Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 24 of 33 FirstFirst ... 142223242526 ... LastLast
Results 231 to 240 of 322

Thread: Stable E8400 OC - stock voltages




  1. #231
    Join Date
    Apr 2009
    Location
    J'Habite En Angleterre
    Posts
    2,431

    Default Re: Stable E8400 OC - stock voltages

    Here is some advice I gave in another thread about how to aproach finding minimum Vcore value without repeated Prime testing (max of 2 potential Prime runs needed, rather than several/half a dozen).

    One useful thing about Intel Burn Test is that at insufficient Vcore, the amount of computational speed (GFlops) and also the peak core temps are noticeably less than if the Vcore is set correctly. You can use this to your advantage. Temporarily set a Vcore a couple of notches higher than you have now, just for certainty and do 2-3 runs. Note down the peak temps and more importantly, the GFlops figure given. Now procede to lower Vcore until you notice a drop in GFlops value and a much cooler temp than normal. Restore the lowest Vcore that gave you "normal" GFlops values and procede to test it in Prime/Orthos/OCCT over night. If it passes 8-12 hours, run IBT again for 10-20 runs, and if all passes I'd call it good (Vcore)
    1 hour of each test is no where enough even to get a rough idea of how stable CPU or RAM is.

    Don't fiddle with priority settings, FFT times etc. You should not be using your PC when priming. There's no point trying to find out if the CPU's stable and web surfing etc at the same time, accessing RAM, Page File, MCH etc etc. That takes any "component isolation" you had away completely.

    Also the statement of "should be fine if you get no errors after 4 hours, especially if you are using my settings." is wrong in so many ways IMO. I can set BIOS settings that'll pass 4 hours of ANY test and yet won't even let me extract a compressed archive without giving me CRC and corruption errors. Michael Jackson himself could beam you BIOS settings from beyond the grave and it doesn't mean they'll work just because "they're his"

    Billy Jean wants a BIOS template
    I posted back and said that "This is the one"
    But the PC would not run!
    Coolermaster CM 690 II advance Case
    Corsair HX750 (CWT, 91%(80+ Gold rated @230V) single 62A 12V rail
    P55A-UD4 v2.0 @ F14
    Core i5 760 @ 20 x 201, 4.02GHz
    TRUE Black with a single Noctua NF-P12 pumping out 55 CFM @ 19db .
    2 x 2GB Mushkin Ridgeback (996902), @ 7-10-8-27, 2010-DDR, 1.66v
    2 x Gigabyte GTX 460 1024MB in SLI (Pre OC'd to 715MHz core and 1800MHz VRAM) @ 850 Core / 4100 Mem.
    Intel X25-M Boot Drive (OS and Programs) 200MB/s Read & 90MB/s Write
    Corsair X32 200MB/s Read & 100MB/s Write
    WD Caviar Blue 640GB C (Steam, Games, Storage, Temp Files & Folders, etc)
    Samsung F3 500GB Backup/Images
    Noctua 1300RPM 19dB case fan (rear extraction)
    3 x 140 MM Coolermaster LED fans (one front intake, one top extraction, one side intake)
    Dell Ultra Sharp 2209WAf E-IPS @ 1680x1050

