Tnx for the nice wishes :) but i dont have that much to spend on a cpu.. :/Quote:
Originally Posted by borntowin
Just got myself a surround system which btw kickes ass :).. so i can only afford about 200/300 euros ....
>_<
Printable View
Tnx for the nice wishes :) but i dont have that much to spend on a cpu.. :/Quote:
Originally Posted by borntowin
Just got myself a surround system which btw kickes ass :).. so i can only afford about 200/300 euros ....
>_<
It's not a big difference, maybe the current ambient temperature affected the results or something else (for 21*217-1.375v with 0011 and 1.39v with 1102), but also I have higher CPU score in Vantage with 0011.Quote:
Originally Posted by donmarkoni
I tested several BIOS and every time I go back to 0011 :).
Found the source of my GPU driver crashing. Took my machine all the way apart last night to find that my R3E has damaged my second GTX 275:
http://sphotos.ak.fbcdn.net/hphotos-...3_847279_n.jpg
I ran these cards in my 780i P5N-T, and my 790i Striker 2 Extreme.
Now I'm worried that the pci-e socket itself is bad. Time to RMA both the GPU and motherboard? Mind you, I've NEVER overvolted my PCI-E, and I have one of the EZPLUG connectors hooked up.
^
Were you running your signature overclock before this happened? What voltages were being used across the board (VCore, CPU PLL, QPI, DRAM, IOH)?
You've never overvolted PCIE, but how about overclocking? Always set at 100MHz or ever go over?
I can't remember which thread it was but basically a couple months back there was a discussion regarding the EZ Plugs and I believe it was concluded that they should be necessary only if you are running two or more power hungry GPUs e.g. GTX 480s. And even then I believe it was only necessary if the multiple 480s were also being heavily overclocked. So for a couple of 275s, although still power hungry but not as much as 480s of course, you may be better off not using the EZ Plugs. Maybe in your case it may have been more detrimental than beneficial :confused:
I hope you get everything sorted out. Starting to feel worried about my setup :(
@Fire4Wire
A few weeks back you mentioned there would be an updated version of CPU-Z which supports ROG boards and displays their core voltage. Any news or updates on this? Is it possible for you to share the revised 1.56 version you have which does show core voltage?
Also what are your thoughts on the QPI PLL dilemma? Can you help us out in getting the table of values for this setting from Asus? A bios with a QPI PLL setting would be the first actual solid improvement in R3E bios history.
@Nickless, donmarkoni
Thank you very much for your informative comparison between the different bioses. With so many versions available and not enough time to test them all it really helps when someone shares such valuable information.
I just made the switch to 0011 from 0901 and it definitely 'feels' a lot better. Also it seems as if temps dropped a couple degrees across the board as well. I don't like having to go with a beta bios over an official one but who cares :p:. Thanks guys for your input much appreciated :up:
Yes, that's the settings I was running.Quote:
Originally Posted by spoolindsm127
vcore: 1.3875 max, CPU PLL: 1.8125, QPI 1.5125, DRAM 1.656, IOH 1.1660
I did *try* 102, 104, and 107. Nothing for longer than a single loop of LinX. It always resulted in more instability.Quote:
Originally Posted by spoolindsm127
I saw that thread too. I actually connected the ez plug *after* this issue started happening, thinking it might have been a power issue.Quote:
Originally Posted by spoolindsm127
Honestly, I've had really bad experiences with my last 3 asus motherboards. P5N-T Deluxe, Striker 2 NSE, and now R3E.Quote:
Originally Posted by spoolindsm127
The P5N-T literally shot a fireball at me and blew up my psu and 3 hdd's. It took asus three RMA rounds to get me a working P5N-T, which I then sold off.
My Striker 2 NSE's killed themselves, twice... Literally wake up and my board is just flat out dead(and it was on the night before). Then Asus sent me a fu*ked up replacement, which I sent back to them the same day, it was so obviously broken(damaged traces, visible water spots, missing headers).
