• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

bigadv project 2686

jebo_4jc

[H]ard|DCer of the Month - April 2011
2FA
Joined
Apr 8, 2005
Messages
14,572
Got one of these on my Sr2. Looks good, similar to 2685 or 2682.

Getting 82000 ppd with 17:42 tpf
 
21:30 on Nexus.

I've clocked it down to 3.6GHz to keep it stable. Even then it's good for 62k ppd with these WU.
 
21:30 on Nexus.

I've clocked it down to 3.6GHz to keep it stable. Even then it's good for 62k ppd with these WU.
It wasn't stable? Heat?
Posted via [H] Mobile Device
 
It wasn't stable? Heat?
Posted via [H] Mobile Device

Low 70's on temp.

I'm not quite sure. There are a bunch of option on this board and I'm sure I could get more of an OC out of this, I just don't know what to tweek. I followed the guide and it got me to 3.8 and 4GHz but it wasn't stable.
 
My stats
i7 920 21x191=4011
2:8 ram at 1528
Project: 2686 (Run 7, Clone 14, Gen 4)
31:36 TPF
34,539.5 PPD
75,795 credit+bonus
(on dedicated cpu only boxen)
 
I'm not quite sure. There are a bunch of option on this board and I'm sure I could get more of an OC out of this, I just don't know what to tweek. I followed the guide and it got me to 3.8 and 4GHz but it wasn't stable.
Could be the RAM and/or memory settings in the BIOS. I always found memory-related issues to be the most limiting factor with all the system architectures I've used over the years. It's also the most difficult to determine as a potential OC obstacle. Sometimes only changing modules will reveal the problem.
 
Could be the RAM and/or memory settings in the BIOS.

QFT
You should be able to run 2:6 or 2:8 divider on the ram.
Is your QPI frequency set to 18 (or 36x depending on the bios option)?
Uncore (UCLK) should be set to 16x.

I had oc troubles recently on my GB's because I had both of these set to auto and it was putting uncore at 20x. Once I set it to 16x I was able to go a lot higher, stay stable, and also run lower volts as well.
 
SR-2's don't have an Uncore setting, strangly enough. The QPI freq setting is 4.8/5.2/6.4 if I am not mistaken. I am not sure how that translates to your 18/16 setting. Screenshots for some of the settings are here:

http://www.evga.com/forums/tm.aspx?m=268280

I am thnking that we are setting IOH voltage way too high based on these recommendations. I am stable at 199 blck, but the thing is running hot (mid 80's.) The X58 boards seem to work well with high FSB without cranking up IOH, so I am speculation that this 5520 board should as well. We will see in a bit.
 
Reportedly an Uncore setting was added to the BIOS, but then removed in the most recent BIOS version. There are some posts on the EVGA forum about it. Maybe it will return one day.
 
I've seen 2 guides say to crank up IOH but I've found no reason to do so myself (this is on x58, however).
The 3 way oc.com guide says 1.3 IOH for 1 gpu and 1.35 IOH for 2+ gpus, and linuxrouter's i7 oc guide says to use 1.2 IOH (no specification on # of gpus).

I set mine on 1.12 following someone else's bios setting template for 4ghz and I haven't had any issues. Being that the stock juice is 1.10, putting 1.12 I see no reason why that would be harmful in any way.

If anything, try it, see if it's stable, then if it is, dial it back down and see if the stability is gone again and you'll know whether it makes a difference or not. Again though, some of the volt settings I see on guides I think people are just upping volts and throwing darts at settings just for the sake of trying things out. I see people using 1.3 Vtt for higher bclks and once I changed my uncore to 16x I was able to hit 200+ bclk with only 1.235 Vtt on both of my 920s. Long story short following oc guides to the 'T' will get you a lot of overvolting. You really shouldn't have to touch IOH, PLL, or DRAM off what stock values are. Vcc you will have to bump, and Vtt you will probably have to bump a notch or two for higher bclks, but people don't need 1.3+ Vtt for 200+ bclk, they just need to lower their uncore. If you run 18x uncore or higher, yeah, you'll need 1.3+ Vtt because the frequency is so much higher, but there's no reason to run that high of an uncore anyway...the most you should run (from the consensus I've seen) is double the ram+1, or 17x. Anything higher and it's hard to get stable without going overboard on juice.
 
I'm a big fan of Gigabyte's 4ghz oc guide for when I need a refresher.
 
I'm liking these new -bigadv!!

From top to bottom is 2686, 2685 and 2682

Capture-1.jpg
 
Nice points guys....Gonna have to get off of Linux and back on the Windoze train! ;)

 
Nice points guys....Gonna have to get off of Linux and back on the Windoze train! ;)
LOL, I still have one box running native Linux since X-mas time. Going to try to wean it off 'nix today and see if it can make the -bigadv deadline in XP...
 
Magny Cours box before I took it (temporarily) down to change out heat sinks tomorrow. It was folding for my other team (NCIX), but Wednesday to be a member of the [H]orde because I'm a two-timing folding w[H]ore:

P2686 R12, C16, G9:

Number of Frames Observed: 23

Min. Time / Frame : 00:22:04 - 59,413 PPD
Avg. Time / Frame : 00:22:15 - 58,680 PPD
Cur. Time / Frame : 00:22:08 - 58,883 PPD
R3F. Time / Frame : 00:22:08 - 58,883 PPD
All Time / Frame : 00:22:15 - 58,575 PPD
Eff. Time / Frame : 00:22:35 - 57,710 PPD

Giggidy.
 
Magny Cours box before I took it (temporarily) down to change out heat sinks tomorrow. It was folding for my other team (NCIX), but Wednesday to be a member of the [H]orde because I'm a two-timing folding w[H]ore:

P2686 R12, C16, G9:

Number of Frames Observed: 23

Min. Time / Frame : 00:22:04 - 59,413 PPD
Avg. Time / Frame : 00:22:15 - 58,680 PPD
Cur. Time / Frame : 00:22:08 - 58,883 PPD
R3F. Time / Frame : 00:22:08 - 58,883 PPD
All Time / Frame : 00:22:15 - 58,575 PPD
Eff. Time / Frame : 00:22:35 - 57,710 PPD

Giggidy.
I'm a w[h]ore too. but this team needs it to fight off the onslaught.
 
We take folders of all reputations and backgrounds.

Hell look at SmokeRngs. He has been manwhore-ing it for years. It's how he pays the electric bill :p
 
Magny Cours box before I took it (temporarily) down to change out heat sinks tomorrow. It was folding for my other team (NCIX), but Wednesday to be a member of the [H]orde because I'm a two-timing folding w[H]ore:

P2686 R12, C16, G9:

Number of Frames Observed: 23

Min. Time / Frame : 00:22:04 - 59,413 PPD
Avg. Time / Frame : 00:22:15 - 58,680 PPD
Cur. Time / Frame : 00:22:08 - 58,883 PPD
R3F. Time / Frame : 00:22:08 - 58,883 PPD
All Time / Frame : 00:22:15 - 58,575 PPD
Eff. Time / Frame : 00:22:35 - 57,710 PPD

Giggidy.

Damn that looks fine. Just imagine what you will be able to do if we ever get an overclockable Magny motherboard. With eVGA's motherboard team having gone to Sapphire, I hope they have something like that in the works.

We take folders of all reputations and backgrounds.

Hell look at SmokeRngs. He has been manwhore-ing it for years. It's how he pays the electric bill :p

I wonder where he has been lately, and whether he ever got his a/c fixed........
 
Back
Top