Page 2 of 9 FirstFirst 1 2 3 4 ... LastLast
Results 11 to 20 of 82
  1. #11
    ROG Enthusiast Array
    Join Date
    Jan 2022
    Reputation
    10
    Posts
    36

    I was told to use that to test stability. If it crashes after 15 minutes in Prime, doesn’t that make it unstable and likely to crash on normal use? If I am wrong please correct me and I will use it and see.
    Last edited by toby12f; 02-04-2022 at 03:16 PM.

  2. #12
    New ROGer Array
    Join Date
    Nov 2021
    Reputation
    33
    Posts
    108

    Quote Originally Posted by toby12f View Post
    Edit: MemTest86 passed with 0 errors.
    This is good, because it shows that your RAM is not faulty. But it does not mean that your system is stable after booting into the OS. To test memory stability you need to run the P95 "large FTT" or the "blend test". Both tests do utilize all of your RAM to run the calculations.

    I have almost an identical system like you, same CPU, GPU, board, and also the same Corsair AIO. I have seen it with my own Gaming-F, and on this forum as well with at least a dozen user reports: This board seems to have memory related problems. If you do not get the system stable and you are sure it is not the RAM, at last resort consider RMA. This is what I did with my Gaming-F. The second board was finally working.

    Quote Originally Posted by toby12f View Post
    I was told to use that to test stability. If it crashes after 15 minutes in Prime, doesn’t that make it unstable and likely to crash on normal use? If I am wrong please correct me and I will use it and see.
    You are not wrong. If P95 is crashing your system or producing errors, your system is not stable, even if it does run less demanding applications or games for a while. If you can't run P95, it is just a matter of time that the system will hang or crash with other applications as well. I think the mentioned 15 minutes are the minimum time to run P95. Many ppl are deluding themselves about system stability. An alternative for testing is Intel's XTU test and oc utility. It allows to test the CPU and the RAM separately, in stages. XTU however is running the same tasks as P95.
    Last edited by Chainbold; 02-04-2022 at 04:20 PM.

  3. #13
    ROG Enthusiast Array
    Join Date
    Jan 2022
    Reputation
    10
    Posts
    36

    I will try to contact Asus support. They don’t make it easy. It’s impossible to use the “chat” function from Canada, and their offices are almost never open.

  4. #14
    ROG Enthusiast Array
    Join Date
    Jan 2022
    Reputation
    10
    Posts
    36

    This was resolved. I took it to the retailer and they fixed stuff in BIOS, apparently the XMP mode was wrong and he did it himself. It worked for at least 25 minutes without crashing on Prime95 so that’s good enough for me. As long as I don’t have any BSODs during normal use I think that’s all. Thank you for the help.

  5. #15
    New ROGer Array
    Join Date
    Nov 2021
    Reputation
    33
    Posts
    108

    Quote Originally Posted by toby12f View Post
    This was resolved. I took it to the retailer and they fixed stuff in BIOS, apparently the XMP mode was wrong and he did it himself. It worked for at least 25 minutes without crashing on Prime95 so that’s good enough for me. As long as I don’t have any BSODs during normal use I think that’s all. Thank you for the help.
    Good to hear! Once you get the Strix-F running, it's a fine machine.

    I wonder what the vendor fixed with the XMP mode? The only thing they could do was either resetting the BIOS, and/or selecting non-XMP, default settings.

  6. #16
    Administrator Array Silent Scone@ROG's Avatar
    Join Date
    Mar 2016
    Reputation
    202
    Posts
    2,641

    Quote Originally Posted by Chainbold View Post


    You are not wrong. If P95 is crashing your system or producing errors, your system is not stable, even if it does run less demanding applications or games for a while. If you can't run P95, it is just a matter of time that the system will hang or crash with other applications as well. I think the mentioned 15 minutes are the minimum time to run P95. Many ppl are deluding themselves about system stability. An alternative for testing is Intel's XTU test and oc utility. It allows to test the CPU and the RAM separately, in stages. XTU however is running the same tasks as P95.
    Hello,

    This philosophy is so outdated now that it's borderline flat-out incorrect. So much so that by leading users up this path you're potentially impeding their overclocking potential for no good reason. Unless it was built to run Prime. By endorsing this method other people will always be able to clock higher than you and be as stable as they need to be.

    Remember, no overclock is ever 100% stable - that's something some users fail to wrap their heads around. For some reason, there are still Pro-Prime camps of users who like to subject their CPUs to copious amounts current that otherwise, their normal workload won't ever see. Depending on the workload, all you are doing is impeding your own overclocking range in real-world tasks. The fact Intel quickly acknowledged this by implementing the AVX offset function speaks volumes.


    Even with all that aside, as far as stress testing memory is concerned there are far better tests than running large FFT (or small) Prime that isolate the memory subsystem. I and others in-house haven't used Prime to test stability since Haswell-E and use our systems crash-free on a daily basis. Of course, what you choose to take away from that depends on your mindset.
    Last edited by Silent Scone@ROG; 02-05-2022 at 10:03 AM.