  2. #232
    Join Date
    May 2009
    Location
    Canada
    Posts
    178

    Default Re: Stable E8400 OC - stock voltages

    Psycho, I actually love your IBT method of a more expedient technique of stability testing for final or interim settings, and was going to make reference to it. However, I opted out of mentioning it based on SBMongoos's responses on this thread. I just wanted to make it as simple as possible for him.
    1) Gigabyte EP45-UD3P (rev 1.1, bios F9) E8400 @445x9@1.2625v bios, TRUE 120, Scythe UltraKaze 44cfm, 4x2gb GSkill 8500, Visiontek 3870, Corsair TX850, Antec 900, Win7-64 Home.
    2) Gigabyte EP45-UD3P (rev 1.1, bios F9) E8400 @500x8.5@1.2875v bios, TRUE 120, Schythe UltraKaze 44 cfm, 4x2gb Buffalo FireStyx 8500, XFX 5770, Corsair AX850, CM Scout, WinXP Pro/Win 7-64 Home.
    3) Gigabyte P55A-UD4P (rev 1.0, bios F5) i5 750 @190x20@1.30625v bios, CM Hyper 212+, 4x2gb G.Skill Ripjaw 1600, 7900GS, Corsair TX750, CM Scout, Win7 Pro-64.
    4) Asus P8Z68-V Pro (bios 0801), i5-2500K @44x100@1.248v cpuz, Megahalem rev.B, 4x4gb GSkill Sniper 1600, Corsair AX850, Antec 902, Corsair Performance 3 128GB, Win7-64 Home.


  3. #233
    Join Date
    May 2009
    Location
    Canada
    Posts
    178

    Default Re: Stable E8400 OC - stock voltages

    Quote Originally Posted by SBMongoos View Post
    I see I need to use CPU-Z for the VCore info per your notes. However, what do you mean by "what is your vcore after droop as per cpuz". I see that CPUZ has Core Voltage on the CPU tab so it seems clear to me this is what you're referring to. Need details on the "after droop" thing.
    Your vcore value will likely be lower when the cpu is under load (eg. priming) then when it is idle. What is your vcore value as per cpuz when it is under load?
    1) Gigabyte EP45-UD3P (rev 1.1, bios F9) E8400 @445x9@1.2625v bios, TRUE 120, Scythe UltraKaze 44cfm, 4x2gb GSkill 8500, Visiontek 3870, Corsair TX850, Antec 900, Win7-64 Home.
    2) Gigabyte EP45-UD3P (rev 1.1, bios F9) E8400 @500x8.5@1.2875v bios, TRUE 120, Schythe UltraKaze 44 cfm, 4x2gb Buffalo FireStyx 8500, XFX 5770, Corsair AX850, CM Scout, WinXP Pro/Win 7-64 Home.
    3) Gigabyte P55A-UD4P (rev 1.0, bios F5) i5 750 @190x20@1.30625v bios, CM Hyper 212+, 4x2gb G.Skill Ripjaw 1600, 7900GS, Corsair TX750, CM Scout, Win7 Pro-64.
    4) Asus P8Z68-V Pro (bios 0801), i5-2500K @44x100@1.248v cpuz, Megahalem rev.B, 4x4gb GSkill Sniper 1600, Corsair AX850, Antec 902, Corsair Performance 3 128GB, Win7-64 Home.


  4. #234
    Join Date
    Apr 2009
    Location
    J'Habite En Angleterre
    Posts
    2,431

    Default Re: Stable E8400 OC - stock voltages

    I wasn't belittling any suggestions made to the OP. I did however think that using IBT would be easier than just using prime. I know it's another program to use, but it'll save alot of adjusting in the long run. Unlike Prime95, this should also truely let you isolate the CPU Vcore first. Any prime failures after are then much more likely to be because of RAM/MCH.

    I also wasn't referring to you about BIOS templates. You have said in all your posts (as do most here) that they are suggestions only.

    Note that for even more ease, the Vcore after each run of IBT (3 runs = ~1.5 min at ~26-28GFlops) you can use ET6 to increase the voltage step by step or decrease step by step depending on findings. I would suggest setting Vcore nice and high but not dangerously slow to get the GFLops value for the current clock speed to compare with all lower voltage runs.

    BoombasticSloth is correct. Vdrop = difference between what you set in the BIOS and what you see in CPU-Z at idle.

    VDroop = difference between idle reading and 100% CPU load reading.

    Example: PC 1's CPU Vcore is set in the BIOS to 1.3625V. When booted to Windows, CPU-Z shows 1.349V. This is VDrop and is acceptable.

    When the user loads PC 1's CPU using Prme/IBT the CPU-Z voltage drops further, to 1.322V. Again, this is normal.

    LLC will not stop VDroop (difference between BIOS settings and idle) but will dramatically reduce VDroop at load.

