|
Thanks for your help Factory Factory. I ended up ordering a Cooler Master Hyper 212 EVO to replace the retail heatsink. I just finished installing it; the temperatures seem a lot better so far, but I'm having trouble achieving a stable overclock. Ideally I'd like to hit 4.5 GHz or higher, but I'd be fine with 4.2-4.4 if that's not feasible for me. I tried the TLDR recommendation in the OP to just change multipler to 42 in BIOS and call it a day, but I end up getting BSODs (always error code STOP 0x24) when running Intel Burn Test on Max with this method. Since the automatic voltage method doesn't seem to work for me, I've been trying to manually specify the CPU voltage instead. (In case it matters, I'm using an MSI P67A-G45 board.) As I mentioned in my previous post, there are a fair number of settings in the BIOS that don't seem to match the terminology from the OP, and I'm not sure if I should be adjusting any of them. As a result, I've mostly just been playing around with the multipler & CPU voltage and leaving everything else alone, but I haven't had much luck achieving a stable overclock. If it's just a matter of me just needing to plug away more, I'm fine with that, but I figured I would post to make sure there's not something obvious I'm missing. What's the max CPU core voltage I should be willing to try? (I want to achieve a stable OC for daily use) Also, I don't see any way to use the "offset" voltage method in my BIOS. Is this feature not available for MSI, or am I just being dumb and not seeing something right in front of me? chronofx fucked around with this message at 23:00 on Mar 1, 2012 |
# ? Mar 1, 2012 22:56 |
|
|
# ? May 16, 2024 19:05 |
|
1.38 is I believe the 24/7 safe max. Voltage adjustment seems to fluctuate on MSI boards by board and BIOS release, so... I dunno. Can you set it to auto? Make sure if you do that it's not juicing you with 1.4+ or the like. Make sure your power saving features (C1E, EIST) are enabled.
|
# ? Mar 1, 2012 23:14 |
|
Dogen posted:1.38 is I believe the 24/7 safe max. I can set the voltage to auto, but it doesn't seem to be stable (BSOD'd during intel burn test @ max stress level). Also, that brings up another question; what is the recommended method for testing stability? So far, I've been doing 4 thread blended test in prime95 & max stress level in intel burn test, but I'm not sure if these are correct settings for these programs to test stability or if there is another program I should be using instead.
|
# ? Mar 1, 2012 23:20 |
|
Those are both fine. I think you might have to just deal with a fixed voltage if you want to go higher, this is why I hate MSI boards.
|
# ? Mar 2, 2012 00:30 |
|
Run IBT in administrator mode, set number of threads to your number of threads (e.g. 2500K, 4 threads; 2600K/2700K, 8 threads). For a super quick crash test, run a default standard stress check 5 times. For a more thorough stress test, run "Very High" 5 times, then maximum stress for 2 runs to give your CPU, RAM and integrated memory controller a workout. Maximum stress eats all available RAM, if you're unstable due to memory reasons it can make that show up sooner than Prime95. Sometimes you need to increase the voltage to RAM or the VCCIO very slightly when overclocking. When using IBT, don't let temps get too close to/above 80ºC. If you can pass these, you're stable enough to do the real stability testing (and you should probably not run IBT anymore, since high temps and full exercise of the processor logic is hard on the CPU and supporting hardware and nothing else will ever get it that stressed, pretty much ever). Stability testing after that is just Prime95. Make sure to run in administrator mode so Prime95 can make sense of your processor, or it'll make assumptions instead. Blend mode for 12+ hours should tell you if your system is stable or not. The higher heat consumption and small FFT modes are both sort of "specialty" tests, good for stressing the CPU specifically, but Blend is nice because it systematically works through your processor and RAM (and the relationship between them in fetching and execution) and so will root out low-level instabilities nicely. Monitor temps throughout the process.