  7. #17
    ROG Guru: Orange Belt Array rjbarker PC Specs
    rjbarker PC Specs
    MotherboardMaximus Z690 Apex BIOS 0811
    Processor12900K w EK Velocity 2
    Memory (part number)32 GB DDR5 - Dominator 5600 / Cl36
    Graphics Card #1Evga RTX 3080Ti FTW3 Ultra w EK Vector
    Sound CardOnboard USB Realtek
    MonitorDell Alienware 34" Ultra wide 120Hz Gsync
    Storage #1Samsung 980 Pro 512G M.2 NVMe
    Storage #2Samsung 980 Pro 1 TB M.2 NVMe
    CPU CoolerEK Velocity 2
    CaseCorsair 900D
    Power SupplyCorsair AX 1600i
    OS W11 HP x64 - 21H1
    Accessory #1 480 * 360 * 240 XSPC Radiators (15 Fans) w 4 Channel 200W Fan Controller
    Accessory #2 XSPC Photon Rez / Vario Pump combo
    Accessory #3 Fesser Flex Tubing - Bits Power Fittings & EK SLi Terminal
    rjbarker's Avatar
    Join Date
    May 2013
    Reputation
    49
    Posts
    406

    Quote Originally Posted by Silent Scone@ROG View Post
    Hello,

    This philosophy is so outdated now that it's borderline flat-out incorrect. So much so that by leading users up this path you're potentially impeding their overclocking potential for no good reason. Unless it was built to run Prime. By endorsing this method other people will always be able to clock higher than you and be as stable as they need to be.

    Remember, no overclock is ever 100% stable - that's something some users fail to wrap their heads around. For some reason, there are still Pro-Prime camps of users who like to subject their CPUs to copious amounts current that otherwise, their normal workload won't ever see. Depending on the workload, all you are doing is impeding your own overclocking range in real-world tasks. The fact Intel quickly acknowledged this by implementing the AVX offset function speaks volumes.


    Even with all that aside, as far as stress testing memory is concerned there are far better tests than running large FFT (or small) Prime that isolate the memory subsystem. I and others in-house haven't used Prime to test stability since Haswell-E and use our systems crash-free on a daily basis. Of course, what you choose to take away from that depends on your mindset.
    This perfectly said...I have been OC'ing for many years and my previous Gen stopped using P95 and Intel Burn Test, its archaic .

    Besides...I game primarily, I certainly dont "play" Prime...the advice above looks like it was copied n pasted from an OC'ing thread in 2015 - 2017

    There are far better n faster methods to check for "General Stability"....get thru a few of these then game or carry on with whatever tasks you normally do and watch for instability.
    Looping R23 for an hour...Realbench...IXTU ...and some Memory tests ....is fine and will catch instability pretty quick.

  8. #18
    New ROGer Array
    Join Date
    Nov 2021
    Reputation
    33
    Posts
    108

    Quote Originally Posted by Silent Scone@ROG View Post
    Hello,

    This philosophy is so outdated now that it's borderline flat-out incorrect. So much so that by leading users up this path you're potentially impeding their overclocking potential for no good reason. Unless it was built to run Prime. By endorsing this method other people will always be able to clock higher than you and be as stable as they need to be.
    I disagree. Intel itself is using P95 for their overclocking and stressing utility. They know why.

    There are of course different ways to test stability, even with a hardcore utility like P95 - depending on the purpose of your system and the stability level you want to achieve. And there is no doubt that P95 (or similar y-cruncher) in particular running with AVX-512 enabled is stressing the system in a way that is far above regular usage. If you want to go soft with P95, disable AVX instructions and run small FFT. But one thing is for sure: No matter how "soft" you intend to test your system, if the P95 blend test is producing errors, freezes, or crashes after only a few minutes running (with AVX instructions disabled) it is 100% for sure you are not stable and will eventually crash. It's a guaranteed event - and telling users here a different story may create unrealistic expectations and eventually frustration.

    Also, as I said before: A passed Memtest is not indicating system stability under OS conditions. It only shows that that memory is running fault free under the given frequency and timings. E.g. Memtest is running my DDR5 at 6000MHz 32-32-32 for hours. If I run the P95 memory test I get errors within a few minutes. And I suppose this is the reason why, for example, when gaming I got occasional freezing or crashes with the memory at 6000MHz.

    Quote Originally Posted by Silent Scone@ROG View Post
    The fact Intel quickly acknowledged this by implementing the AVX offset function speaks volumes.
    AVX instructions have been implemented more than 10 years ago, with Sandy Bridge. They are used by many applications. If you want to test without AVX instructions because you think they are stressing your system too much, you can run P95 or Intel XTU without them.

    Quote Originally Posted by Silent Scone@ROG View Post
    Of course, what you choose to take away from that depends on your mindset.
    Exactly.

