Arghhhh.......lost another -bigadv :(

Thanks for your advice on this issue as it is nice to hear from someone who is using the exact same ram, mobo, and chip as me. I am currently running the Cogage True Spirit cooler on this processor. I believe it is the case that is responsible for those temps as taking the side panel off lowers temps by about 6 degrees C almost immediately. Even at 3.46ghz, it is still running around 80 degrees. Just for fun, I ran my computer without the side panel last night and temps were hovering between 69-72 degrees when I woke up this morning.

In about 6 weeks time, I plan on moving everything over to a decent case with good airflow like the CoolerMaster HAF932. At that time, I am seriously considering moving to the Corsair H50 as I hear really good things about it. From reading peoples' sigs on this forum, it seems like half of the hardforum i7 folks use that cooler. I would do it sooner, but my entertainment budget is empty.

I have everything in red set exactly as you have it except for uncore clock ratio which is at 16x(it would not post when I set it at 17x like I read in an OC guide), and PCIx bus which I have locked at 100mhz.
 
I would stick with the True Spirit, both the H50 and that are supposedly pretty similar in performance, with the True Spirit probably actually getting slightly better numbers on average.

If you don't have any pets or small children or similar then you can just keep running without the panel for a while, it shouldn't cause any harm. Heck, some of us around here used to run with a regular fan pointing straight into our open cases to get massive OCs back in the day!
 
In about 6 weeks time, I plan on moving everything over to a decent case with good airflow like the CoolerMaster HAF932. At that time, I am seriously considering moving to the Corsair H50 as I hear really good things about it. From reading peoples' sigs on this forum, it seems like half of the hardforum i7 folks use that cooler. I would do it sooner, but my entertainment budget is empty.
I would like to save you some bucks in recommending not to change your cooler unless you can purchase something truly superior like a full blown WC setup. I haven't seen any reports of the H50 being vastly cooler than what's in your setup now. The True Spirit has even demonstrated to be a bit better by some people. Your money is best spent on a new case or in optimizing what you have now, such as replacing the fan you on the cooler with two high-performance fans in a push/pull config. Also, I strongly recommend reinstalling the HSF and purchase the best compound you can find before contemplating on a totally different cooler. It's a lot cheaper and could reveal the source of the problem. That might be all you need (and a new case). :)
 
I would stick with the True Spirit, both the H50 and that are supposedly pretty similar in performance, with the True Spirit probably actually getting slightly better numbers on average.

If you don't have any pets or small children or similar then you can just keep running without the panel for a while, it shouldn't cause any harm. Heck, some of us around here used to run with a regular fan pointing straight into our open cases to get massive OCs back in the day!

I just don't like running cases without a cover on it. It just doesn't feel right, and this room has a lot of stuff piled up around there. I'm just going to deal with the high temps until I can move to a HAF932 case.

I would like to save you some bucks in recommending not to change your cooler unless you can purchase something truly superior like a full blown WC setup. I haven't seen any reports of the H50 being vastly cooler than what's in your setup now. The True Spirit has even demonstrated to be a bit better by some people. Your money is best spent on a new case or in optimizing what you have now, such as replacing the fan you on the cooler with two high-performance fans in a push/pull config. Also, I strongly recommend reinstalling the HSF and purchase the best compound you can find before contemplating on a totally different cooler. It's a lot cheaper and could reveal the source of the problem. That might be all you need (and a new case). :)

Yeah, I keep forgetting that my computer has the Cogage in it, possibly because of those temps. I was considering the H50 at one time for my gaming rig, until I realized that it requires you to switch the rear exhaust fan into an intake fan, thus you blow a bunch of hot air inside your case off the rad. That would be bad since my graphics cards would be working with hotter air.

I meant to re-seat the Cogage and use some Artic Silver paste, but never got around to it. I started playing FFXIII for the X360 and it ended up taking up my entire weekend. Frickin addictive that game is............

Before going to a new case, i'm gonna try putting in a bigass expansion bay intake fan, and replacing the rear exhaust fan with a much higher output fan.
 
I just don't like running cases without a cover on it. It just doesn't feel right, and this room has a lot of stuff piled up around there. I'm just going to deal with the high temps until I can move to a HAF932 case.

I just realized that my number one threat, looking to overtake me in about 3 days, is none other than you sir, Mr. amdgamer.

Therefore, please feel free to take your time getting you system stable and running -bigadv's at max speed!

