ColonialRaptor
Member
sk3tch said:
Thanks man... you knocked me down a spot
Race Your PC Thread of Benchmarking to find out who has the fastest PC on GAF!
- Thread starter ColonialRaptor
- Start date
Thanks man... you knocked me down a spot
Yeah, no.sk3tch said:
I can do 4.7Ghz with 1.40V and even that high I don't like.
Raptor if you can keep it inclusive it will help people estimate their system to others (say the PC thread and GTX460/6850 users) and let them know where they stand.
I'm sure you can get a mod to have a top 100, 100-###, and the main list?
Thanks.
ColonialRaptor
Member
RukusProvider
Banned
sk3tch said:
I just bought another stock fan that comes with the 212+
http://www.newegg.com/Product/Product.aspx?Item=N82E16835103069
I saw below that you're running 1.44v on the 2500k with just a Hyper 212+ and 1 fan? Your temps in IntelBurnTest must be really high, no? Are you using a fixed VCore or offset?
Hazaro said:
Personal choice. No risk with high volts as long as you monitor temps. My temps are A-OK. It's a $200 processor. The absolute worst that could happen is it could blow up. But I monitor it all the time and have ran it through Prime95/IBT/etc. for hours and hours.
RukusProvider said:
Thanks for the link!
My temps max out in the 70s under load. No problems. Fixed 1.44v in the BIOS. I turned off a ton of settings. I need to get in there and document them as I do not recall off the top of my head. But all the power saving/etc. related stuff. Let me know if you want it, no prob I can dig it up.
RukusProvider
Banned
ColonialRaptor said:
Weird. It still shows my old score in the OP.
Either way, I'm going to graciously bow out at this point and see where I end up. I'm happy with my setup and I'm not wanting to tinker anymore. Thanks for the thread! Inspired a lot of us to learn and play with our systems.
Now I'll wait for the real test: BF3 beta
ColonialRaptor
Member
Updated, and I am TOTALLY done for the day 
Please call out any errors you find.
I think I just updated it, I posted my comment before I'd hit save changes.
Please call out any errors you find.RukusProvider said:Weird. It still shows my old score in the OP.
Either way, I'm going to graciously bow out at this point and see where I end up. I'm happy with my setup and I'm not wanting to tinker anymore. Thanks for the thread! Inspired a lot of us to learn and play with our systems.
Now I'll wait for the real test: BF3 beta
I think I just updated it, I posted my comment before I'd hit save changes.
MisterNoisy
Member
ColonialRaptor said:Done
Okay all - Ladders are updated, we're getting mighty close to the character limit for the OP, but I think we'll make it to a Top 100, I put my feelings about this at the bottom of the first post, I'll have to remove that text at some point in order to fit the Top 100 in the first post.
Today's updates are all done and we're up to 89 entries, things are coming along well... once again, please call out any mistakes you see or find, it's getting harder to maintain and keep track of everyone's scores... spend 3 hours on this bitch today... it'll die down once we get to a Top 100 for me, but it's been a lot of work to maintain so far so I hope everyone has been enjoying it and getting a lot out of it so far
You might want to move this to a Google Docs spreadsheet before it gets too much bigger - it'd be a hell of a lot easier to add/update that way.
Black_Stride
do not tempt fate do not contrain Wonder Woman's thighs do not do not
A Human Becoming said:
Dont worry. Laptop vs Desktop, the Desktops always win.
Im sure if there was a Laptop Ladder you'd be way up top.
Thaedolus said:
Damn it!
Damn 2600K.....no real way to kill that...except overclock the CPU slighting higher.
But right now im not in the mood to kill my chip......but if it did die id definetely be buying a 2600K replacement.
To WAR!
Still salty over my RAM. That 6950 spot is rightfully mine :[
I'm not comfortable now changing around the other voltage values (VCC, NB, PLL, etc.) and no way in hell do I trust AUTO settings when my CPU voltage is already high, everything set at extreme, CPU power 130%, etc.
I'm also not comfortable with the possibility of causing excess wear on my mobo/CPU and really wouldn't like to replace either. Heck, I wanted 1.25V on my CPU, but I ended up at 1.30V.
At least that's what I'm holding over from the last time I did OC stuff. As you said it is very personal, but I'd like someone who is clocking to know that past ______ is getting riskier and they might not be sure of what they are doing. That's all.