    For real, hard core overlocking, testing and validating see here, how they tested:

    https://www.igorslab.de/en/supercool...-alder-lake/4/

    They used y-cruncher with AVX-512 enabled. Their target was to demonstrate highest possible CPU frequency with their direct-die waterblock. They got an 12900 up to 5.5 GHz. For bragging, they certainly could have shown a higher value with a softer stability test.

    Quote Originally Posted by rjbarker View Post
    Looping R23 for an hour...Realbench...IXTU ...and some Memory tests ....is fine and will catch instability pretty quick.
    Intel's XTU stress test is by P95. In the same way as you can do with P95, XTU let you disable AVX instructions, if you want to run a softer test.
    Last edited by Chainbold; 02-06-2022 at 03:39 AM.

  9. #19
    New ROGer Array Jimbo93 PC Specs
    Jimbo93 PC Specs
    Laptop (Model)ABS Master Gaming PC
    MotherboardASUS Prime B560M-A AC
    ProcessorINTEL i5 10400F
    Memory (part number)G.SKILL F4-3200C16D-16GVKB x 2
    Graphics Card #1ASUS DUAL-RTX3060TI-O8G-V2
    Sound CardRealTek HD
    MonitorSAMSUNG 40" LCD
    Storage #1INTEL 660p NVMe SSD PEKNW512GB
    CPU CoolerTHERMALTAKE UX100 ARGB
    CaseTUF GAMING GT301
    Power SupplyASUS TUF GAMING BRONZE 650W
    Keyboard TUF GAMING K1
    Mouse TUF GAMING M3
    Headset/Speakers Cambridge Works PC
    OS Win 11 Pro 21H2 22000.652
    Accessory #1 Windows Feature Experience Pack 1000.22000.652.0

    Join Date
    Oct 2021
    Reputation
    82
    Posts
    543

    Quote Originally Posted by Chainbold View Post
    I disagree. Part of the oc community was always over-optimism and self delusion. Each to its own. Intel itself is using P95 for their overclocking and stressing utility. They know why.

    There are of course different ways to test stability, even with a hardcore utility like P95 - depending on the purpose of your system and the stability level you want to achieve. And there is no doubt that P95 (or similar y-cruncher) in particular running with AVX-512 enabled is stressing the system in a way that is far above regular usage. If you want to go soft with P95, disable AVX instructions and run small FFT. But one thing is for sure: No matter how "soft" you intend to test your system, if the P95 blend test is producing errors, freezes, or crashes after only a few minutes running (with AVX instructions disabled) it is 100% for sure you are not stable and will eventually crash. It's a guaranteed event - and telling users here a different story is only creating unrealistic expectations and eventually frustration.

    Also, as I said before: A passed Memtest is not indicating system stability under OS conditions. It only shows that that memory is running fault free under the given frequency and timings. E.g. Memtest is running my DDR5 at 6000MHz 32-32-32 for hours. If I run the P95 memory test I get errors within a few minutes. And I suppose this is the reason why, for example, when gaming I got occasional freezing or crashes with the memory at 6000MHz.



    AVX instructions have been implemented more than 10 years ago, with Sandy Bridge. They are used by many applications. If you want to test without AVX instructions because you think they are stressing your system too much, you can run P95 or Intel XTU without them.



    Exactly.

    For real, hard core overlocking, testing and validating see here, how they tested:

    https://www.igorslab.de/en/supercool...-alder-lake/4/

    They used y-cruncher with AVX-512 enabled. Their target was to demonstrate highest possible CPU frequency with the with the direct-die waterblock. They got the 12900 up to 5.5 GHz. For bragging, they certainly could have shown a higher value with a softer stability test.
    I actually agree with Chainbold here. But that is because I demand stability. Im still not sure I am going to keep the memory at XMP1, although it has now been set that way for maybe 2 months or so. I do occasional important work on my system, I cant afford corruption of any kind when I do. If I cant get good reliability scores in windows the system maybe cannot do what I need when I need it. But I can certainly see people with gaming systems that do not need that. Now if its just a software bug causing crashes that is different, of course. Just dont run that software. Sometimes the only way to be sure is to run the system on bios defaults.

    Ive been using pc s for business and pleasure a long time and started using them in the early days of micro computers.

  10. #20
    New ROGer Array
    Join Date
    Nov 2021
    Reputation
    33
    Posts
    108

    Quote Originally Posted by Jimbo93 View Post
    I do occasional important work on my system, I cant afford corruption of any kind when I do. If I cant get good reliability scores in windows the system maybe cannot do what I need when I need it. But I can certainly see people with gaming systems that do not need that.
    This is also my point. I'm willing to forsake some additional MHz in return for higher stability. In addition to gaming, I use my system constantly for video editing for business purposes, I cannot afford freezing or crashes. But do not underestimate the stability requirement for gaming. Some of the most popular games, like Far Cry 5 or Rainbow 6 are very CPU demanding.

Page 2 of 9 FirstFirst 1 2 3 4 ... LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •