 4.0 Beyerdynamic DT770 80ohm Removable Cable Mod 
CPUGraphicsRAMHard Drive
Intel Core i5 M 480 Nvidia GeForce 310M 4gb Samsung ram 640gb Samsung HM641JI 
OSOther
Windows 7 Home Premium Samsung QX410 Laptop 
Keyboard
Which one? :P 
  hide details  
Reply
post #34 of 43
Yep! I only run it for about 30 - 40 minutes max. I just need to know my temps and running it any longer is not going to make any difference unless room temp changes.

And for the normal user 24 hr testing in my eyes is a waste of time. Come on, when are we really likely to use 100% cpu all the time?

And then there are those that read about this extensive testing and think its the norm to do so, sheep follow sheep. They may feel insecure when it does pass but later on computer fails to boot up or bsod and other things that go bump in the night. Guess what? They run prime again biggrin.gif

Seen it , heard it , with mates and so on. There are a number of reason why our systems go haywire and prime95 is not going to give you the answers - prefer Aida64 thumb.gif
post #35 of 43
Quote:
Originally Posted by ladcrooks View Post

Yep! I only run it for about 30 - 40 minutes max. I just need to know my temps and running it any longer is not going to make any difference unless room temp changes.

And for the normal user 24 hr testing in my eyes is a waste of time. Come on, when are we really likely to use 100% cpu all the time?

And then there are those that read about this extensive testing and think its the norm to do so, sheep follow sheep. They may feel insecure when it does pass but later on computer fails to boot up or bsod and other things that go bump in the night. Guess what? They run prime again biggrin.gif

Seen it , heard it , with mates and so on. There are a number of reason why our systems go haywire and prime95 is not going to give you the answers - prefer Aida64 thumb.gif
Don't forget that overclocks can fail at idle as well, due to LLC/offset settings. And yet nobody ever does a 24h idle test.
post #36 of 43
It sounds like your buddy is just lazy, OP. Synthetic tests will find errors before "real world" (whatever that means...like what, watching porn?) tests will, because they stress the system more than you, as a user, really ever do. Sure, you'll find oddities anywhere and everywhere, but the bottom line is that if you can pass 10 runs of IBT and 3 hours of small FFT and blend, you're probably ok. An unstable system usually crashes out in 5 minutes from those tests.

I happen to prefer IBT, MSI Kombustor, Prime95 blend, and cinebench as my OC testing gauntlet. Everyone has their own style.
Edited by Axon14 - 4/3/13 at 6:42am
    
CPUMotherboardGraphicsRAM
i7 6700k ASUS z170-e EVGA 1080 SC ACX 3.0 G.Skill 32 GB  
Hard DriveOSMonitorPower
Samsung 950 m2.a SSD  Windows 10 Acer Predator xb271hu Corsair 850ax 
Case
Phanteks Eclipse P400w 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 6700k ASUS z170-e EVGA 1080 SC ACX 3.0 G.Skill 32 GB  
Hard DriveOSMonitorPower
Samsung 950 m2.a SSD  Windows 10 Acer Predator xb271hu Corsair 850ax 
Case
Phanteks Eclipse P400w 
  hide details  
Reply
post #37 of 43
Quote:
Originally Posted by almighty15 View Post

That maybe so but for me personally I've always used 50 runs of IBT on max RAM and I've never had a crash or BSOD...... Would my over clock pass Prime95? Who knows but I'm going to waste 24hrs trying to find out smile.gif

If your method is stable enough for you, then it's a good enough method for you.

I'm just saying that I have personally encountered situations where even extremely long IBT runs were not enough to isolate stability issues, even in some of my real world programs, and that some errors are found faster with some tests than others. Nothing is one-size-fits all.
Quote:
Originally Posted by almighty15 View Post

I remember watching a video on youtube from inside Intel and even they said there would be certain codes and situation where stock CPU's would crash, they just try to reduce them as much as possible.

This is very true.

Both AMD and Intel (as well as others) release whole volumes of documented errata, detailing conditions that will cause their parts to behave incorrectly. There are always other errors they don't catch as well, and beyond this, Intel's validation doesn't always catch more random and mundane instabilities (though this is rare, as they leave enormous headroom).
Quote:
Originally Posted by Dyson Poindexter View Post

Gamers don't need it, pro benchmarkers never have it, HPC people don't want to risk their applications.

