Jump to content

My experience with Prime95 on Haswell

FunkyFeatures

Hello

 

I am not sure if you can call this a review, if it does not count, please by all means move this to the proper subforum :)

About 2 months ago, I built a badass system(IMO) including the I7 4790k. I got it up to 4.5ghz at 1.175vcore, called it a day after a couple of hours of AIDA64.

A month later, I saw that the BSODs might actually be my cpu overclock, and my AIDA trial ran out.

 

Had to use Prime95, and what does everybody say? HECK NO! What do I say? Why didn't I do this in the first place?

I had it on manual voltage, always had because adaptive wouldn't work. 3ish minutes in, bsod. 44x multiplier, couple of hours? Same temps as AIDA64, yet no BSODs, and my system has been rock solid ever since.

I seriosly think Prime95 has a bad reputation, but after this experience, I would never trust AIDA again. I cannot believe that I had trusted AIDA, and not Prime95. Anyways, stable now, and would just like to tell people, that (at least for me) AIDA is not always right when it comes to BSOD, and you should try Prime95 with manual voltages. 

 

It may be an incredibly unpopular opinion, but I trust Prime95 over AIDA now.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

Prime puts more stress on the system than Aida, but it's more less reasonable stress, which is why makes Aida64 the better stress tester since it puts a realistic max load on the cpu. 

PSU Tier List | CoC

Gaming Build | FreeNAS Server

Spoiler

i5-4690k || Seidon 240m || GTX780 ACX || MSI Z97s SLI Plus || 8GB 2400mhz || 250GB 840 Evo || 1TB WD Blue || H440 (Black/Blue) || Windows 10 Pro || Dell P2414H & BenQ XL2411Z || Ducky Shine Mini || Logitech G502 Proteus Core

Spoiler

FreeNAS 9.3 - Stable || Xeon E3 1230v2 || Supermicro X9SCM-F || 32GB Crucial ECC DDR3 || 3x4TB WD Red (JBOD) || SYBA SI-PEX40064 sata controller || Corsair CX500m || NZXT Source 210.

Link to comment
Share on other sites

Link to post
Share on other sites

It was unstable with aida 64 because you had it at 4.5GHz

When you turned it down to 4.4 then it was stable.

 

This has nothing to do with which stress testing program you used.

NEW PC build: Blank Heaven   minimalist white and black PC     Old S340 build log "White Heaven"        The "LIGHTCANON" flashlight build log        Project AntiRoll (prototype)        Custom speaker project

Spoiler

Ryzen 3950X | AMD Vega Frontier Edition | ASUS X570 Pro WS | Corsair Vengeance LPX 64GB | NZXT H500 | Seasonic Prime Fanless TX-700 | Custom loop | Coolermaster SK630 White | Logitech MX Master 2S | Samsung 980 Pro 1TB + 970 Pro 512GB | Samsung 58" 4k TV | Scarlett 2i4 | 2x AT2020

 

Link to comment
Share on other sites

Link to post
Share on other sites

Prime puts more stress on the system than Aida, but it's more less reasonable stress, which is why makes Aida64 the better stress tester since it puts a realistic max load on the cpu. 

That may be right, but if AIDA won't find the occasional BSOD problems, then what is it good for? Running games, BSOD. Aida - nothing, Prime - BSOD.

 

AIDA might be more realistic, but then my games aren't(like games have ever been realistic duuuh)

 

:)

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

The reason most people say not to use Prime95 is because the majority of people have voltage set to adaptive. Both Aida and Prime are good, but for most Haswell users Aida is the safer option. 

 

Aida shouldn't have anything to do with a BSOD, it would be the OC. Really you should get it to pass both benchmarks without BSOD to ensure complete stability. 

Link to comment
Share on other sites

Link to post
Share on other sites

It was unstable with aida 64 because you had it at 4.5GHz

When you turned it down to 4.4 then it was stable.

 

This has nothing to do with which stress testing program you used.

I might have been confusing, It was rocksolid stable on AIDA with 4.5ghz, but 4.5GHZ was unstable in Prime95 and every now and then in gaming.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

I might have been confusing, It was rocksolid stable on AIDA with 4.5ghz, but 4.5GHZ was unstable in Prime95 and every now and then in gaming.

Well, how long were you running Aida for? "A couple hours" isn't enough to prove stability; you really need to run a stress test for 12-24 hours to ensure complete stability. 

PSU Tier List | CoC

Gaming Build | FreeNAS Server

Spoiler

i5-4690k || Seidon 240m || GTX780 ACX || MSI Z97s SLI Plus || 8GB 2400mhz || 250GB 840 Evo || 1TB WD Blue || H440 (Black/Blue) || Windows 10 Pro || Dell P2414H & BenQ XL2411Z || Ducky Shine Mini || Logitech G502 Proteus Core

Spoiler

FreeNAS 9.3 - Stable || Xeon E3 1230v2 || Supermicro X9SCM-F || 32GB Crucial ECC DDR3 || 3x4TB WD Red (JBOD) || SYBA SI-PEX40064 sata controller || Corsair CX500m || NZXT Source 210.

Link to comment
Share on other sites

Link to post
Share on other sites

I might have been confusing, It was rocksolid stable on AIDA with 4.5ghz, but 4.5GHZ was unstable in Prime95 and every now and then in gaming.

And that's because you're not following the proper overclocking guidelines.

