Quote:
Originally Posted by sierra_bound
It actually hurts performance in my testing. ;) I saw the board running at DDR2-1333 (5-5-5-18) yesterday, the memory is not available yet. :slobber:
Printable View
Quote:
Originally Posted by sierra_bound
It actually hurts performance in my testing. ;) I saw the board running at DDR2-1333 (5-5-5-18) yesterday, the memory is not available yet. :slobber:
I recently found a solution to a problem I have benn having when overclocking my evga 680i board.
When I first got the board I was running 1750mhz QFSB with out any problems whatsoever but once I get into the 1900's I get almost like a cold boot issue. Basically at 1909 QFSB I can boot into windows and 3d bench for hours but once I reset my computer the system wont boot up and will automatically go back to default speeds. So I found that when it dosent want to boot up I have to turn the psu off BEFORE the system changes back to defaults and wait for 10-15 seconds for the motehrboard to completey power down and then I turn the power supply back on and power on the motherboard and then she boots up just fine at 1909 QFSB for some more benching.
Has anyone else have this problem?
Kentsfield has major problems with this board. I was not stable above 310 fsb with a stepping 7 B3. I have since sold it for even money and went back to P5B which hits 440 on Kentsfield with everything at auto...
Quote:
Originally Posted by techstyled
My 1.6375 is actual 1.58 in bios and also on phase. The board undervolts I believe although I have not used a digital multi meter to verify bios readings.
I noticed the Pi thing also. Your p5B Pi score http://img122.imageshack.us/img122/1006/4653gd3.jpg at 11 at 4653mhzQuote:
Originally Posted by sierra_bound
On the 680i my 4500mhz score was 12.18, although I had backfround stuff running, they should be closer, no?
Anyone else have this problem? When I set the boot order to usb floppy/cdrom/hard drive, I get no boot to hard drive. just sits at black screen after post. When I set boot order at hard drive first, all is fine. So to boot to a usb floppy or cdrom, I have to chnage the 1st boot device to that one and diable the 2nd and 3rd. basically the boot order is f'ed and this happens on ship bios and P20 bios.
X6800 retailQuote:
Originally Posted by vmsein
Oh, you mean like this.;)Quote:
Originally Posted by bingo13
http://img.photobucket.com/albums/v3.../DDR2-1333.jpg
The purpose of running Orthos is to test stability, not performance. I'm sure you know that.
Damn so what is evga waiting for new bios release now! I'm getting mine by tomorrow, unfortunately i'll not be able to test it, this commets make me feel blue, i thouhg i did a GDMF purchase but i'm starting to doubt.
lets hope problems will be solved soon and thanks for sharing everyone! ;)
It's Nvida waiting, not EVGA. The same bios will go to BFG etc. Should be out sometime this week.Quote:
Originally Posted by FEEXX
i hope RD600 puts all to shame
My 680i also tops out at exactly 449FSB .. any higher does not work ( 466,480,502). I cannot try ntune as that does not run, just dies when trying to open the overclocking pannel. If I could try ntune I think I would be ok, similar to how UGuru helped the nforce3 on the Abit back of yore :)
Also, if the system is not run in Linked mode and not in sync then performance really suffers.
975 12x300 RAM 800Mhz 4-4-4-12 7900GS 540/750 2001 = 481150
680 12x300 unlinked RAM 800Mhz 4-4-4-12 = 47400 2T and 47800 1T
680 12x300 linked sync 800Mhz 4-4-4-12 = 48250 IT
If you are running 2T and not linked and in sync then it is no wonder SuperPi times are low.....
I'm going to wait for BIOS and ntune revisions or somebody on the forums to find the solution, just got to wait. I'll continue testing tomorrow.
Regards
Andy
I'm having this wierd issue with the 680i on an e6600. When I set the multiplier down to 8x and the FSB to 400MHz, the BOIS startup screen, Orthos, and Nvidia Monitor will report it as 3600Mhz (450Mhz x 8) but CPU-Z and Nvidia Control Panel will report it as it should be, 3200Mhz (400Mhz x 8).