Volts kill, temperature is just a side effect. If you keep temps way way down you can boost voltage a tad.sk3tch said:
I'm not comfortable now changing around the other voltage values (VCC, NB, PLL, etc.) and no way in hell do I trust AUTO settings when my CPU voltage is already high, everything set at extreme, CPU power 130%, etc.
I'm also not comfortable with the possibility of causing excess wear on my mobo/CPU and really wouldn't like to replace either. Heck, I wanted 1.25V on my CPU, but I ended up at 1.30V.
At least that's what I'm holding over from the last time I did OC stuff. As you said it is very personal, but I'd like someone who is clocking to know that past ______ is getting riskier and they might not be sure of what they are doing. That's all.
TheVampire
Banned
Dizzy-4U said:sk333333333333tch!!!!!!
4.7Ghz, wow. What cooler are you using? I can't get over 4.2 due to my crappy water cooling system. I need to upgrade to an H70 or something to raise it up to 4.5.
crappy water cooling? So not a custom one? So your 580s do 970 on AIR ?
Hazaro said:
I run up to 1.4v daily which is about the max on just about any kind of cooling imho
I did 1.5v for this short 5.153 GHz 3dmark run, as that's about the max that is safe for short amounts imho (if your cooling can handle it). And yes that with everything extreme, extra PLL, 140% overcurrent and a lot more.
I have run above 1.4v daily in the past to do 5 GHz daily, but I quickly noticed the chip was degrading (needing more voltage for the same overclocks), i.e. now needed a larger offset for something as simple as 4.6 GHz even... I've read about this online as well, basically they clock better when they're still "fresh" then degrade a bit which causes them to need more voltage.
Black_Stride said:
wtf is with these ridiculously high clocking 570s? they didn't go that high when they first came out
are these benchmark stable speeds only? What can they run at 100% stable?
edit: also you can improve your score by running the latest Nvidia BETAs, Thaedolus does and he beats your graphical score as well by a small margin because of those drivers
ColonialRaptor
Member
I just added a further 8gig of ram to my computer today, it's so cheap and so worth it.
Even though at 8gig I usually wasn't using more than 60 - 70%, with 16gig it really feels like the computer has real room to breathe or something, feels relaxed.
I also got my CPU overclock up to 4.0ghz.
In BC2 at 2560x1440 with full graphics and 8xAA it doesn't skip a beat and is locked at 60fps constantly, looks absurdly good.
I will do a run without vsync to see what the true frames it can muster are later, but I'm hoping it's like 120+ or something. It's cool to see just a flat line for the fraps display on my g13.
Even though at 8gig I usually wasn't using more than 60 - 70%, with 16gig it really feels like the computer has real room to breathe or something, feels relaxed.
I also got my CPU overclock up to 4.0ghz.
In BC2 at 2560x1440 with full graphics and 8xAA it doesn't skip a beat and is locked at 60fps constantly, looks absurdly good.
I will do a run without vsync to see what the true frames it can muster are later, but I'm hoping it's like 120+ or something. It's cool to see just a flat line for the fraps display on my g13.
MisterNoisy
Member
Code:
nVidia GPUs 60 60.61%
AMD GPUs 39 39.39%
4 GPUs 0 0.00%
3 GPUs 3 3.03%
2 GPUs 19 19.19%
1 GPU 77 77.78%
Intel CPUs 89 89.90%
AMD CPUs 10 10.10%
Max Score 14,841
Mean Score 6,378
Median Score 5,621
Score Total %
>9000 (lol) 17 17.17%
7201-9000 13 13.13%
6001-7200 15 15.15%
5001-6000 17 17.17%
4001-5000 18 18.18%
<=4000 19 19.19%
Mobo Vendors Total %
ASUS 41 41.41%
Gigabyte 23 23.23%
ASRock 10 10.10%
MSI 6 6.06%
EVGA 4 4.04%
Other/N/A 15 15.15%
GPU Vendors Total %
EVGA 20 20.20%
ASUS 12 12.12%
MSI 12 12.12%
XFX 7 7.07%
Sapphire 7 7.07%
Gigabyte 5 5.05%
Zotac 3 3.03%
Other/N/A 33 33.33%
RAM Vendor Total %
Corsair 31 31.31%
G.Skill 29 29.29%
Kingston 7 7.07%
Mushkin 4 4.04%
Patriot 2 2.02%
Crucial 1 1.01%
Other/N/A 25 25.25%
CPU Cooling Total %
Air 86 86.87%
Water/Other 13 13.13%Lots of ASUS fans here.