And now my R3E has killed my GPU. I'm going to send the board back, as well as the GPU, since there's no way to tell which was the failing component without risking damage to other parts. I highly suspect it was the motherboard since I've had both of these gtx275's for well over a year, and I've taken them across the P5N-T and the S2NSE, and they were perfectly fine then.
It's such a shame that no other motherboard manufacturer can make nice layouts like ASUS. Although Gigabyte and EVGA are catching up. I've always gone with ASUS after looking at the boards, and basically seeing that the ASUS layouts were way better/easier to work with. I'm really thinking about putting ASUS on my $#!+ list and never buying another product of theirs.
@zoson
Your voltages seem fine for your setup, nothing "extreme".
I have heard people running PCIE up to 110MHz so I don't believe that is the cause. Did you experience problems before raising PCIE freq.? I'm assuming you raised PCIE freq. to resolve the crash issue right?
Same thing with the EZ Plug, can't be that since you had GPU problems before connecting it.
I'm leaning towards this being a power issue seeing what happened to your video card. I've read somewhere that if the ATX12V or PCIE power cables have tension at the connector (as in when people twist and bend cables to route them a certain way in their case) this can cause improper contact and increased resistance on certain pins in the plug, causing physical damage from severe heat.
I don't know man I feel bad you've had such bad luck with Asus. I build PCs once a year at most, so maybe I don't buy enough hardware to experience the failing rate.
I've realized that lately, quality control on motherboards has gone WAY downhill. It's almost expected nowadays for me to return a mobo at least once before getting a fully functional one. But surprisingly enough my first (and current) R3E board performed well right from the start. So far it's been holding up well.
BTW in relation to bioses, compared to 0901 which required 1.37250v QPI to stabilize my signature overclock, bios 0011 needs only 1.36250v :D.
Clearly some folks don't have the best luck with hardware. Comparatively my experience has been quite good. I'm looking at 4 motherboards right now 2 new and in use and 2 old, 3 ASUS and 1 XFX all have and still perform flawlessly. As far as i can remember the most common reasons for my previously owned board's demise, simply because they were outdated or failed due operator induced meltdown.. :( But hey I had fun! The only ASUS board I RMA'd was due to the famous non back flash-able soldered in cmos chips.. But I have to say before flashing to a newer BIOS (clearly was a bad idea) the board performed great.
Seems to me that all manufactures have a Lemon from time to time, the R3E definitely does not fall into that category! I've never been more stoked with a motherboard :) My only issues are with cooling and a GPU which I'm working on. the R3E has amazed me !!
If you work at it, any board can be made to under perform ;)
I went to micro center last weekend to get a cheap 950, and while I was there looked at motherboards. I was very tempted to grab another RIIIE. I keep wondering if I am doing something wrong to not be able to reach anything above 215bclk in fast mode, seeing that many people have been able to do much better, and because this is an Extreme board. Instead though I grabbed the X Power MSI board, mainly because of price and the layout, and it does 220 easy, although it requires some tricks to get any higher, max bclk is 233 fast mode. On 3 X58 boards I have been able to reach above 230, so I find it hard now to think its an operator error. I would try to RMA this R3E now but I broke one of the white clips off on one memory slots that lock the memory in. It still works fine and the stick of memory stays in tight, and reads 6gb in windows. I like the looks of the board and love the layout, it would be nice to be able to use it again. But I have a few X 58's and with a new platform comming soon it wouldnt make sense to buy another.
Operator error is in reference to the times for example when through the application of dangerously high voltage a MOSFET burned up leaving a hole in the motherboard where it previously existed !! Or as in the case above when the memory retainer broke off !! I hate that when that happens..