|
# ? Mar 2, 2012 00:33 |
|
Alright guys, after reading this quote from a Maximus IV review:quote:The BIOS has a nifty "Load Extreme Overclock Profile" setting which allowed us to easily push the 2600K to 4.6GHz with a 1833MHz / 1.65v RAM speed / voltage with Auto voltages set doing nothing but tweaking the processor multiplier. OS would load but it did not like the memory speed under load, so I pushed the memory back to 1600MHz. From there I pushed the CPU farther to 4.9GHz but it required me bumping the vCore to 1.44v manually. As noted this was done with only the tweaking outlined here. I know it's [H] but if the above is true I can just punch in 46 in the CPU Ratio and don't even mess with the voltages? We talked about this before were passed 43x, the voltage needs to be adjusted to get it higher. However with reading the above, then I can just leave things at auto?
|
# ? Mar 2, 2012 03:02 |
|
Yinzer posted:Alright guys, after reading this quote from a Maximus IV review: Two things: 1) 42-43x is the zone because almost every processor will do it on almost every motherboard at a not-to-outrageous voltage. At 46x, you get a lot more exceptions (especially re: voltage limits), so you need more in-depth tweaking. 2) Those voltages they use are way above 24/7 safe limits.
|
# ? Mar 2, 2012 03:17 |
|
I gotcha. I didn't think they manually changed the voltages until they hit 4.9Ghz. Of course that RAM voltage is stupid. I think I'm going to give 45x a go, follow your guide and run it through IBT and Prime95.
|
# ? Mar 2, 2012 03:36 |
|
I went the easy way and just set the multiplier to 42 (i5 with a ASUS board) however after I did so, the motherboard adjusted the BCLK to 103...should I change that back to 100?
|
# ? Mar 2, 2012 15:03 |
|
Yeah. Dunno why Asus boards will do an auto-overclock pass after the first time (per flashed BIOS) you set them to overclock, but just setting your settings a second time will make them stick.
|
# ? Mar 2, 2012 15:07 |
|
They have a history of doing stuff like that to improve their benchmarks for reviews.
|
# ? Mar 2, 2012 18:27 |
|
Just to bring this thread back to life, I just installed a Seasonic Platinum 1000. Upgrading from a Corsair AX850. I think I have a problem
|
# ? Mar 7, 2012 05:43 |
|
Dogen posted:Just to bring this thread back to life, I just installed a Seasonic Platinum 1000. I have the same Seasonic PSU. What's the problem?
|
# ? Mar 7, 2012 10:29 |
|
Spending too much money on power supplies, especially when I already had an extra fancy one. I blame finally having disposable income after several years of school.
|
# ? Mar 7, 2012 13:59 |
|
Unnecessary large/expensive power supplies are probably the biggest wastes of money in the entire PC component world. Somehow my cheapo Corsair 650W has held up just fine for many years on heavily overclocked setups with four HDs and overclocked SLI across multiple machines, and that's probably overkill anyway. I'm sure the people who build $200+ power supplies are laughing their asses off every day when they walk back and forth between the bank and their mansions.
|
# ? Mar 7, 2012 20:03 |
|
PUNCHITCHEWIE posted:Unnecessary large/expensive power supplies are probably the biggest wastes of money in the entire PC component world. Somehow my cheapo Corsair 650W has held up just fine for many years on heavily overclocked setups with four HDs and overclocked SLI across multiple machines, and that's probably overkill anyway. I'm sure the people who build $200+ power supplies are laughing their asses off every day when they walk back and forth between the bank and their mansions. They're magnificently built and engineered (well, some of them are), but in large part unless you are running three video cards, incredibly unnecessary. Also the people who actually build the things live in Foxconn-esque dorms, you are probably thinking of the people who own the companies that sell the things. But your point is well taken.