I must provide fair warning though, my first -bigadv should hit this evening, and I am going to be setting up another i7-920 system late this week or early next week, so good luck catching me. ;)
 
Last edited:
I just don't like running cases without a cover on it. It just doesn't feel right, and this room has a lot of stuff piled up around there. I'm just going to deal with the high temps until I can move to a HAF932 case.
/QUOTE]

I just realized that my number one threat, looking to overtake me in about 3 days, is none other than you sir, Mr. amdgamer.

Therefore, please feel free to take your time getting you system stable and running -bigadv's at max speed!

I must provide fair warning though, my first -bigadv should hit this evening, and I am going to be setting up another i7-920 system late this week or early next week, so good luck catching me. ;)

Heh, I may never be able to pass you then. My next -bigadv isn't due to finish(assuming it finishes without any problems) until late tomarrow night. As of right now, I can finish a 2681 WU in about 2.6-2.7 days with a TPF of almost 36 minutes so long as it is uninterrupted.

I won't be setting up a second folding rig till likely the upcoming Christmas season. I am hoping that we will have native Windows bigadv folding available by then, and that we get some cheaper Gulftown derivatives on the market since I don't think AMD's Bulldozers will hit till late 2011 from what I hear.
 
I must provide fair warning though, my first -bigadv should hit this evening, and I am going to be setting up another i7-920 system late this week or early next week, so good luck catching me. ;)

Early congrats on your first bigadv and may there be many more with your 2nd i7 coming on-line.
 
Early congrats on your first bigadv and may there be many more with your 2nd i7 coming on-line.

Thanks! When I left the house this morning HFM was predicting about an 8:40pm finish time (PTZ), so I have just over 5 hours to go if all goes well.

What is really annoying is that I lost all my work on the ones I started Friday night and Saturday night because, get this, my UPS was bad and kept cutting power to my PC. I have had this issue for a little while, but usually it would only kick in once every couple months because I wasn't actually using my PC full-out that often. So I wasn't sure if the UPS was causing it or saving me from it (thought I might be pulling too much juice from the wall or something). Further testing this weekend revealed it had to be UPS because when I used a totally different circuit I still had the problem. So I have removed the UPS from the system and so far I am going strong with no problems at a 3.9ghz OC.
 
Well, I guess I get to join the club. I just lost a Big Adv at 99.5%!!! I was 15 mins from completing :( :( :(
 
Thanks! When I left the house this morning HFM was predicting about an 8:40pm finish time (PTZ), so I have just over 5 hours to go if all goes well.

What is really annoying is that I lost all my work on the ones I started Friday night and Saturday night because, get this, my UPS was bad and kept cutting power to my PC. I have had this issue for a little while, but usually it would only kick in once every couple months because I wasn't actually using my PC full-out that often. So I wasn't sure if the UPS was causing it or saving me from it (thought I might be pulling too much juice from the wall or something). Further testing this weekend revealed it had to be UPS because when I used a totally different circuit I still had the problem. So I have removed the UPS from the system and so far I am going strong with no problems at a 3.9ghz OC.

That right there is one reason why i'm currently hesitant to spend good change on a UPS, as i've always heard that they have a useful service life and can do stuff like this when they get old.

Well, I guess I get to join the club. I just lost a Big Adv at 99.5%!!! I was 15 mins from completing :( :( :(

Welcome to the club. Sorry you had to join as I know that feeling when you lose a bigadv that you spent so much time on.
 
This one really hurt...
Code:
[18:21:41] Completed 240000 out of 250000 steps  (96%)
[18:55:59] Completed 242500 out of 250000 steps  (97%)
[19:30:13] Completed 245000 out of 250000 steps  (98%)
[20:04:30] Completed 247500 out of 250000 steps  (99%)
[20:08:33] - Autosending finished units... [May 6 20:08:33 UTC]
[20:08:33] Trying to send all finished work units
[20:08:33] + No unsent completed units remaining.
[20:08:33] - Autosend completed
[20:38:45] Completed 250000 out of 250000 steps  (100%)
[20:38:54] DynamicWrapper: Finished Work Unit: sleep=10000
[20:39:04] 
[20:39:04] Finished Work Unit:
[20:39:04] - Reading up to 52544928 from "work/wudata_02.trr": Read 52544928
[20:39:05] trr file hash check passed.
[20:39:05] - Reading up to 46183832 from "work/wudata_02.xtc": Read 46183832
[20:39:07] - Checksum of file (work/wudata_02.xtc) read from disk doesn't match
[20:39:07] 
[20:39:07] Folding@home Core Shutdown: FILE_IO_ERROR
[20:42:26] CoreStatus = 64 (100)
[20:42:26] Unit 2 finished with 52 percent of time to deadline remaining.
[20:42:26] Updated performance fraction: 0.567724
[20:42:26] Sending work to server
[20:42:26] Project: 2683 (Run 13, Clone 12, Gen 29)
[20:42:26] - Error: Could not get length of results file work/wuresults_02.dat
[20:42:26] - Error: Could not read unit 02 file. Removing from queue.
[20:42:26] Trying to send all finished work units
[20:42:26] + No unsent completed units remaining.
 