Gary I just want a simple request, I don't need my CPU to do 10 million jigahertz, I would just like the board to post at all with a public bios with xpander ;)
Some people take stuff for granted, like the ability to display video at all.
http://chew.ln2cooling.com/qdig-file...g_IMG_3168.jpg
^ +1Quote:
Originally Posted by spoolindsm127
You won't get it due to using a MSI card Brian :p Shame on you :rofl:Quote:
Originally Posted by chew*
Update:
I tested the 'suspect' card alone in the second pcix slot on my R3E(originally it was in slot 3), to my surprise it went through stress for 12 hours with no issues!
However, when I run SLI with my cards in slots 1 and 2 the GPU crash still happens after about an hour, instead of under 5 minutes.
It looks like there's just something wrong with my board...
Sadly my R3E will be going back to Frys this weekend, I've had it all of 4 days. Thing wouldn't even boot at 210BCLK without Slow Mode. The $89 open box R2E I bought can at least do that (same CPU, memory, PSU, video card). Very strange board.
did you up the ioh voltages to
1.26
1.6
set qpi jumper?
if you did yours is more pityfull than mine was:shakes:
Quote:
Originally Posted by Shiranui Gen-An
QPI jumper was set, I bumped IOH vcore as high as 1.40v and I think I did try IOH PCI-E 1.6v, no dice.
Hi guys...
I am having a bit of trouble running an Xeon X5570 in this board...
After a few successful hours of testing the cpu is only being recognised as 1 core 1 thread when in OS which sucks.
I seem to remember a few having this problem on some boards with a 980x so maybe this is related?
Tried with 1102 bios and 601 bios with same result.
Bios is set as all cores on and HT on before you ask :D
Bios reflashed and defaults loaded several times with same result.
I7 920 detected fine as 4 c 8 t
X5570 is known good and detects fine on other boards....
Going to put in X5570 after 920 just for kicks to see if that rectifies the issue...
ps I would like to test 1203 on this issue ;)
Edit: is anyone else running Xeon X55xx or X56xx series in this board without issue?
@Biker
Looks like a Windows issue.
Go to: Start/Run/MSConfig/Boot/Advanced options/Number of processors, is it checked, and if so what is set?
i wonder how many are like thatQuote:
Originally Posted by Shiranui Gen-An
most of the overclocks i see in this thread is 4.0-4.2
i was gonna buy lowfat's rexIII but it was only good for 228:(
my e760 has no problems @ 21x231blk:D but boy does it get hot withought the a/c lol
I could do 4.2GHz as 21x200BCLK.
It was unchecked / greyed so tried to force it with 4 and 8 with no success.Quote:
Originally Posted by donmarkoni
This does not explain how an i7 920 is detected fine but the 5570 is 1c 1t every time...
I doubt this is OS it looks like hardware / bios to me...
@Biker
I'm currently running an X5690 ES @4,35GHz on my board with no problems, bios 1005.
"Resistance is Futile"Quote:
Originally Posted by Leeghoofd
Seriously I would have used the 4th ASUS card for far less compatibility headaches if some idiot had not killed it ( idiot being me ).
After finally trouble shooting some interesting issues I managed to get this run off.
Gary you have mail........
http://chew.ln2cooling.com/qdig-file...%20vantage.jpg
Hi guys,
The Rampage 3 extreme has the 2 marvell SATA 6 GB/s connectors and then 6 x sata 2 connectors.
I will set up a raid 0 with 2 SSD's and am wondering if I will benefit from the 2 6GB/s ports or would it be the same if I use the 3 GB/s ones? Is this an option?
If so, can I than use the 2 marvel ports to run with Sata3 disks in raid 1? or what ever? ..Thanks
I found some differences after Sep. production REX3.
The difference is CPU socket. (FOXCONN to LOTES)
And maybe improve BCLK wall problems.
I get 2 REX3 and both are over BCLK 245MHz without slow mode.
Good one is over 250MHz.:eek:
(My 980X BCLK limit is around 250MHz, so it's not mobo limit, I think.)
But PCB rev does not change, it's still 1.04G. (CPU-Z rev is still 1.xx)
http://img43.imageshack.us/img43/5403/cpu3001.jpg
arent they all lotes :confused:
The Marvell ports are slower than the Intel ones with two SSDs in RAID 0.Quote:
Originally Posted by buccaneer
for the sound chipset of the R3E, which driver you use?
thanks ;)
realtek r2.55
ok thanks ;)
EDIT: I integrate the drivers to the ISO of Windows 7, which file I select?
there is too much .inf file, and I do not know which one is good
Thanks CX-RAY. tel me; how about normal sata 3 on the marvel?
The Marvell SATA III is a bit of a trade-off compared to the ICH10R SATA II. With a single Crucial C300 256GB running AS SSD benchmark, the sequential reads are about 80MB/s faster than on the SATA II ports, seq. write about 10MB/s slower, with 4K write also about 10MB/s slower. The remainder is about the same.
The best driver setup for the SATA III ports is the minimum install of the Marvell driver. This will let the device manager recognize the controller properly. Then, let Windows install the standard Microsoft msahci driver.
Nope, this crappy R3E I'm about to take back to Frys uses Foxconn.Quote:
Originally Posted by dejanbinladen
could you post a pciture please
Well it's all boxed up ready to go now. There are no markings on the socket retention mechanism itself, but rather on the backplate where "Foxconn" is clearly printed. Like this.
Right. I should have read it better, but I was tired. So, I'll try harder to help, as an excuse. :DQuote:
Originally Posted by Biker
This is what I would do, except trying other BIOS as Mydog said, which would be the first thing. Go in BIOS and check if enabling/disabling HT changes anything. Then try disabling/enabling cores. Next, you can try enabling 2D Performance Booster to see if it helps, so you'll know that it's a BIOS issue.
My board uses a Foxconn socket. Board rev. 1.04G (1.xx in CPU-Z). BCLK limit 221 w/o Slow Mode.
Since my board booted a 930 accidentally at 4730Mhz (215x22) I really had not pushed any higher due to some cooling issues :( I think it was coincidental that I stopped at 215 bclk, because later attempts to find an upper limit of the bclk in 5Mhz increments I found the board would not post at 216 or even up to 225 :( have yet to try higher. Not looking good, but I've said that before...
Have you tried using the 21 multi to find your max blck? I have a 930 and 920 D0 and have gotten up to 21x228 but couldn't get over 210-212 with the 22 multi on my 930.Quote:
Originally Posted by Ian[Frozen]
Here is my new REX3 photo.Quote:
Originally Posted by dejanbinladen
It's serial number is "A9M0AC20xxxx".
http://img529.imageshack.us/img529/9692/rex3lotes.jpg
I cannot see how a CPU socket would limit the ability of the board to overclock. A lot of boards use the foxconn sockets...
I do not think CPU socket change improve OC ability.
It's just info REX3 has some minor change after Sep. lot.
Maybe it's chgange NB chip X58 to X58E, too.
(I hear some gigabyte mobo changing NB chip from X58 to X58E.
http://www.fudzilla.com/motherboard/...t-remains-x58e)
So it improve BCLK limit after Sep. lot.
Does anyone has newer lot REX3 & check its BCLK limit.
My new one has very improved BCLK limit.
Yes when finding the max bclk I usually set the lowest multiplier of x14, however have also tried x18 & x20 with no joy :(Quote:
Originally Posted by deachus
then you are noob :D NHF you have to use i dont know how to say it on english x15 17 19 21 ...
EDIT ok google help me its "odd" numbers :D
You say you used x14, 18 and 20 but have you used x19 or x21? I have heard alot on these forums that odd numbered multi overclocks better and higher than even numbered multi. Try 21 or 19 if you have not already.Quote:
Originally Posted by Ian[Frozen]
lol go figure the only multipliers not tried may infact be the fix!! lol
Odd numbers kinda go against the grain for me, but will try when home in a few days thanks :)
The socket in my not so good r3e is foxconn.
Which bios version did the board come with?Quote:
Originally Posted by kurumi_Japan
My new board came with 1102 from the factory, so must be latest batch.
My board come with 1005,Quote:
Originally Posted by bito
So I think your board is newer than mine.
well, there was a problem with foxconn sockets on the p55 (socket 1156) platform. hopefully this turns not over to x58 also:Quote:
Originally Posted by kurumi_Japan
lotes sockets are higher quality than foxconn sockets. there might be some misaligned contacts on foxconn sockets preventing them from making full contact between the cpu and the socket. this might result in increased contact resistance as less pins contribute to the total contact and in fact might burn the socket, the processor or both due to higher total contact resistance.
there are some reports of users and reviewers who observed this:
P55 Extreme Overclockers: Check your sockets!
P55-UD6 socket burn
was a bit surprised to see foxconn sockets on the r3e as the usage of lotes sockets was one of the arguments asus used to position the r3e over competitors. may be they run out of lotes sockets and made a batch of boards with foxconn sockets.
anway, from 3 boards i had in hand, two of them used lotes sockes, the third board is still unpacked will check when i'm at home again ...
after all, the described problem was discovered on P55 boards and might be no issue on x58 boards, though it would be interesting to see if there is a correlation between max bclk and socket manufacturer!
regarding x58e: does anybody know how to check if the board is equipped with x58 or x58e? are there any x58e variants of the r3e out at all?
You do understand even on my R2E with the new lotes socket it also says foxxconn? And I'm pretty sure my RE3 says the same too. Check the main socket, and take a picture of it. if the two extended prons/tabs are large, and has a LARGE endent then that means its a LOTES socket, if its only a little endent that its a foxxconn.Quote:
Originally Posted by Shiranui Gen-An
The lotes socket does improve oc ability by a little bit. My old R2E died, first revision and 1st release in Taiwan had a foxxconn, had many problems, asus swapped it out, got a lotes socket, bclk and Oc ability improved a little bit on my 920 C0 one of the 1st few batches.
PS Most boards with LGA1366 Backplates has the marking with Foxxcon, but IT DOES NOT MEAN THE SOCKET ITSELF IS FOXXCON. Check the retention bracket and see if it has LOTES marking, or check the two tabs that hold down part of the chip and headspreader.
My R3E socket has neither foxconn markings on the back or lotes markings on the socket cover...
http://sphotos.ak.fbcdn.net/hphotos-..._3645000_n.jpg
Foxconn & Lotus
My socket looks like neither of those... In fact it sort of looks like a combination of the two...Quote:
Originally Posted by cx-ray
Also my R3E socket looks like none of the above on this or the last page :( On reflection perhaps like zoson, a combination of all three types shown above..
Only pic I have for now is this one taken during a cleaning, shows a lobe similar to zoson's but from what I remember it had LOTES stamped on the cover. If it's a LOTES then clearly the manufacturing process uses inconsistent stamps.. I'll have to inspect it more closely on the next evap removal...
http://db.tt/LafyLWt
So does anybody want to point me to a good guide for working with this board since i moved from evga options completly different. I only have basic Oc knowledge on X58 never really messed with my 920 after i had it where i wanted.
And best bios to use came with 0902 installed 1102. Temps seemed to jump but i was after playing with voltages for ram.
Sent from my X10a using Tapatalk
Hi guys, i got this motherboard assembled with:
Core i7-950
Corsair CMP6GX3M3A1600C7 6Gb 7-8-7-20-2T @ 1.65V
Sapphire Radeon HD5870 Vapor-x
2xSamsung spinpoint F1 raid0
Enermax HX650W
The problem: the mememory kit is sold to be compatible with X.m.p. profile, but if i enable it i can't ever enter in windows. I discovered after various tests that i must set the Vqpu>1.25V and it must be 1.3V to pass linx tests, even lowering the memory voltage@1.5V! Now i have the memories at 7-8-7-20-1T@1.5V but with a Vqpi=1.3V, while the x.m.p. profile tries to set the Vqpi=1.2V.
Is that normal?
P.s. memories passed memtest without problems with Vqpi=1.25V and settings like above.
Thankyou!
I have the dom 2000 kit i can't even run it off default settings without problems. I have no idea what to mess with to fix it.
Sent from my X10a using Tapatalk
You need to start playing with CPU PLL. It defaults to 1.8v but I had to lower mine to 1.42-1.45v depending on CPU and memory speed to get 20 runs LinX stable. Asus has stated QPI PLL is directly tied to CPU PLL but we don't know the values it sets. Also, I was able to get memory stable after switching the QPI Load Calibration jumper on the Board. I suggest you do this if you have not already.Quote:
Originally Posted by MagisD
I'm starting to understand this architecture. Seems like i've gained full stability with Vcpu=1.2V, Vqpi=1,325V and Vram=1.5V (1600Mhz 7-8-7-20-1T). Now i have my cpu @3600Mhz. If i would like to raise my cpu clock raising the max ratio, for example to 19 to obtain 3800Mhz, it's enough to raise (eventually) the Vcpu parameter, isn't it?
Thankyou!
I'll mess with it in a couple of weeks its currently my work comp and i need it up. Need to read more.
Sent from my X10a using Tapatalk
although the board supports 2000mhz speed, it only supports that as an overclock, hence the 2000mhz OC, i have the same kit, all i needed to do to get the kit running at 2000mhz 8-9-8-24 was to raise the qpi voltage to 1.4Quote:
Originally Posted by MagisD
any1 else have benchmark issues over 215 BCLK? mine seems to go upto 219, but the benchmarks get worse (graphically). using 1005 BIOS, any 1 got a template for 220+ w/o issues?
I have been able to bench up to 228 so far but need to get some cold on the CPU to go any higher at this point. I can boot into windows up to 232 but cannot bench. That is without slowmode. I have not tried slowmode yet.Quote:
Originally Posted by xdave
Same here buddy.. get in line :shrug:Quote:
Originally Posted by xdave
In my experience vqpi will need a bump tooQuote:
Originally Posted by DrSlump
For me vQPI needed above 225bclk is in the 1.52v - 1.58v range depending on Uncore, memory frequency and timings.
vQPI 1.45 for 230blck. Havent try more yet.
try with low memory settings.
ok i had many many problems with this board as you can read on the forum but one that took me down is broken pin in the socket, so i ordered new one from ebay LINK and i am hoping to get a new start with this unfinished board
opinions of the socket please ?!?!
any chance you could post your template for 225 bclk?, gives me a bit of a starting point to work from... cheersQuote:
Originally Posted by deachus
How hard is it to replace the socket? I threw away a new P55A-UD7 with bad pens.Quote:
Originally Posted by dejanbinladen
Quote:
Originally Posted by xdave
Sure, the next time I do some benching I will take some BIOS pictures;)
as i saw on youtube not so hard if you have the equipment LINKQuote:
Originally Posted by OC Nub
something like that the machine costs about 1000$
Hi guys whats the best bios to use on this board now.
I have not updated the bios since I got the board and want to start on the OC
thanks
Try the one you are using first, a newer bios isn't always better. I'm still on 0704 for daily usage, not many issues here...
Here's a little test I did last night, BCLK at 230MHz, No Slow Mode. ;) Maybe I could tighten some latencies on RAM, but 2300MHz is good anyway, and it was all about 230 BCLK.
http://img819.imageshack.us/img819/2...oslowspi32.jpg
:clap: Nice Job :clap:
37.000v CPU!