|
# ? Mar 7, 2012 21:17 |
|
Dogen posted:Just to bring this thread back to life, I just installed a Seasonic Platinum 1000. I have a AX750, which I probably do not really need. It was on sale though! And 80 PLUS Gold! (really, my parent's machine needed a new PSU so I decided to snag a new one to replace my 500W NeoPower that had 3 12V rails)
|
# ? Mar 7, 2012 21:28 |
|
I pulled a 1000 watt power supply from a Dell machine with a bad motherboard that was getting thrown out. It's still with me 4 years later. The actual reason I came to post was to talk about noise. If you're trying to quiet down your case by replacing fans don't ignore your older "quiet" fans. I replaced almost every fan in my computer before I thought to try unplugging an old Sunon that was originally nearly silent. Visually it wasn't doing anything weird but it made a terrible whine that I couldn't pinpoint. The worst part was I cut up the fan shrouds on my videocards so I could disable the stock fan and run really slow 120mm fans on them. At least they're 20 degrees cooler now.
|
# ? Mar 8, 2012 00:02 |
|
movax posted:I have a AX750, which I probably do not really need. It was on sale though! And 80 PLUS Gold! I ended up with the 850 in probably the same sale because I said "what the hell, it's only $10 more!" craig588 posted:I pulled a 1000 watt power supply from a Dell machine with a bad motherboard that was getting thrown out. It's still with me 4 years later. If you buy pretty good fans in the first place, especially with a sealed bearing (FDB type) you avoid a lot of these problems over the life of the fan. Pretty good advice when de-noising an old system, though
|
# ? Mar 8, 2012 00:29 |
|
Microcenter (if you have one near you. In store only.) is running a crazy deal on Z68 mobo+2500k. 179 in store (already a 45 dollar savings) and 50 dollars off any board in stock.
|
# ? Mar 9, 2012 09:51 |
|
So after a little bit of trial and error on a chain of BSODs, I think I've got a bad ~factory overclocked~ gtx570. A quick GPU stress test on OCCT crashes my system at (an approximated)75C. (No bluescreen, but a minidump file. dxgkrnl.sys, dxgmms1.sys, ATKDispLowFilter.sys) Are there any other tools to test defects outside of benchmarks? Should I RMA this card, or try lowering the clockspeed with the manufacturer's software? (That'd be Asus SmartDoctor, in this case.) It's worth noting, Aida64 and 3DMark 11 both run tests without issue, only OCCT can 'replicate' these regular under-load crashes..
|
# ? Mar 9, 2012 16:05 |
|
They sold it to you as being capable of running at the stock clocks, RMA that sumbitch
|
# ? Mar 9, 2012 17:28 |
|
Same problem with my MSI one. Up the voltage, problem solved. Also going to stock clocks would probably fix it. Seems to be an inherent problem with factory OC 570s. Mine would crash in Crysis 2 but nothing else really.
|
# ? Mar 9, 2012 22:40 |
|
I'm not sure if this is exactly the right place to ask, but since I'm overclocking and having the problem, figure I'll ask. I recently built a computer with the following specs: Intel i5-2500k @ 4.3ghz Asus p8z68V-LE 8gb Gskill DDR3-1333 Ram 128gb Crucial M4 SSD HIS Radeon HD 7950 And I don't think I'm getting the performance I should be. All I've really played so far is Skyrim (My resolution is 1080p) and though I am running it on the max settings, it does seem to slowdown a bit, which I didn't think it would with this card. But my biggest concern was when I decided I was going to OC my video card today; I ran 3DMark11 and my score came out way lower than it should be. Looking at comparable systems on their database, people are getting easily above 8,000 points (http://3dmark.com/3dm11/2843486). I realize his card is overclocked some, but I'm running mine at 925 core and 1575 memory and only getting 5587 marks. In addition, my graphics score particularly is much lower than it should be (its at 5478). Any ideas what could be going on?
|
# ? Mar 11, 2012 06:10 |
|
I'm pretty sure none of the regulars around SH/SC give any shits about 3DMark scores, so that doesn't really give us information. Your card should definitely pretty solidly wreck unmodded Skyrim, though. Silly question, which slot is the video card installed in? The blue slot on top, or the brown slot?