Black_Stride
do not tempt fate do not contrain Wonder Woman's thighs do not do not
n0n44m said:wtf is with these ridiculously high clocking 570s? they didn't go that high when they first came out
are these benchmark stable speeds only? What can they run at 100% stable?
edit: also you can improve your score by running the latest Nvidia BETAs, Thaedolus does and he beats your graphical score as well by a small margin because of those drivers
My Golden Sample runs at these speeds stable 24/7... well i game with it for hours upon hours, never actually run a 24/7 run (if that actually exists), max temps about 85 degrees when benching
Less when im gaming.
Considering at stock speeds the card would do about 77 on load im not too worried running ~80
I can take the card higher but it wont be real world, it would be one panic run (which with my luck will die just when i need to show off)
P.S Thanks for the tips on Beta drivers , i will try them and hope for the best.
I just did a run with the Nvidia "Battlefield 3 enhanced" BETA drivers that just released, and they score just as good as the BETA ones from 2 weeks ago, and both score significantly higher than the last 275.xx series drivers.
the last 280 WHQL ones were terrible imho but these two BETA releases seem fine and everyone has to upgrade to them anyway for the BF3 beta it seems 
also all these 950+ MHz results from 570s/580s on air make me want to upgrade again lol
the last 280 WHQL ones were terrible imho but these two BETA releases seem fine
and everyone has to upgrade to them anyway for the BF3 beta it seems also all these 950+ MHz results from 570s/580s on air make me want to upgrade again lol
Oh for fucks sake why do I even try to bother ? >__<
Right, so at first the test always always always crashed between test 4 and 5.
The solution to this was to turn of "scan system info", right done. Now it doesn't crash. The test finishes and I'm presented with
" Problem receiving your result "
yeah screw this.
Right, so at first the test always always always crashed between test 4 and 5.
The solution to this was to turn of "scan system info", right done. Now it doesn't crash. The test finishes and I'm presented with
" Problem receiving your result "
yeah screw this.
ss_lemonade
Member
Weird, I just noticed that my 3dmark 11 wasn't up to date so I downloaded all the necessary updates and started some gpu overclocking. Despite running OC'ed, I'm starting to score like 300 or so points lower than before. Is it because of the latest 3dmark version? I'll post my updated score again once I finalize my OC settings.
Black_Stride
do not tempt fate do not contrain Wonder Woman's thighs do not do not
P7323 - 1 * Gainward Golden Sample GTX 570 - 950/2300/1150mV - i5 - 2500K @ 5.0GHz - 8Gb G.Skill no one cares - AsRock Z68 Extreme 3 - Black_Stride - Air - link
Opened a window and pushed my 2500K to what i believe is the limit....probably not stable, been running a couple tests for about an hour now, no issues so far but i can smell them coming.
The 570 well nothing i tried beyond 950/2300 was stable, so i think this is the limit at this voltage and with this cooling.
I was so hoping i could take that Liquid cooled GTX480, but no such luck....beyond volt modding(which i probably wont do) this is my final entry in this race....its been fun and exciting, but i must bow out.
In case anyone was curious here are my temps and volts after running 3D Mark 11 five times.
Opened a window and pushed my 2500K to what i believe is the limit....probably not stable, been running a couple tests for about an hour now, no issues so far but i can smell them coming.
The 570 well nothing i tried beyond 950/2300 was stable, so i think this is the limit at this voltage and with this cooling.
I was so hoping i could take that Liquid cooled GTX480, but no such luck....beyond volt modding(which i probably wont do) this is my final entry in this race....its been fun and exciting, but i must bow out.
In case anyone was curious here are my temps and volts after running 3D Mark 11 five times.
Vulcano's assistant
Banned
Black_Stride said:
Wow, that can't be safe. Incredible performance nonetheless.