    Example, the same PC now has LLC on . The CPU-Z voltage at idle is still 1.349V. When loaded, the VCore stays exactly the same at 1.349V. On occasion some PC's Vcore may oscillate between two values with LLC on. This is nothing to worry about. The PC in the example may load at a voltage that changes between 1.349V and 1.342V once every second/CPU-Z refresh interval.

    I know it can get confusing and sometimes be like walking through treacle as to what to do next etc.

    Here is a nice clear to do list.

    1. Download IBT.
    2. Open it
    3. Open ET6
    4. Open CoreTemp/RealTemp
    5. Using ET6, set core voltage a couple of clicks higher than it currently is.
    6. Watching the temps, run 2 passes of IBT on Standard.
    7. Write down the GFLops values given by the program
    8. Set Vcore to the original value
    9. Again run IBT and compare GFlops values. If they are around the same (within ~1.5GFlops) procede to decrease Vcore a click and repeat IBT
    10. Repeat this until you see a noticable fall in GFlop value.
    11. Restore the lowest Vcore value that gave you a "normal" GFlops value.
    12. Add a single click of Vcore


    Note down the exact Vcore value you are now using and restart the machine. Enter this Vcore value manually in the BIOS before testing any further.

    Now test over night with Prime set to Blend, all standard settings should be fine.
    Coolermaster CM 690 II advance Case
    Corsair HX750 (CWT, 91%(80+ Gold rated @230V) single 62A 12V rail
    P55A-UD4 v2.0 @ F14
    Core i5 760 @ 20 x 201, 4.02GHz
    TRUE Black with a single Noctua NF-P12 pumping out 55 CFM @ 19db .
    2 x 2GB Mushkin Ridgeback (996902), @ 7-10-8-27, 2010-DDR, 1.66v
    2 x Gigabyte GTX 460 1024MB in SLI (Pre OC'd to 715MHz core and 1800MHz VRAM) @ 850 Core / 4100 Mem.
    Intel X25-M Boot Drive (OS and Programs) 200MB/s Read & 90MB/s Write
    Corsair X32 200MB/s Read & 100MB/s Write
    WD Caviar Blue 640GB C (Steam, Games, Storage, Temp Files & Folders, etc)
    Samsung F3 500GB Backup/Images
    Noctua 1300RPM 19dB case fan (rear extraction)
    3 x 140 MM Coolermaster LED fans (one front intake, one top extraction, one side intake)
    Dell Ultra Sharp 2209WAf E-IPS @ 1680x1050

  5. #235
    Chike is offline Senior Member
    Join Date
    Jun 2009
    Location
    Israel
    Posts
    6,531

    Default Re: Stable E8400 OC - stock voltages

    Quote Originally Posted by Psycho101 View Post
    1. Again run IBT and compare GFlops values. If they are around the same (within ~1.5GFlops) procede to decrease Vcore a click and repeat IBT
    IMO 1.5GFlops is a huge difference, I have XP Home Edition and run mine at complete idle, and with 1624MB my results have no more than 0.05GFlops different between high and low result. I figure if they are higher something is wrong, vtt and CPU ref maybe.
    I don't know it might be OS service than run at the the time.

  6. #236
    Join Date
    Apr 2009
    Location
    J'Habite En Angleterre
    Posts
    2,431

    Default Re: Stable E8400 OC - stock voltages

    1.5GFlops is a good indicator of there something being wrong. Even if Vcore is out by only one increment, the drop will be more than 1.5GFlop. Background tasks such as Windows defender can drop GFlops values by up to 1GFlop in extreme circumstances. Also background downloads etc (I recommend stopping Limewire and other P2P/torrent programs) can have an effect.

    I have variances of up to one and half a GFlop at times. For me I know it's nothing to worry about.