Here's a screenshot showing the issue:
http://img364.imageshack.us/img364/7517/mhzpk6.jpg
Sounds like thisQuote:
Originally Posted by cleanerupper
http://www.xtremesystems.org/forums/...&postcount=150
Regards
Andy
Quote:
Originally Posted by NapalmV5
hope is all we have..:clap:
That sounds exactly like the issue I'm having. Thanks! So does this sound like a bug that should be fixed with a new BIOS release?Quote:
Originally Posted by zakelwe
This happened on my P5B too. Windows, Orthos, and other programs thinks its (default multiplier x FSB = Speed posted) while cpu-z will post the correct numbers.
so it is posting 3600mhz for you because it thinks its 9 (default for E6600) x 400 = 3600mhz.
Yes, it's posting 3600Mhz, but instead of saying 9x400Mhz at bootup, it will post 8x450Mhz. All the while, it is actually set at 8x400Mhz in the bios.Quote:
Originally Posted by damnnit
There must be some mismatched multiplier settings in the bios somewhere.
Got it today evga 8800 x2 and evga 680i -
http://service.futuremark.com/orb/pr...ojectId=634484
no case on air :)
Just starting to play with this board
so far fsb = 471.1 and rising
where do you see the bios version?
mine says 2.053.57 , but zi only see it when booting up and an error occurs, otherwise it flashes by.
evga board btw
interesting... a broad range of results on this chipset.
Hi, I'm also playing with eVGA's NF680i/P19 BIOS in these days:)
My board/CPU goes upto 466MHz.
http://222.151.144.121/c-board/file/...80i_FSB466.png
Mouse(via PS/2) jumping problem was resolved by using USB mouse...
...of course it's a FUGGER's suggestion, many thanks!!!:)
>I cannot try ntune as that does not run, just dies when trying to open the overclocking pannel.
--
I suspect that this OC pannel dying problem is derived from language/localization immaturity.
Yes I'm a japanese and using WinXPSP2 Japanese edition, and I also got same problem.
Next I change region/language option at Windows controll panel from Japanese to English/USA,
then I tried nTune...OC pannel appears correctly.
But still I can't change FSB by nTune...it returns error code:(
Thanks for that information Kyosen. I and other people also have it with English version, however I am the other person who mentioned it are in the UK not USA and so have English UK set I think... maybe English Uk is not proper English also ! :D I am at work but will try it tonight....Quote:
Originally Posted by kyosen
Regards
Andy
Weird because I have no crashing or closing issue with nTune. Also my GPU overclocking screen is different from others'. :confused:
Interesting! Do you have web space to load your ntune files up to so we can download and try ?Quote:
Originally Posted by lopri
Regards
Andy
Quote:
Originally Posted by cleanerupper
Thats exactly the issue I am having
Hi got a question about Bios settings, How does affect performance to have the memory in SLI mode or not? any noticiable difference just using a single video card??
Thx
I've also got a question about overclocking on the 680i boards (evga in particular):
By default, under ht multiplier,
nforce spp -> nforce mcp is 5x
nforce spp <- nforce mcp is 5x
What's the max speed these can handle? Meaning, when OCing, can you leave them at 5x or do you eventually have to lower them to 4x, and when?
Also, can the HT spread spectrum be left on or should it be turned off when OCing?
good question, i would like to hear an answer to that also...i keep mine on 4x with a clock of 410x9...but i am not sure what it really does...Quote:
Originally Posted by Celcius
tRRD
tRC
tWr
tWTR
tREF
setings for 500+ ram speed???
I do have space but can't use it right now. But I downloaded nTune and ForceWare drivers from NVDIA.com so the drivers files should be the same. Below is my GPU Overclocking screen.Quote:
Originally Posted by zakelwe
http://img126.imageshack.us/img126/8090/g80ocei2zv4.png
I think it's just that nTune/NV Control panel is buggy in current state.
You don't have to change HT multipliers. Just leave it at X5. I think it's generally recommended to disable spread spectrum thingy. I disabled them except the SATA spread spectrum, which is set at "Down" something.Quote:
Originally Posted by Celcius
I'm loving the evga 680i so far, but I do hit a hard wall at 449fsb. This is with a 626B E6600 conroe. I have seen several other people at this same wall. Hopefully the new bios will fix this.
so far I am at 445:
http://s2.supload.com/thumbs/default/fsb.JPG
p20 bios better be good!
I have got 1 question...........what memory slots are people running their ram in? Blue/Blue or Black/Black? unless you have 4 sticks....*-)
It is recommended to use blue slots if only using 2 sticks.Quote:
Originally Posted by Grinch
680i seems a bit slow at SuperPi, hopefully better at 3D.
Quote:
Originally Posted by charlie
If you use 1T and have the memory half the FSB speed ( 1:1 in effect) then it is slightly quicker than 975 clock for clock Charlie. You also get the benefit for FM approved drivers for 3dmark of course.
Regards
Andy
Quote:
Originally Posted by zakelwe
Personally I would love to see some 1t action. Here's something you 680i guys can compare to. Untweaked XP Pro with everything in my sig. You can use any fsb you like just match cpu speed. I only have a X1800XT so I won't waste my time with 3d.
Question, how high will 1t go?
http://home.comcast.net/~phippy1/SPI3825.JPG
With a bit of luck I should be able to match that exactly with 1T so I will test tonight .. I doubt it will be as fast though, only seems to be better in 3D ..but we shall see.
Regards
Andy
On my Corsair twin2x-6400c4 @ DDR2-800 I could not do 1T at 4-4-4-12 or 5-5-5-18, 1T gives me thousands of errors in memtest and hangs windows and orthos.Quote:
Originally Posted by zakelwe
New Bios out for EVGA 680i over at there website
sweet
oh yea anyone tested the new bios? p20?
I downloaded P20 and it installed just fine. Is it better? No idea. I am stuck at 3500...
Need some information on configuring that BIOS....
You guys at 4000..... show us some BIOS settings..
Anyone using Mushkin 6400? DDR800 (green) 5-5-5-18
How high did you them to go?
Quote:
Originally Posted by NightRaven
After some testing my FSB went from 475mhz to 479mhz x 7. Not very much but still something.:)
Someone mentioned spread spectrum.
SS is for EMI reduction.(That's electromagnetic interference for those unaware.)
System stability and performance may be slightly compromised with SS enabled. Especially with timing critical devices.
For overclocking even a 0.25% jitter can cause a temporary boost in clockspeed which may cause an overclocked processor to lock up.
If you aren't having EMI problems keep it off.
I've got up to 480FSB with my new E6400. (8x480=3840)
Any stability screenshots?Quote:
Originally Posted by lopri
This for now. Will add more up later on.
FSB 1920
http://img139.imageshack.us/img139/7...1920nn3rr9.png
FSB 1900 with tighter timings
http://img139.imageshack.us/img139/8627/sp8mbhg4jy9.png
@ lopri which memory u got ?
TeamGroup. But this board doesn't love my memory that much. (No 1T love @800MHz and beyond)
Edit: Here is a dual 32M run.
http://img136.imageshack.us/img136/2285/sp32m2ng1.png
Hi malficarQuote:
Originally Posted by malficar
I run out of a case with a tech station so I am not EMI shielded, in my case should I disable it or not, and does it efect temps of the cpu?
You can run it in blue or black:fact:Quote:
Originally Posted by Grinch
I'd say leave it off unless you are getting popping or odd distortion from your sound card due to EMI. I do not believe it would effect CPU temperature on a significant scale.Quote:
Originally Posted by palese
I was running Orthos for a couple of hours at DDR2-1311, but forgot to save the screenshot.:( Here's DDR2-1309
http://img182.imageshack.us/img182/8...dr21309fn2.jpg
Using Corsair 6400C'3 @ 5-5-5-9
I got this board for the SLI support but for an overclocker this board is horrible, even for mild overclocks. Poor effort on Nvidia's part:fact:
Corsair 6400C3's at DDR2-925 (462.5MHz) with 3-3-3-9 timings. Memory voltage at 2.5v. Linked and Sync modes.
http://img167.imageshack.us/img167/6...2925333qn6.jpg
Sierra, do you think this board undervolts the memory just like it undervolts the cpu? Just curious since 2.5 volts through the memory seems like quite a bit to me. Thanks!
Hi Guys, I Got a quick question on the EVGA 680i with a E6600, In the bios there are 2 options in question
CPU CORE
CPU FSB
Which one should i set higher when overclocking ???
Currently i have CPU CORE at 1.425 which shows 1.4v and i have the CPU FSB at auto whichs shows 1.19v, the reason i ask is that these processor was on the 975x/g running at 375fsb at 1.4 volts orthos stable for 36hours. on this board the max i get is 350fsb (1400 divided by 4 =350x9=3150) i have it on unlinked so i'm wondering if i am raiseing the wrong voltage, and yes i put it back in the 975x/g to make sure that the settings i had and it goes right to windows at 375 stable Thanks for your help :)
For cooling i'm using the CoolIt Unit
So, it looks like the 6400C3's turned out to be some good sticks after all. :D When you first started posting some results with the Team memory I was confused b/c I thought the highly binned C3's were superior.Quote:
Originally Posted by sierra_bound
Thanks again for posting your findings as I found your result are reproducible. This has saved me hours in tweaking. While on air (at least until my new block arrives) my rig is running nice and smooth at 1440fsb (3.6GHz@1.4625v - in bois/1.42v in Windows) and 1200MHz (DDR2 600) 4-4-4-9 2.5v prime stable. I was surprised to find the memory to be so cool to the touch at 2.5v.
Anyway, nice work with the 680i board. I think it is a winner. :banana:
Hi, i want to buy a 680i MoBo, but i dont know what vendor? eVGA or Asus Striker? What is maximum Vcore on eVGA?
I have not measured voltages yet with a DMM, so I don't know. What I do know is that this board is great for memory overclocking. It's definitely the board's strongest asset.Quote:
Originally Posted by vmsein
Quote:
Originally Posted by OBR
The Striker is a nice board however; it is still based upon the 680i chipset and has not done any better than the reference Foxcon clones. eVGA is a good choice given the advance warranty replacement and availability.
Oh...did I mention price? You stand to save some cash with the eVGA as well. If the Striker were performing leaps and bounds above the reference boards (and it were $200 cheaper) it would be an easy choice.
So...BFG is you live near a Comp USA if not perhaps an eVGA if you live near a Fry's. Club IT is selling the eVGA for $249 IIRC.
edit: 1.850v in bios is max vcore w/ the P20 bios.
If I remember right,when I mesured @2.5v set in bios,the DMM was reading 2.49v inside the dimm slot.Pretty nice.Quote:
Originally Posted by sierra_bound
EDIT:I am talking for the EVGA board.
Orthos @ DDR2-930 3-3-3-9
http://img242.imageshack.us/img242/5...2930333ii0.jpg
Ignore the date. I forgot to change it after I flashed to newer bios.:D
This board is a mystery. With a Xeon 3060 (E6600), 32M (not to mention dual 32M) results are horrendous. I could also see other folks getting around 14 ~15 mins at CPU frequency 3.6~3.8GHz. Switching to E6400 got me 13 mins 48 secs. Sandra bandwidth is even more ridiculous. :confused:
Maybe BIOS isn't tuned to 4MB cache chips yet?
http://img294.imageshack.us/img294/7...msandrarz3.png
This board is the worst overclocker when it comes to memory OC.
How do you get your ram on this board to boot at 500mhz 4-4-4-8 2.5 VOLT, my ram is stable at 555mhz on P5B Dlx 2.40v, so if this problem isn't fixed i will just send it back to Evga for an RMA as it has the Sound issue when i try to use a X-Fi sound card.
Thanks for your answer, but "Striker" have better board layout, no extra Molex power connector, clean CPU area with no capacitors ... but Striker is limited in Vcore only to 1.6V and it is big problem .... i will buy EVGA and i will see.Quote:
Originally Posted by ShoNuff
PS. - it is important to connect that Molex power connector on EVGA board?
Can anyone give me some pics where I find the messuring points for the DMM?Quote:
Originally Posted by Lestat
I would like to know what this molex connector is for as well. I haven't had mine plugged in all this time, but I wonder if it would help...Quote:
Originally Posted by OBR
Also, when overclocking, should the mcp and spp voltages be left at stock or raised to 1.5v? My fsb is 1134 and mem is running at 850mhz.
i have mine plugged in even though i am not running sli.
hmm my my there's a whole range of results.. lopri.. i hope my xeon 3060 will get here soon and then i can confirm ur findings~
also i think many of us raise voltages to 1.5V or 1.4V. that includes fsb, nforce and aux accd to ntune~ think ht will be abt 1.25V
How do you think my G.skill 2x1GB ddr800 "HZ" ram will do with the 680i chipset? Currently with my DS3 im @ 462FSB using 4/4/4/12 timings @ 2.2v. But with the damn D9 bug i cant go any higher.... Im really interested in buying the EVGA version, but really need some input on how this ram might react with this motherboard. Thank you
__________________
i thought coolarer showed a pic.. with vcore up to 1,9 on the striker .Quote:
Originally Posted by OBR
The molex connector on the eVGA board is only to be used if you are running an SLI set up according to the manual.Quote:
Originally Posted by OBR
edit: I would caution you against the EVGA board ATM. The more problems EVGA is having with this board the better the Striker looks.
So, will the Storm water block fit on the eVGA board? I take it there are NO issues of CPU socket clearance with the Striker board?
I want to build a PC that can SLI without hack drivers.
I'm looking to game with a 30" LCD and that means having the graphics horsepower to run 2500x1600 and not worrying about frame rates. EVER. SO a pair of 8800GTX in SLI with a 6600 running at say - 3300MHz on up should get me there. Right?
As long as any of these 680i boards can take a Storm water block. Because water cooled - it shall be.
Quote:
Originally Posted by icon57
same here. figured can't hurt. ;)
How hot is it safe for the mcp to get?
Here's my temps after a few hours of benching and gaming:
http://xs209.xs.to/xs209/06472/hotmcp.jpg.xs.jpg
Here are my latest results on using the EVGA N68..
My max stable overclock is so far is 3520. Boots into Windows just fine at 3600 but not stable. Crashes in Orthos and jumpy mouse running idle. Even at 1.60 Vcore.
1.52V is just fine at 3520.... Orthos stable. My RAM MUSHKIN 6400 is running in sync mode just under 800 5-5-5-18 (1T).
I tried unlinked at upto 820 for the RAM with little or no performance difference.
The last Intel I overclocked was a 300 Celeron so I am not very familiar with Intel ... Anyone know what else I should be doing to get it stable at 3600 +:slapass:
My RAM wont overclock either. I have no idea what it can do or not do being my 1st Nvidea board... Even with high voltage and lower timings it just will not boot..Quote:
Originally Posted by X.T.R.E.M.E_ICE
Bios P20
If you are using a PS/2 mouse, switch to a USB one.;) Running DDR2-800 at 1T is not easy with some CPU's and this board. You may be better off running 2T with higher memory speed.Quote:
Originally Posted by Badge56
Measured VDIMM with a DMM. Overvolts 0.05~0.06V.
my asus 680i
fsb517
http://home.pchome.com.tw/club/qk4723/680i/5146.PNG
:slobber: :slobber: :slobber: :slobber: :toast: :banana:
:clap: :clap: :clap: :banana: :toast: :woot:
@qk4723Deluxe Is that a retail board? Looking good m8 :) Nice FSB!
http://www.newegg.com/Product/Produc...82E16813131074
is that for real $$$$?
My BFG is doing some screwy stuff with memory and FSB readings when using lower than default CPU multipliers.
This was the maximum bootable FSB that I could get from early testing. You can see the differences between the different reporting programs. I think that only being able to boot at 400FSB is RAM related as I am nearly certain that the BIOS estimate of DDR800 is wrong and the true value is DDR1200.
http://slidog.bigblog.com.au/data/2/...1121230057.jpg
Same boot up within windows. The CPU-Z values are what was set in BIOS though only nvtune reports ram values at this time.
http://slidog.bigblog.com.au/data/2/...1121230057.png
BFG has a new BIOS available on their website which I'm going to try.
Is this mcp temp safe as long as the system is stable?Quote:
Originally Posted by Celcius
If the voltage is set to AUTO then I guess so.Quote:
Originally Posted by Celcius
You know the 2-digit display on the side of the mobo that says "FF"?
Is it possible to change it to display the cpu temp?
Can someone please answer this question: I'm getting an E6600 to OC to see if I can beat my E6300 (currently @ 3.23GHz) However, im currently using a DS3 motherboard, and this D9 issue is pissing me off. I have funds to buy the EVGA 680i motherboard, should I do it? I have 2x1GB sticks of g.skill DDR800 HZ ram, and with any hope, do you think this combo will be good for the 680i board? I know its not that great with the DS3.. Thank you
Id stay with the DS3 for now. The replies in this thread have not impressed me in the least. Maybe the 680i will improve with future bios revisions, but the DS3/D9 issue has been improved greatly with the F8i revision.Quote:
Originally Posted by newls1
Loving this board(EVGA680i). Sure it has its quirks as almost every single new motherboard has with early bios releases. You have to appreciate that this is a new platform and not a revised version of intels chipset. I can't understand people moaning about the 680i, expecting it to perform miracles and suddenly turn an average cpu into a WR beater:rolleyes:. Its expensive sure (especially here in the UK) but compared to the price of a W64/WS2 from Asus it aint that bad. The bios is a little tricky to get to grips with but you don't need to be an overclocking guru to work things out and once you do its actually easier to get good clocks than its 975/965 counterparts due to the memory being able to run unlinked. Those who worried about the EVGA bios support need worry no more as it seems that EVGA actually listen to there customers and solve problems with bios releases quickly which is more than can be said with ASUS. The XFI issue has now been solved (beta bios update) and so has the ntune problem (software fault not hardware). I'm not trying to justify the ££'s spend on this board it just gets under my skin when it gets bashed in such at such an early stage. Take a look at the p5b/p5w problems and fixes thread yet those two boards have turned out to be great and havn't nearly recieved the amount of crap thrown at the 680i . /rant
Comparing it to my P5WDH, heres what I have found :
Boot time : 680i by a long way
Gaming : 680i
3dmark : 680i
FSB : 680i
Memory : 680i
Overall Overclock : 680i (by 100mhz so far 3.7vs3.8 orthos stable)
SuperPI : P5WDH (beats 680i in 1m by about 0.5 secs @3.7-possibly related the the tighter memory subtimings of the P5WDH which are not available).
CMOS reset : 680i - simply becuase its now available without having to remove the graphics card!
Looks : 680i (imo):D
However there are a few problems I have encountered with the board :
CPU temp calibration seems a little low (show me a motherboard that is accurate!).
The chipset fan(optional) is very noisy on 100% setting (near silent on 50%).
Not all the onboard diagnostics are sufficiently explained in the manual.
Fitting of aftermarket coolers on the back of the board can be a pain due to the smd's/solder. I used some foam backing on a H-plate to counter this but those with solid plastic plates may find they need to modify them slightly to get them to fit without hitting the smd's.
So far I've managed to get 3.8 semi stable (4 hrs orthos) with this board with the same setup as I had with the P5WDH which despite having a mature bios wouldn't boot past 3.75 nevermind run orthos for any decent length of time. I put this down simply to the better memory overclocking abilities of the 680i which asus seemed reluctant or unable to incorporate. The abilities of this board/chipset have some great possibilities and with new bios releases can get even better (hopefully).
But what are the benefits of better memory ocing if it doesn't translate to better performance.
the evga board sux at memory overclocking. P5B is better.Quote:
Originally Posted by MrDeeds
Because it allows a higher FSB to be attained = more cpu clocks. If you're comparing clock for clock performance via SuperPI then yeah it is slower in 1m but thats down to the differing chipsets imo, sure it would be nice to be hitting low 1m PI times to give an extra inch to the e-penis but as the tests get longer the scores get closer and the evga actually betters the P5WDH over 8m, see here : http://sg.vr-zone.com/?i=4215&s=11. SP has no 'real world' value anyway imo. With the P5WDH I couldn't get my cpu over 3.7 orthos stable in ANY of the tests. Now with the evga I can reach 3.8 (not tried higher yet). This was probably down to the tight memory subtimings mentioned in other threads that held back the P5WDH. I'm not bashing the P5WDH, far from it because it is a great board but it seems there are many who are bashing the 680i without actually having used it or have simply forgotten what problems the intel boards had when they were first released.Quote:
Originally Posted by MrDeeds
see what I mean?Quote:
Originally Posted by X.T.R.E.M.E_ICE
im with iceman....this evag board is way better than my asus ws pro.
...too bad i keep corrupting my os by pushing my mems too high...3rd install in 2 weeks...lol