DUDE, WTH!!!

It finished and THEN crapped out? I feel for you.
 
Yep, it apparently couldn't write the results file when it finished, so it wiped the whole thing. I'm not exactly thrilled about it...but what are you going to do? Such is life in the bigadv world...
 
That right there is one reason why i'm currently hesitant to spend good change on a UPS, as i've always heard that they have a useful service life and can do stuff like this when they get old..

I use a UPS on every tv and computer in the house and I've never had a problem. The small chance of one going bad is heavily outweighed by the protection you get from crappy electric service, storms, and a host of other electronics killing stuff. The secret with a UPS is that the UPS rarely ever goes bad. It is the battery inside that has a 3 or so year lifespan. So every three years, I take the UPS to a battery store and have them change the battery for around $40. It is totally user replaceable, I just let the battery people do it because their battery prices are good and they do the swap for free. So for $40, I get a brand new UPS. of the 10+ UPS I have owned, I've never had the electronics inside go bad--just batteries. Also, if you decide to get one, I recommend the Cyberpower series with LCDs. Here's an example. Once you have the LCD screen showing wattage output, battery power, current AC voltage, runtime, estimated shutdown, etc., it becomes very addicting. Every couple of months, Newegg, Buy.com, and Amazon put the 1000VA (or higher) versions on sale for under $100 and with free shipping.

Regarding your continuing problems with your system, have you run any kind of disk utility on your hard drive to see if you have a bunch of bad sectors or something? How old is the hard drive?
 
I use a UPS on every tv and computer in the house and I've never had a problem. The small chance of one going bad is heavily outweighed by the protection you get from crappy electric service, storms, and a host of other electronics killing stuff. The secret with a UPS is that the UPS rarely ever goes bad. It is the battery inside that has a 3 or so year lifespan. So every three years, I take the UPS to a battery store and have them change the battery for around $40. It is totally user replaceable, I just let the battery people do it because their battery prices are good and they do the swap for free. So for $40, I get a brand new UPS. of the 10+ UPS I have owned, I've never had the electronics inside go bad--just batteries. Also, if you decide to get one, I recommend the Cyberpower series with LCDs. Here's an example. Once you have the LCD screen showing wattage output, battery power, current AC voltage, runtime, estimated shutdown, etc., it becomes very addicting. Every couple of months, Newegg, Buy.com, and Amazon put the 1000VA (or higher) versions on sale for under $100 and with free shipping.

Regarding your continuing problems with your system, have you run any kind of disk utility on your hard drive to see if you have a bunch of bad sectors or something? How old is the hard drive?

I may try it, but boy they arn't cheap. I do trust lead acid battery technology as it is the most reliable battery tech on the market even today(IMHO).

Can an overclock interfere with the proper operation of a SATA hard drive?
 
Folding won't hurt you hard drive, but drives naturally degrade over time. If you have a lot of bad sectors, there's a good chance that your WU work file is being written to one of them and resulting in corruption. It may not be your issue, but it is a simple thing to check.
 
I may try it, but boy they arn't cheap. I do trust lead acid battery technology as it is the most reliable battery tech on the market even today(IMHO).

Can an overclock interfere with the proper operation of a SATA hard drive?

A bad memory overclock or too high a PCI-E overclock can cause hard drive data corruption.
 
Well my points are going to continue their "in the toilet" trend, seems like now NEITHER of my i7's wants to run bigadv stable. I've been dropping avg PPD over the last week because I keep losing several hours at a time on each machine (due to crash/reboot then sitting idle unbeknownst to me), then the bonus pts get the shaft. [sigh] Looks like I'm in for a weekend of bios, reboot, LinX, repeat ad nauseum. TWICE :mad:
 
So every three years, I take the UPS to a battery store and have them change the battery for around $40.

What is this battery store you speak of, and where might I find one (especially in the Puget Sound region)?
 
What is this battery store you speak of, and where might I find one (especially in the Puget Sound region)?

The one I go to is Batteries Plus, but there are a few battery chain stores out there. The closest one to you looks to be in Seattle. But, the batteries in a UPS are not anything special. Just take them out and google the model number and you will find a bunch of places selling them. I think the ones in my UPS are just marine batteries, so I was going to get them at a boat store until I found Batteries Plus.
 