This is not necessarily true, and there are many users that do not primarily fit into any of these categories.
Quote:
Originally Posted by Dyson Poindexter View Post

And yet nobody ever does a 24h idle test.

Long idle tests (often well in excess of 24 hours), resuming from S3, and cold boot tests are part of my standard test procedure.

So, wrong again.
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
post #38 of 43
Here is my stressing system input.

I do 1 hour of Prime95 custom 1792FFTs min/max with 14400mb of my 16gb used and then I run HyperPI 32m which takes about 12 minutes. Good enough for me. As others have said, you could run 24 hours of prime95 and 100 runs of IBT, but still get bsods.
    
CPUMotherboardGraphicsRAM
8700k @5.1GHz Apex X TITAN X (Pascal)  Trident Z 4400MHz 
Hard DriveCoolingCoolingOS
750 NVMe 1.2TB EK HF/rx360/rx120 EK FC/GTs/mcp655 Windows 10 Pro  
MonitorPowerCaseAudio
BL3200PT 1600 T2 800d Pro Audio 2.1 + DragonFly Red 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
8700k @5.1GHz Apex X TITAN X (Pascal)  Trident Z 4400MHz 
Hard DriveCoolingCoolingOS
750 NVMe 1.2TB EK HF/rx360/rx120 EK FC/GTs/mcp655 Windows 10 Pro  
MonitorPowerCaseAudio
BL3200PT 1600 T2 800d Pro Audio 2.1 + DragonFly Red 
  hide details  
Reply
post #39 of 43
Had a browse in Google on prime 95 and came across this from another site -

Stress tests are good to test your cooling and not much else. Sad to say this, but all they prove is that your system can pass that particular stress test for that period of time. No certainty that the next time your run the test it will pass. And we have heaps of threads of people crying my system passed 12 hours of Prime and then failed in WOW after 10 minutes. I kid you not.

If your system does the things you need it to do and does not fall over, fail or BSOD. Your system is stable.

One of the most demanding things your system will do is resume from sleep. Open heaps of applications that are memory heavy, have about 30 tabs open in your browser and put your system to sleep. If after a couple of hours it wakes no problem and all is good then mate you have a stable system.

Respect goes to those folks who are able to run Prime or any of similar stress tests for days on end at the lowest voltages possible. That is a skill all on its own an done I do respect. That is their thing, just like benching, gaming or what ever is other folks thing.

I totally agree. I have never used Prime, Linx etc to test the stability of an OC. I've lost count of the number of times I've heard people say they ran Prime for 12+ hours only to have the system crash after gaming for a few minutes.

After I set up an overclock, I run Aida64 for an hour. It's the closest stress test to real world usage that I've come across and I've been doing this for a lot of years (way back when it was called Everest) If my system can handle an hour of Aida64 it can handle everything I do with no WHEA logger warnings (which too many people ignore or don't know about)



What I have agreed with all along! biggrin.gif You may differ, that's up to you thumb.gif
post #40 of 43
I once saw a guy on a forum who used Prime95 for 5 minutes to test stability. And then he was wondering why his games would crash the PC. Stability is sometimes really subjective. I use 8-12 hours of P95 to check stability. After moving from X58 to X79, Prime95 was much harder to get stable than LinX. wink.gif I've thought that I _never_ use PC for more than 10 hours a day, or stress it. Never biggrin.gif

In the past, when I overclocked my P120@133, P2-350@392 and P3-600@810 and so on, there were not stress programs available like we have now or at least I didn't know about them. Sometimes the PC would just do something weird while gaming and then I thought - hey, maybe it's the OC again...but in the long run, I used games and general usage to check stability.

Also I have started seeing guys on forums doing 4.8 GHz on 3930K with air cooling, and I simply cannot believe they would be fully stable - in some particular games, maybe? They often post screenies of IBT passing like 10 loops, with lower GFlops than the processor would do at stock! Which makes me believe their gaming performance would be actually worse than at lower OC. But, since technology moves forward so fast and no enthusiast keeps his PC for more than few years...why not overclock it so that you can fully use it without crashes for the things you do without fear or data loss or corruption?

I still feel like to be 8-10 hours P95 stable though. thumb.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Friend argues Prime 95 is useless. I need others opinions to ease it.