    I personally doubt the usefulness of higher than normal GFlop readings. Put it down to good luck on the part of the CPU. If it were linked to CPU Term and CPU Ref, then it would be like saying that incorrect values of VTT and CPU Ref can increase CPU Performance (GFlop measurements are refering to how fast the CPU solves the IBT set problem). CPU Term and VTT IMO would more than likely also cause values values lower than normal OR cause the PC to freeze when running IBT.

    The main reason why I said ~1.5GFlops difference is that if Vcore is too low, the difference will for sure be way more than that. the 1.5GFlops covers every eventuality.

    Example: My Q8200

    At 1.3625V with a core speed of 3.4GHz, Average GFlops value was 37.
    This seemed little improvement, maybe even a step back over 3.2GHz, so I increased Vcore until I got an average of ~44GFlops. 43.5914GFlops was the minimum result and 44.8188 was the maximum, a variation of 1.2274GFlops. This was with a minimal system config. No indexing, Windows defender etc. Msconfig was used to disable all startup programs. Further Vcore increases gave nothing additional. I tried two ticks higher still, testing with IBT 50 runs on each. Everything was virtually the same, bar some minute natural variation.



    How the board handles power distribution, having LLC on or off, any and all PC components used can all reasonably cause variations between an individuals results.

    My E7400 behaves in roughly the same way, lower GFlops ofc, but the max variance is ~1GFlop, +/_0.5 The 1.5GFlops buffer seems pretty valid for me. Any smaller, and certainly as tight as 0.05 would have people belting up the VCore without need until eventually realising their results are normal. No harm done, but certainly time wasted IMO.
    Coolermaster CM 690 II advance Case
    Corsair HX750 (CWT, 91%(80+ Gold rated @230V) single 62A 12V rail
    P55A-UD4 v2.0 @ F14
    Core i5 760 @ 20 x 201, 4.02GHz
    TRUE Black with a single Noctua NF-P12 pumping out 55 CFM @ 19db .
    2 x 2GB Mushkin Ridgeback (996902), @ 7-10-8-27, 2010-DDR, 1.66v
    2 x Gigabyte GTX 460 1024MB in SLI (Pre OC'd to 715MHz core and 1800MHz VRAM) @ 850 Core / 4100 Mem.
    Intel X25-M Boot Drive (OS and Programs) 200MB/s Read & 90MB/s Write
    Corsair X32 200MB/s Read & 100MB/s Write
    WD Caviar Blue 640GB C (Steam, Games, Storage, Temp Files & Folders, etc)
    Samsung F3 500GB Backup/Images
    Noctua 1300RPM 19dB case fan (rear extraction)
    3 x 140 MM Coolermaster LED fans (one front intake, one top extraction, one side intake)
    Dell Ultra Sharp 2209WAf E-IPS @ 1680x1050

  7. #237

    Default Re: Stable E8400 OC - stock voltages

    Quote Originally Posted by BoombasticSloth View Post
    Your vcore value will likely be lower when the cpu is under load (eg. priming) then when it is idle. What is your vcore value as per cpuz when it is under load?
    Looks like either 1.28v and 1.296v. I've seen both under load. But seems to level at 1.28. Running the Blend test in my earlier notes. Not sure if it's influenced as I'm writing this while the test is running.
    Gigabyte EP45-UD3R (rev 1.1, bios F12), E8400 @ 3.83Ghz, Xigmatek S1283, 2x2Gb OCZ2RPR11504GK Reaper PC2-9500 HPC 1150Mhz, PNY 9600 GSO 768MB PCI-e OC , CoolerMaster 590, WinXP SP3 32bit and Win7 Pro 64 Bit (swappable HDD's via Icy Dock)


  8. #238
    Chike is offline Senior Member
    Join Date
    Jun 2009
    Location
    Israel
    Posts
    6,531

    Default Re: Stable E8400 OC - stock voltages

    Quote Originally Posted by Psycho101 View Post
    No harm done, but certainly time wasted IMO.
    We all have different definition for waste of time, stability, safty etc. Only took me ~3 hours (3rd attemt) at 3.8GHz, not that bad is it?
    I agree difference in vcore during the test should be looked into, e.g. LLC enabled vs disabled.
    Another thing that might be a factor is how far the CPU is pushed. It's expected it will get less stable toword the edge.
    I'll experiment a bit with CPU Ref (my vtt options are very limited) and post the results.