Black_Stride said:
w00t victory
although purely because my 2600Ks hyperthreading (and 5.153GHz clockspeed) eats that Physics test lol, you do beat me in the Graphical tests (7199 vs 6938)your CPU temps look pretty good, GPU is double that of mine
Vulcano's assistant said:
fine for just a few benches, probably for longer if he keeps the voltage at 1.4ish
RukusProvider
Banned
Black_Stride said:P7323 - 1 * Gainward Golden Sample GTX 570 - 950/2300/1150mV - i5 - 2500K @ 5.0GHz - 8Gb G.Skill no one cares - AsRock Z68 Extreme 3 - Black_Stride - Air - link
Opened a window and pushed my 2500K to what i believe is the limit....probably not stable, been running a couple tests for about an hour now, no issues so far but i can smell them coming.
The 570 well nothing i tried beyond 950/2300 was stable, so i think this is the limit at this voltage and with this cooling.
I was so hoping i could take that Liquid cooled GTX480, but no such luck....beyond volt modding(which i probably wont do) this is my final entry in this race....its been fun and exciting, but i must bow out.
In case anyone was curious here are my temps and volts after running 3D Mark 11 five times.
I have the same board (extreme 4). Mind sharing hour bios settings?
LabouredSubterfuge
Member
n0n44m said:I just did a run with the Nvidia "Battlefield 3 enhanced" BETA drivers that just released, and they score just as good as the BETA ones from 2 weeks ago, and both score significantly higher than the last 275.xx series drivers.
the last 280 WHQL ones were terrible imho but these two BETA releases seem fine
also all these 950+ MHz results from 570s/580s on air make me want to upgrade again lol
I tried the betas last night like you suggested and I was getting better performance but my overclock was now unstable (where it was stable under 275) and would get crashes. Should this be happening?
Black_Stride
do not tempt fate do not contrain Wonder Woman's thighs do not do not
Vulcano's assistant said:
Well im about to find out how long i can stick it at 5Ghz, but so far its been happy, temps arent too bad for on Air, and the Vcore is locked at 1.44.
After one hour of benchs(benches?) ill officially be part of the 5Ghz club.
The Graphics card is no problem, its safe there ive been using it like this forever pretty much.
And its only slightly hotter than at stock clocks
n0n44m said:w00t victory
your CPU temps look pretty good, GPU is double that of mine
I just realized all the scores above mine on the NVIDIA ladder are 2600Ks....damn Hyperthreading!
Depending on how this setup runs Battlefield 3 my PC might get be getting an upgrade.
Else, ill wait till my next major upgrade and ill be sure to go Hyper
sk3tch said:
No, there is no cut and dry between the two. Increasing either area can lead to a CPU death as well as shorts, internal issues, condensation, leaks, and others.Hazaro said:
I'm not sure where you read this but this effect is miniscule, usually only being a factor over years time with persistent use (if ever). You aren't going to see degradation after a few days of overclocking. Chips actually can have burn-in periods as well similar to RAM. I have had processors that accept greater voltage and maintain better stability after intensive overclocking over time.n0n44m said:
Just an FYI for people, modern Intel CPUs are designed to handled thresholds 120c+. I've run an i7 without a heat-sink for a duration of time without issue. I put a strong warning of not approaching this number but these CPUs are built to handle high TDP.
LabouredSubterfuge said:
if it doesn't happen at clock speeds then the overclock probably never was fully stable, but the card wasn't pushed enough to reveal it. New drivers that make more efficient use of the card have revealed it.
if it also happens at clock speeds then the drivers don't play nice with your setup, which sucks if you like BF3 because the BETAs released this morning are recommended if not required to play the BF3 beta
Black_Stride said:
Hyperthreading is just Master-Master-Race
seriously if 7200+ 3dmarks doesn't run BF3 well I'm gonna
Corky said:
with enough (i.e. a lot or more) voltage, sure!
although a 2600K needs more voltage than 2500K in general because the Hyperthreading is more stressing. You can turn it off but that costs you 3dmark points so that would be somewhat pointless...but it's only fun for benches, 4.5ish is more than enough for any kind of daily use right? Not to mention that you're more or less forced to use fixed voltages instead of the voltage offset, so it never drops voltage on idle and your motherboard's VRM components are much more stressed as well, let alone the fact that your CPU will probably degrade soon and require even more voltage...