 

It seems like 4.5GHz was the highest you got it stable on aida and then you left it there. That's not what youre supposed to do. When you find the maximum clock that your CPU will reach, you either increase the voltage by a few hundredths of a volt, or you downclock the CPU by 100-200MHz

 

You should not leave a CPU at the verge of instability unless you do a thorough stress test. This includes multiple stress test programs, benchmarks, benchmarking GPU at the same time, ram at the same time, and many other things for long periods of time (12+ hours)

NEW PC build: Blank Heaven   minimalist white and black PC     Old S340 build log "White Heaven"        The "LIGHTCANON" flashlight build log        Project AntiRoll (prototype)        Custom speaker project

Spoiler

Ryzen 3950X | AMD Vega Frontier Edition | ASUS X570 Pro WS | Corsair Vengeance LPX 64GB | NZXT H500 | Seasonic Prime Fanless TX-700 | Custom loop | Coolermaster SK630 White | Logitech MX Master 2S | Samsung 980 Pro 1TB + 970 Pro 512GB | Samsung 58" 4k TV | Scarlett 2i4 | 2x AT2020

 

Link to comment
Share on other sites

Link to post
Share on other sites

i used prime 95 and the thermals jst kept creeping u to over 100c and i get idle of 30c

CA-RIP, Cz up hoes down, bang these hoes out with my rounds.

Link to comment
Share on other sites

Link to post
Share on other sites

Well, how long were you running Aida for? "A couple hours" isn't enough to prove stability; you really need to run a stress test for 12-24 hours to ensure complete stability. 

In my brain I included it, apparently not when I wrote it. I was supposed to say 5 hours.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

In my brain I included it, apparently not when I wrote it. I was supposed to say 5 hours.

5 hours isn't long enough. I've had Prime proc a BSOD after 12 hours. 

PSU Tier List | CoC

Gaming Build | FreeNAS Server

Spoiler

i5-4690k || Seidon 240m || GTX780 ACX || MSI Z97s SLI Plus || 8GB 2400mhz || 250GB 840 Evo || 1TB WD Blue || H440 (Black/Blue) || Windows 10 Pro || Dell P2414H & BenQ XL2411Z || Ducky Shine Mini || Logitech G502 Proteus Core

Spoiler

FreeNAS 9.3 - Stable || Xeon E3 1230v2 || Supermicro X9SCM-F || 32GB Crucial ECC DDR3 || 3x4TB WD Red (JBOD) || SYBA SI-PEX40064 sata controller || Corsair CX500m || NZXT Source 210.

Link to comment
Share on other sites

Link to post
Share on other sites

And that's because you're not following the proper overclocking guidelines.

 

It seems like 4.5GHz was the highest you got it stable on aida and then you left it there. That's not what youre supposed to do. When you find the maximum clock that your CPU will reach, you either increase the voltage by a few hundredths of a volt, or you downclock the CPU by 100-200MHz

 

You should not leave a CPU at the verge of instability unless you do a thorough stress test. This includes multiple stress test programs, benchmarks, benchmarking GPU at the same time, ram at the same time, and many other things for long periods of time (12+ hours)

Did not seem to find that when I found an overclocking guide, will keep that in mind next time - the highest I got was 4.6 at 1.175vcore, decided to back off to 4.5 and run the stress test.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

i used prime 95 and the thermals jst kept creeping u to over 100c and i get idle of 30c

Manual or adaptive voltage?

 

5 hours isn't long enough. I've had Prime proc a BSOD after 12 hours. 

I know now then. Thanks for letting me know though.

 

 

 

 

But the point of this post was - Prime95(for me) was much more effective at making the machine BSOD possibly saving people time doing dirty overclocks like I did.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

Manual or adaptive voltage?

 

I know now then. Thanks for letting me know though.

 

 

 

 

But the point of this post was - Prime95(for me) was much more effective at making the machine BSOD possibly saving people time doing dirty overclocks like I did.

manual had it set to 1.149v

CA-RIP, Cz up hoes down, bang these hoes out with my rounds.

Link to comment
Share on other sites

Link to post
Share on other sites

had it running for 30 min

CA-RIP, Cz up hoes down, bang these hoes out with my rounds.

Link to comment
Share on other sites

Link to post
Share on other sites

manual had it set to 1.149v

 

had it running for 30 min

Did you monitor volts to see if Prime95 actually caused the voltage to go bananas?

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

Did you monitor volts to see if Prime95 actually caused the voltage to go bananas?

lol it was manual it was static at 1.149v

CA-RIP, Cz up hoes down, bang these hoes out with my rounds.

Link to comment
Share on other sites

Link to post
Share on other sites

lol it was manual it was static at 1.149v

Sounds more like a cooler issue than prime95 doing funky stuff do your pc.

CPU: I7 4790K(4.6@1.252v)                               Case: Fractal Design Define R5 Windowed(Black)           Cooler: CM 212 EVO + NF F12 iPPC

RAM: HyperX Fury 1600MHZ CL10 2x4GB      Storage: Samsung 850 EVO(250GB) + WD Red(2TB)      PSU: Corsair RM750 (and no, it hasn't blown up!)

MoBo: Asus Maximus VII Ranger                      Graphics: MSI GTX 970 TwinFrozr (1494MHZ Core)       OS: Windows 10 Enterprise

Link to comment
Share on other sites

Link to post
Share on other sites

Yea, Prime95 has a bad reputation because a lot of people don't realize that you have to set the voltage to manual before stress testing, this is the case with any stress test, but especially P95 because it is so strenuous.  When I test with P95, if I am stable in P95 after just 20 minutes, I am rock hard stable.

 

People who use P95 with Haswell/DC that don't set their voltage to manual/static/constant/override before stress testing are going to run into overvolting and overheating problems.  Just be a pro, and use manual voltage when stress testing.

"I genuinely dislike the promulgation of false information, especially to people who are asking for help selecting new parts."

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×