  9. #239

    Default Re: Stable E8400 OC - stock voltages

    Quote Originally Posted by Psycho101 View Post
    I wasn't belittling any suggestions made to the OP. I did however think that using IBT would be easier than just using prime. I know it's another program to use, but it'll save alot of adjusting in the long run. Unlike Prime95, this should also truely let you isolate the CPU Vcore first. Any prime failures after are then much more likely to be because of RAM/MCH.

    I also wasn't referring to you about BIOS templates. You have said in all your posts (as do most here) that they are suggestions only.

    Note that for even more ease, the Vcore after each run of IBT (3 runs = ~1.5 min at ~26-28GFlops) you can use ET6 to increase the voltage step by step or decrease step by step depending on findings. I would suggest setting Vcore nice and high but not dangerously slow to get the GFLops value for the current clock speed to compare with all lower voltage runs.

    BoombasticSloth is correct. Vdrop = difference between what you set in the BIOS and what you see in CPU-Z at idle.

    VDroop = difference between idle reading and 100% CPU load reading.

    Example: PC 1's CPU Vcore is set in the BIOS to 1.3625V. When booted to Windows, CPU-Z shows 1.349V. This is VDrop and is acceptable.

    When the user loads PC 1's CPU using Prme/IBT the CPU-Z voltage drops further, to 1.322V. Again, this is normal.

    LLC will not stop VDroop (difference between BIOS settings and idle) but will dramatically reduce VDroop at load.

    Example, the same PC now has LLC on . The CPU-Z voltage at idle is still 1.349V. When loaded, the VCore stays exactly the same at 1.349V. On occasion some PC's Vcore may oscillate between two values with LLC on. This is nothing to worry about. The PC in the example may load at a voltage that changes between 1.349V and 1.342V once every second/CPU-Z refresh interval.

    I know it can get confusing and sometimes be like walking through treacle as to what to do next etc.

    Here is a nice clear to do list.

    1. Download IBT.
    2. Open it
    3. Open ET6
    4. Open CoreTemp/RealTemp
    5. Using ET6, set core voltage a couple of clicks higher than it currently is.
    6. Watching the temps, run 2 passes of IBT on Standard.
    7. Write down the GFLops values given by the program
    8. Set Vcore to the original value
    9. Again run IBT and compare GFlops values. If they are around the same (within ~1.5GFlops) procede to decrease Vcore a click and repeat IBT
    10. Repeat this until you see a noticable fall in GFlop value.
    11. Restore the lowest Vcore value that gave you a "normal" GFlops value.
    12. Add a single click of Vcore


    Note down the exact Vcore value you are now using and restart the machine. Enter this Vcore value manually in the BIOS before testing any further.

    Now test over night with Prime set to Blend, all standard settings should be fine.

    I'm out the door right now but this looks interesting and will be back to read more. Where can I grab the current version of IBT? I'm running two "Prime" one for each core right now. Stress Prime v0.40.95.13.
    Gigabyte EP45-UD3R (rev 1.1, bios F12), E8400 @ 3.83Ghz, Xigmatek S1283, 2x2Gb OCZ2RPR11504GK Reaper PC2-9500 HPC 1150Mhz, PNY 9600 GSO 768MB PCI-e OC , CoolerMaster 590, WinXP SP3 32bit and Win7 Pro 64 Bit (swappable HDD's via Icy Dock)


  10. #240
    Chike is offline Senior Member
    Join Date
    Jun 2009
    Location
    Israel
    Posts
    6,531

    Default Re: Stable E8400 OC - stock voltages

    Latest version of all programs can be found in the programs sticky. Note that different versions may give different results for the same settings so they may not be comparable.
    Prime95 will run workers for all cores automatically, no need to run multiple instances.

Page 24 of 33 FirstFirst ... 142223242526 ... LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

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