|
# ? Mar 11, 2012 06:12 |
|
Yeah, I didn't assume so, it's just an easy way to isolate out that there at least is some sort of problem. And I just ran Skyrim with Fraps on, and it was barely holding steady at 40fps - clearly not what I should be getting. And it's installed in the top slot.
|
# ? Mar 11, 2012 06:25 |
|
Okay, top slot is the correct slot. Try installing the Catalyst 12.2 drivers released this past week. You might also want to post a screenshot of GPU-Z's main page, and one of the sensor page that's been open while you're playing Skyrim, so we can give that a sanity check. If neither of those is productive, though, you might want to post a thread to Haus of Tech Support.
|
# ? Mar 11, 2012 06:43 |
|
I actually may have fixed it...I had been using 12.2 drivers since I built my computer, but I uninstalled them and rolled back to 11.12. 3DMark score jumped up by 2000 and Skyrim runs at a constant 60+. Weird that the newer drivers would cause so many problems though. Anyway, here's my GPU-Z. Does anything look amiss? http://imgur.com/ooHz9
|
# ? Mar 11, 2012 08:11 |
|
Zacmaniac posted:I actually may have fixed it...I had been using 12.2 drivers since I built my computer, but I uninstalled them and rolled back to 11.12. 3DMark score jumped up by 2000 and Skyrim runs at a constant 60+. Weird that the newer drivers would cause so many problems though. Hit the question mark next to the results for Bus Interface. When you do the render test, does the value change from "@ x1 1.1" to "@ x16 2.0"? If you do it full screen, look in the lower left corner.
|
# ? Mar 11, 2012 08:23 |
|
No, when I run the render test it changes to x1 2.0: http://imgur.com/QTqvj
|
# ? Mar 11, 2012 08:30 |
|
Well, dere's yer problem. I have no clue what is causing it or whether you have a faulty motherboard or video card. But the issue is that your video card is getting 1/16th of the PCIe bandwidth it should be getting, and that's not enough for it to perform properly. Do you have another video card you can test in the PC and/or another computer to test that 7950 in?
|
# ? Mar 11, 2012 08:36 |
|
No, I don't. Is there any other way to test it/fix it? I'll go check in my BIOS and make sure there aren't any PCI-E settings set wrong somehow or something. Edit: Yeah, checked through all the possible settings in my BIOS and the only ones even related to PCI-E were talking about the black second slot. Zacmaniac fucked around with this message at 08:51 on Mar 11, 2012 |
# ? Mar 11, 2012 08:41 |
|
Zacmaniac posted:No, I don't. Is there any other way to test it/fix it? I'll go check in my BIOS and make sure there aren't any PCI-E settings set wrong somehow or something. Verify that you have the latest chipset drivers installed for your motherboard.
|
# ? Mar 11, 2012 13:58 |
|
Zacmaniac posted:No, when I run the render test it changes to x1 2.0: http://imgur.com/QTqvj I think your mobo may be defective then; have any other machines to try the card in? I bet either a cold joint on one of the 32 PCIe caps on the card or maybe a cold joint on the mux (on mobo, if it has one, else caps) is at fault, and the link is down training to x1. I don't think chipset drivers would affect it; BIOS generally sets up the PEG port bifurcation and OS gets those details via the ACPI tables. E: on phone, but can that Asus board make the other x16 slot x4 electrically? If so, try moving card there and see what GPU-Z says. I assume that GPU-Z has been updated to properly report info from 7000-series. movax fucked around with this message at 16:42 on Mar 11, 2012 |
# ? Mar 11, 2012 16:39 |
|
movax posted:I assume that GPU-Z has been updated to properly report info from 7000-series. Yeah, it has. Googled screenshots will show "@ <proper lane width> 1.1" during idle and 2.0 or 3.0 at full bore, depending on the board.