gunbo13 said:
I know but this is what happened:
my 2600K did 4.5 without any voltage adjustments on both air and water
then after finishing my water setup and upping it to 5.1 with more voltage (1.48ish I think), it ran stable for around 4 days (good temps ofcourse), then started to BSOD
so I set it back to 4.5 as always, only to BSOD as well D:
I was afraid it was dying, but it turned out it still did 4.5 but now needed a voltage offset to run stable, just like the voltage offset I needed for 4.7 had now increased.
read around a bit on xtremesystems / overclock.net and the like and read some more posts of people who's CPUs needed more voltage after running for some days or weeks on high speeds with high voltage
so now I stick to 4.8 for daily use, with voltage offset that comes down to about 1.4ish Volt
Thermal paste is a variable that can cause changes in clocks speeds over time. The TIM changes structure and positioning after a "burn-in" period. This result in a lot of false positives. This change can also alter heat-sink contact and transfer efficiency. Also, a lot of people re-seat their heat-sinks not realizing that this can completely alter your thermal transfer.n0n44m said:
In your case, I would have to know how you defined your chip being "stable." I used to run prime95 runs on older chips that would fail after 23 hours of testing, passing all the previous. Your chip may appear stable for quite some time and then all of a sudden you BSOD constantly. That doesn't necessarily mean your chip degraded.
Your test period was only 4 days which is definitely a "burn-in" period. Your initial clocking should have handled a 20x linx pass. You should have also tested your memory with a memtest. Most people then leave their chip as is after this and call it stable. However, when you are really pushing processors, this is a terrible practice. You need to account for TIM, re-seats, air flow, and even your ambient room temp. I have had chips fail overclocks simply because my room thermostat was set a bit higher. What I typically recommend is doing 20x linx passes periodically over the first few weeks/month in order to survive your burn-in. If you do not see erratic behavior over time, you can then call it. This is only necessary though for ceiling overclocks, which is basically near 100% what the chip can do per cooling method.
it went from 20 IBT passes and hours of Prime 95 Custom Blend (and 3 months of gaming) stable on 4.5 to failing threads and BSOD within 15 seconds of starting Prime95 on the same clockspeed/motherboard settings (saved the profile). Cooling only improved so that's not a factor. It just wanted more volts all of a sudden ...
edit: it's just a Sandy Bridge thing I guess, here's another post from someone who noticed the same
edit: it's just a Sandy Bridge thing I guess, here's another post from someone who noticed the same
P12206 - 2x EVGA GTX 580 SC 1.5GB (SLI) @ 935/1850/1013 & 1.15v - i5-2500k @ 4.7Ghz @ 1.44v - 16GB Corsair Vengeance LP DDR3 @ 1600 - MSI P67A-GD53 - sk3tch - Air - Link
An extra ~150 points all due to the latest beta drivers (I was on the WHQL ones, prior). Thanks for the heads up, n0n44m.
http://www.geforce.com/Drivers/Results/38108
EDIT: this also serves to illustrate what BS these benchmarks can be, in a way. Both AMD and Nvidia work very hard to optimize their drivers for these specific (and common) benchmark tools. Build your system to play games, people. This is why I love HardOCP.com - they take an objective view with "playability" along with providing pure benchmark results (this is what led me to switch from 6970 CFX to 580 SLI - playability...not raw performance...micro-stuttering was horrible). With that said, I just love tweaking so here I am. I've just seen a lot of people come in here and be disheartened with their results. As long as you can play your games the way you want to play them - don't worry about what any benchmark says.
An extra ~150 points all due to the latest beta drivers (I was on the WHQL ones, prior). Thanks for the heads up, n0n44m.
http://www.geforce.com/Drivers/Results/38108
EDIT: this also serves to illustrate what BS these benchmarks can be, in a way. Both AMD and Nvidia work very hard to optimize their drivers for these specific (and common) benchmark tools. Build your system to play games, people. This is why I love HardOCP.com - they take an objective view with "playability" along with providing pure benchmark results (this is what led me to switch from 6970 CFX to 580 SLI - playability...not raw performance...micro-stuttering was horrible). With that said, I just love tweaking so here I am. I've just seen a lot of people come in here and be disheartened with their results. As long as you can play your games the way you want to play them - don't worry about what any benchmark says.
| Praxis |
Banned
These are my stats now (Not going to bother OC'ing any more), could I be added to the list please
P6763 3DMarks - Inno 3D NVIDIA GeForce GTX 580 (stock) - I7 2600k @ 4290 MHz - 2 x 4096 MB G.Skill 9 @ 667 MHz -Gigabyte Technology Co., Ltd. P67A-UD4 - Air - DeathCubeK
http://3dmark.com/3dm11/1894791
P6763 3DMarks - Inno 3D NVIDIA GeForce GTX 580 (stock) - I7 2600k @ 4290 MHz - 2 x 4096 MB G.Skill 9 @ 667 MHz -Gigabyte Technology Co., Ltd. P67A-UD4 - Air - DeathCubeK
http://3dmark.com/3dm11/1894791
Use linx and not p95. P95 doesn't stress the IMC as well as linx. P95 has been retired.n0n44m said:
http://www.xtremesystems.org/forums/showthread.php?201670-LinX-A-simple-Linpack-interface
http://software.intel.com/en-us/articles/intel-math-kernel-library-linpack-download/
It's possible there were internal changes to the chip but it is rarely the culprit. I've gone through a lot of chips and I've experienced a lot of "all-of-a-sudden" BSODs. However, I have never encountered a single instance where I couldn't fix the situation without pumping more volts. As I said, there are way to many variables to just say it is this or that. It can also be a combination of numerous. The real question is, do you want to redo your current configuration to solve the puzzle or just clock it down, moving on? Typically, users don't even utilize these high clock speeds and it is mostly for kicks. So moving on is the common choice.
Nah...and overclock.net isn't really a great resource...n0n44m said:
mclaren777
Member
Oops, forgot about that.ColonialRaptor said:Need a link for your score dude
http://3dmark.com/3dm11/1890931?show_ads=true&page=/3dm11/1890931?key=zVEmW9Tq3k5pmYfmuE87PT6Bm0BNSf
P7162 - 1x Asus GTX 580 (Ref) @ 893/2004(stock)/1.125gpucore - i7 920 @ 4.20ghz HT ON - 3 x 2 gb corsair @ 1600 mhz @ 9-9-9-24 - Gigabyte X58A-UD7 (v1) - koji - Air (H70) link
Guess I`d probably be able to squeeze some more mhz out of that card but don`t want to spend hours tweaking it... the result of an overclocked GFX in real world applications (games) is minimal imo. But the hell with that we`re benchmarking here, BIGGER NUMBERS = MOAR WIN :lol
Guess I`d probably be able to squeeze some more mhz out of that card but don`t want to spend hours tweaking it... the result of an overclocked GFX in real world applications (games) is minimal imo. But the hell with that we`re benchmarking here, BIGGER NUMBERS = MOAR WIN :lol
LabouredSubterfuge
Member
D
Deleted member 1159
Unconfirmed Member
lowrider007 said:i5 2500k @ 4.5Ghz
GTX 570 @ stock
It seems it's my gpu that's holding me back, I'm using 280.26 drivers atm, do you think it's worth trying the beta's?, I tried overclocking my gpu but it just dies even with a 50Mhz increase to the core, the voltage option is blank in afterburner.
To get the voltage option you have to "unlock" it somewhere in the settings. It's not allowed by default as a sort of fail safe, but it only goes up to 1100mv on my 570 anyway so it's not like you can really do any damage at that voltage...I was getting artifacts at about 50 MHz as well until I adjusted the voltage up some. I adjusted my fan settings to be more aggressive too so I could keep the temperatures down, they really jump when the voltage is boosted. Around 80% fan it's basically loud as hell, but if I'm gaming with my headphones on from 12 feet away (my PC's behind my TV) it's not really noticeable. Bonus space heater for the winter too.
gunbo13 said:
I know but more people post there so if I'm looking for other's people experience there are not many substitutes, xtremesystems has too few people who are actually interested in stable overclocks instead of LN2 benchmarking
most forums either have lots of people where 60% seems to have no clue what they're talking about, or only a few people with lots of experience who usually don't even trust each other lol (watercooling scene is even worse :\ )
IBT uses the same Linpack files that LinX uses right? P95 custom blend with lots of RAM enabled did seem to stress out the IMC enough in my case judging by the crashes when setting the RAM from 2T to 1T
1. rig is built, overclocks to 4.5 fine (IBT/P95/gaming) for months (not a single crash)
2. watercooling goes on, 4.5 runs fine (IBT/P95/gaming) for a week, then I continue overclocking and reach
3. 5GHz at high voltage (1.45v+), runs stable(IBT/P95/gaming), then doesn't after ~5 days (BSOD galore)
4. now the same 4.5 setting from 1 and 2 BSOD's within 20 seconds (P95/gaming) D:
5. 4.5 (and any other speed) now needs more voltage than previously to remain stable.
Temps have been higher (summer) and lower (open windows at night) but the only thing that works is more voltage. Only thing I haven't tried is upping one of the specific voltages (PLL and so on) in the bios, but that would also be upping the voltage right
youcantexplainthat.gif , maybe at the 5GHz+ extreme spectrum but not for what used to be a simple 4.5GHz overclock that had been stable for months (most of the gaming was Bad Company 2 and some Witcher 2, both love CPU power). Same profile loaded from bios as well, and it really went from a week of stable stressing/gaming at 4.5 to crashing within 20 seconds of anything mildly CPU intensive getting loaded, with just those couple of days high voltage 5.0GHz in between.
Could be CPU, could be motherboard, all I know is I was very glad it didn't totally die on me as I bought the CPU abroad because continental Europe was out of stock for the first 6 weeks of Sandy Bridge launch
I've just decided to stick to 4.8 these days ... also just did a 20-pass run of IBT 2.52 at very high 8 threads, perfectly stable as well.
Revolutionary
Member
Holy shit, I broke 10k with the new 285.38 Nvidia drivers! I wasn't planning on updating my score in the thread (ever), but this is too good to ignore.
P10057 - 1x ASUS GTX 590 @ 675/3700/stock voltage - i7 950 @ 4.0ghz - 12GB G.Skill DDR3 @ 1600 - ASUS Rampage Formula III - Revolutionary - Air - Link
P10057 - 1x ASUS GTX 590 @ 675/3700/stock voltage - i7 950 @ 4.0ghz - 12GB G.Skill DDR3 @ 1600 - ASUS Rampage Formula III - Revolutionary - Air - Link
You typically don't need to communicate much on the forums. The stickies rarely ever miss anything. And overclocking these days is so ridiculously easy to do. I was mostly talking about the necessity for p95. Linx + memtest here but it isn't like more is worse.n0n44m said:
I could probably explain it if I had it. But as I said, you either tear down what you got starting over or stick with dialing back. Also, the easiest way to diagnose issues is to have multiple sets of hardware. Then it is just a process of moving things around to isolate the issue. This is not a common though.
That overclock is more then plenty though and going higher won't net anything.
Vulcano's assistant
Banned
I need help. My heatsink arrived today and I finally have the CPU running at 4.5 ghz, but so far I have not been able to increase the GPU clock above 742 mhz. Every time I set it to 843 Mhz , for instance, 3DMark11 crashes ( same happens with the heaven dx11 benchmark).
I have an ASUS GTX 570, i have tried it to overclock it with both asus doctor and nvidia inspector to no avail. this is the clock speed i was aiming for:
Any ideas on what might be causing this?
I have an ASUS GTX 570, i have tried it to overclock it with both asus doctor and nvidia inspector to no avail. this is the clock speed i was aiming for:
Any ideas on what might be causing this?
Not above 742 at all? If you jump to 842 without voltage adjustment it won't run,and there is no guarantee your card does as well as the review sample..
Small steps, 10 mhz at a time, meanwhile up voltage as necessary when needed until you reach the maximum voltage you're comfortable with
Small steps, 10 mhz at a time, meanwhile up voltage as necessary when needed until you reach the maximum voltage you're comfortable with
Vulcano's assistant
Banned
n0n44m said:
I have tried small increments too, just now i run it at 750 Mhz and it crashed after 40 seconds (1.013V), ( it wont even start if i get it closer to 800Mhz).
By the way, some times a window pops saying "display driver stopped and has recovered"
I think this is similar to a crashing problem i had with dx11 games, which i solved by increasing the voltage to from 1.013 to 1.05. Should I try to run it again a voltages of 1.1 or 1.2?