My newest OC which I had stable overnight wiped out sometime today and took down the dual bigadvs I run. I dialed it back to the last good OC and am going to give it a try again after I let a few Win SMP WUs run for a day.
 
My newest OC which I had stable overnight wiped out sometime today and took down the dual bigadvs I run. I dialed it back to the last good OC and am going to give it a try again after I let a few Win SMP WUs run for a day.

I'm sorry to here that, so I guess you are back down to 4.2ghz then?

I find it incredible how folding bigadv work units are literally the ultimate torture test for your processor. Before I knew about folding, I used to think that systems that were Prime95 stable overnight could be considered to be stable. At least for gaming, I imagine that is probably pretty true.
 
I'm sorry to here that, so I guess you are back down to 4.2ghz then?

Yeah I'm back to 4.2 but that's ok because that's a real nice OC and it is not worth it to me to put the time and effort to wring out even more performance for a marginal improvement. I want 24/7 stability and I'm getting it now, especially with the GPUs being happy too.
 
after I let a few Win SMP WUs run for a day.

Same thing I'm going to do.
Of course the bigadv that was around the 80% mark got flushed earlier when my oc crashed again on my main....and when I got it back up and running, I realized it had flushed it. No sense in wasting 2 days of bigadv time when a crash could flush it and waste it all again....just going to tweak the oc, get it to where I think it's stable, and run A3's for a few days until the A2 hits it's expiration date and flushes (again). :eek:

Bigadv is unbelievable man, you get a system that is rock solid and you can crunch whatever but bigadv makes it look like a playskool computer incapable of anything. (I'm flipping around in CS4 while running Media Monkey with a 33k library, doing Ps/Ai work with both gpu clients going....rock solid. Bigadv though, every few days crashes it like i'm running a weak Vcore on one of those fake i7s. :()
 
I had a power flicker about 6:20 this morning, but thankfully the only thing that rebooted was one of my A3 boxen, and not my -bigadv work. Fortunately I had all my other stuff on APC's.
 
Dumb thing bluescreened this morning but I deserved it. I had to goose the vCore a bit and it's been running smp all day long without a hiccup. Back to bigadvs once the current smp unit finishes.
 
Dumb thing bluescreened this morning but I deserved it. I had to goose the vCore a bit and it's been running smp all day long without a hiccup. Back to bigadvs once the current smp unit finishes.

I'm sorry to hear that. Are you still going to keep running things at 4.2ghz? Still an incredible overclock!
 
I'm sorry to hear that. Are you still going to keep running things at 4.2ghz? Still an incredible overclock!

Hell yeah! I think once the dust cleared from my misadventures a little extra vcore seems to have stabilized things and after 12 hours of successful SMP/GPU folding I've switched back over to bigadv/GPU folding. I'm pretty sure all is well now and just in time, I've got too many people in my radar now.
 
4.2 on a 980 is very nice. :eek:
I'm doing final stressing on my primary running 3.9 (21x186). I thought 188 would be the final #, but at some point overnight it crashed while doing a 20 run IBT max mem run. This particular 920 just don't want to run any higher even with a lot more (Vcore) juice. Currently I'm set at 1.325 Vcc and 1.355 Vtt; I can get up to 206blck on 1.355 Vtt stable, but it won't run any stresser for more than a minute at 190+ even with 1.40 Vcc and I don't really want to put that much juice through it for 100mhz. :D Gotta draw a line somewhere LOL.
 
4.2 on a 980 is very nice. :eek:

Thanks. It seems to be quite happy at that speed so I don't see any point in cranking it up any more than that. The entire system is stable and munching on some bigadvs right now.

I'm doing final stressing on my primary running 3.9 (21x186). I thought 188 would be the final #, but at some point overnight it crashed while doing a 20 run IBT max mem run. This particular 920 just don't want to run any higher even with a lot more (Vcore) juice. Currently I'm set at 1.325 Vcc and 1.355 Vtt; I can get up to 206blck on 1.355 Vtt stable, but it won't run any stresser for more than a minute at 190+ even with 1.40 Vcc and I don't really want to put that much juice through it for 100mhz. :D Gotta draw a line somewhere LOL.

Vcore seems high for 3.9, I had my 920 C0 under 1.3 vcc and was running @ 3.8.
 