|
# ? Mar 11, 2012 18:10 |
|
So last night before I went to bed, I restarted my comp, went into the BIOS, didn't change anything and booted into Windows. It showed PCI-E as x16, and games were running well. Then I restarted again this morning and it was back to x1. I returned my card to default clocks, tried putting it in the black PCI-E slot, and my computer wouldn't boot into Windows - it blue screened when I tried. Then I put it back into the blue slot and restarted the computer again, and its been at 16x for the past 3-4 restarts. I'm not sure if it's permanently fixed or not, but it seems to be working well. Is there anything that would be making it change from 16 to 1x seemingly at random? I'm wondering if maybe the Catalyst overclocking software is messing something up. Also, I do have the latest chipset drivers installed.
|
# ? Mar 11, 2012 21:18 |
|
Zacmaniac posted:So last night before I went to bed, I restarted my comp, went into the BIOS, didn't change anything and booted into Windows. It showed PCI-E as x16, and games were running well. Then I restarted again this morning and it was back to x1. I returned my card to default clocks, tried putting it in the black PCI-E slot, and my computer wouldn't boot into Windows - it blue screened when I tried. Then I put it back into the blue slot and restarted the computer again, and its been at 16x for the past 3-4 restarts. I'm not sure if it's permanently fixed or not, but it seems to be working well. Is there anything that would be making it change from 16 to 1x seemingly at random? I'm wondering if maybe the Catalyst overclocking software is messing something up. Overclocking could do it, but it sounds like the link is not consistently training...maybe you didn't seat it that well in earlier installations? I feel like the components on the motherboard side are at the extreme end of their tolerance.
|
# ? Mar 11, 2012 22:21 |
|
I've got an EVGA GTX460 SSC+, which comes at 850mhz core, 3900mhz effective ram speed. The EVGA GTX460 FTW runs at the exact same core speed, but has 4000mhz effective ram speed. Is it worth trying to overclock the ram a little bit, or would it be about the same level of uselessness as buying faster ram for my motherboard?
|
# ? Mar 12, 2012 23:17 |
|
22 Eargesplitten posted:I've got an EVGA GTX460 SSC+, which comes at 850mhz core, 3900mhz effective ram speed. The EVGA GTX460 FTW runs at the exact same core speed, but has 4000mhz effective ram speed. Is it worth trying to overclock the ram a little bit, or would it be about the same level of uselessness as buying faster ram for my motherboard?
|
# ? Mar 12, 2012 23:54 |
|
|
# ? May 16, 2024 19:05 |
|
You also buy some throttling room (without hacked firmware, the cards are pretty insistent about staying within a given thermal envelope, which changes depending on the card and manufacturer). Still, if you hit a core/shaders wall, provided that your memory controller and VRAM chips are well cooled it really can't hurt to raise your GDDR5 speed too, within tested safe and stable limits. It will increase bandwidth, good for certain AA methods for example. Speaking of which, you can get totally lost in nVidiaInspector dicking around with the "hidden" AA modes, altering compatibility flags, etc.; I've finally got my GTX 580 actually flexing its muscles with 4x sparse grid SSAA combined with 8xCSAA 4 coverage 4 color mode in Mass Effect 3 - which is different from what looks the best in ME2, there it's 2x sparse grid SSAA combined with 2x2 SSAA + 4xOGAA... Different for Mass Effect 1, too, but Mass Effect 1 does not take AA gracefully, even for an Unreal Engine game. Been dicking around with ambient occlusion, too, though it's a bit of a performance hog for the looks in my opinion. Has to be on quality mode or it's sloppy, clearly lower precision... and on quality mode, even this heavy hitter of a card can't combine it with the above high-powered AA modes. Sparse grid supersampling is the poo poo. It's my favorite shimmer-reduction technology. Even 2xSGSSAA goes very far to reduce visible shimmer in textures. In games without good, engine-level FXAA implementations, I know what I'm going to be using.
|
# ? Mar 13, 2012 00:48 |