Last edited:
@ tj
Yeah this cpu seems somewhat anal about liking the juice. :eek:
Well it wasn't passing any IBT with >1.325 Vcc @ 3906, and it seemed like (really) I wasn't much more of a notch above what I used to run, which was 21x180 (3780) @ like 1.30 Vcc and 1.295 Vtt. I originally tried to put duplicate settings into both systems (180bclk, 1.2685 Vcc, 1.295 Vtt), the main (12gb) I could never get stable, the dedicated boxen (6gb) I had to go 1-2 bumps up in Vcore and it's pretty much been stable.

I really wanted to hit the mythological "4" number :D but it just wasn't happening. 1.4 Vcore didn't seem like it was cutting it, and while the e-peen in me wants the 4, the conservative part of me decided to forget about it and live with 3.9 @ juice that won't affect lifespan. :) Ah well at least it's not a completely borked chip that doesn't go above 3.6 or something...could be worse. :(

In any event it finished the 20 runs of IBT @ max mem with those settings (21x186 = 3906, 1.325 Vcc and 1.355 Vtt, 2:8 ram 1333 @ 1480 1.66 Vdimm) so I've got it chugging a -bigadv again.

If it weren't for severe storms tonight and general storms the next 2 days I'd start working on re-doing the oc on the boxen...have to wait til later I guess and hope it stays up and running. :eek:
 
12 gigs of RAM made it a bit more difficult for me to OC as well. I slowed down the RAM speed under the 1333 mine was rated and it worked OK.

Good luck when you chase after the magic "4" again.
 
That is still impressive for a 920 as you should be happy. Besides, the internet is too full of people showing off their e-junk and sometimes it is just better to have something that is stable versus trying to go after something that puts you on the ragged edge.

I couldn't get my 930 past 3.46ghz, but may try again in a month once I get everything moved to a decent case. To honest, i'm not disappointed as this is still netting me 22k ppd versus the 7000-8000 ppd my gaming rig was getting me before I got a dedicated folding rig.

@ tj
Yeah this cpu seems somewhat anal about liking the juice. :eek:
Well it wasn't passing any IBT with >1.325 Vcc @ 3906, and it seemed like (really) I wasn't much more of a notch above what I used to run, which was 21x180 (3780) @ like 1.30 Vcc and 1.295 Vtt. I originally tried to put duplicate settings into both systems (180bclk, 1.2685 Vcc, 1.295 Vtt), the main (12gb) I could never get stable, the dedicated boxen (6gb) I had to go 1-2 bumps up in Vcore and it's pretty much been stable.

I really wanted to hit the mythological "4" number :D but it just wasn't happening. 1.4 Vcore didn't seem like it was cutting it, and while the e-peen in me wants the 4, the conservative part of me decided to forget about it and live with 3.9 @ juice that won't affect lifespan. :) Ah well at least it's not a completely borked chip that doesn't go above 3.6 or something...could be worse. :(

In any event it finished the 20 runs of IBT @ max mem with those settings (21x186 = 3906, 1.325 Vcc and 1.355 Vtt, 2:8 ram 1333 @ 1480 1.66 Vdimm) so I've got it chugging a -bigadv again.

If it weren't for severe storms tonight and general storms the next 2 days I'd start working on re-doing the oc on the boxen...have to wait til later I guess and hope it stays up and running. :eek:
 
Congrats! It must feel good :)

Yes it does. My stats were hemorrhaging and people were getting on my radar, so not a good week of folding for me. Posting 116k yesterday makes me feel much better though.
 
Yes it does. My stats were hemorrhaging and people were getting on my radar, so not a good week of folding for me. Posting 116k yesterday makes me feel much better though.

Same here, I've dropped more than 1/2 PPD over the last week (went from around 80k to just over 36k as of right now) and I hate to say it but I'm at my wits end on oc'ing and tweaking and I basically threw in the towel last night and have settled for 3.6 with 3 bumps in juice on both sides (Vcc and Vtt; settings are 21x172, 1.2875 Vcc and 1.235 Vtt). I'm tired of the endless tweaking and stress testing....at this point I could care less about going from 3.6 > 3.8 or 3.9, both of my cpu's are duds, they won't do 4.0 even with 1.4 Vcc and Vtt, and the TPF difference for the extra 200-300mhz (3.6 > 3.8 or 3.9) is practically nothing, just a 100 PPD or so.

My dedicated boxen finished a 20 run max mem LinX run this morning, that has been crunching a -bigadv for the last 4 hours. My primary machine is just finished its own 20 run max mem LinX run and will start crunching -bigadv shortly.

I need to get these people off my threat list...even if they're listed as 1yr +. [smacks] Go away!!!! :D
 
Back
Top