CPUz 1.56 bug shown the wrong voltage for R3E.Quote:
Originally Posted by bassplayer
I have 240+ w/o SLOW mode:yepp:Quote:
BCLK at 230MHz, No Slow Mode
http://img51.imageshack.us/img51/5730/242ky.jpg
http://valid.canardpc.com/show_oc.php?id=1512952
:eek: i got problem with my rampage 3 now... i cant overclock anymore over 4ghz will system crash and i think bios is corrupt all of them...
i have tested several bioses and when i save my oc Profile rampage 3 wont booting and nothing windows 7 will boot either. rampage 3 has 2 bios chips and i switched and tested both of them but same problem there :( OMG !! its really piss me of
Nice. :clap: Inspires me to try harder. :up:Quote:
Originally Posted by Diabolo80
yes realy nice =]
My kit booted up on first try using 2000mhz. around 1.35 QPi should be enough for 3800uncore, and 2000mhz.Quote:
Originally Posted by MagisD
DONT ever use XMP, the QPiV is simply set too low for Anything, stupid corsair..
I tested 1203 BIOS on Air.
Its solve my sleep problem, perfectly.:)
But I think this BIOS does not suitable for LN2 session.
It has CPU temp bug under LN2.
When CPU temp under -100 degree, BIOS reply CPU temp over 250 degree.:eek:
I hope ASUS fix CPU temp bug.
Hi guy's , just wanted to say Hi as I dont post much on this Forum much as I feel a bit out of my depth. lol. I generally just do a lot of research and reading.
Here is my R3E :) O/C so far :
http://img251.imageshack.us/img251/8...ghz138v.th.jpg
I have put 2 x 480's SC in her now and I want to try and O/C her some more to see how far she will go :D
I have the vcore i notch under 1.4v and the Qpi is about 1.35v and PLL is 1.55v and Dram is 1.635v With no LLC enabled ( i just dont like it) for the above O/C .
I have every thing else basically set on Auto.
Can any one sugest some other settings I can tweak in the bios to push it a bit further ??
Thanks dazza
Nice chip man, good temps also.. what is your ambient ?Quote:
Originally Posted by Top_cat2
Dam hot at th mo , mate , Im In Sydney Australia , and the Temps and humidity are creeping up now:( hit about 30c today and very humid :(
I have my Rig in my sig , but its not showing up :confused:
Any how got any tips on what I should do to O/C this little Mutha ..the Bclock will go much further and I want to use some of the functions in the bios but not sure what to touche :shrug:
regs Dazza
I would say your gflops a bit low.. on 4.4 im on 65.8 gflops @ 1.57 Qpi/Dramm.. maybe notch it up a bit see if you get improvement.
You also need to use Maxmem in LinX. I see you only used 2048MB for that LinX run when your running 6GB in Triple Channel. Hit the "ALL" button and then run LinX again. When I hit "ALL" LinX uses between 4800MB - 5048MB. That also might be why your GFLOPS are low.
Hey there, does anyone have any pointers or good starting points on the 980x(my first extreme chip) or just a nice R3E bios template (other than the one on the first page) to start out at I'm all ears.
also what is the max safe long term vcore for the 980-x? I thought I saw some people in the R3E thread saying over 1.4 was highly dangerous where on my old i7-920 over 1.5 was dangerous.
It's depend on the chip what CPU V is safe or not.Quote:
Originally Posted by Xel_NaGa
Air or water cooling, it will be change, too.
The first, you start 4.0GHz stable setting.
I think it's very esay to make it.
And step up 4.2GHz / 4.4GHz.....
This is my opinion, QPI & DRAM V is more dangerous than CPU V.
When I tested new chip, I set QPI V @ 1.45v & check uncore mem limit.
If CPU is not so hot, I step up QPI V 0.025V & check uncore mem clock, again.
If your chip uncore is good, you can use high DRAM & uncore (1:2 or more) setting.
But it's not so good, you had better choice 2:3 DRAM/uncore setting nor 1:2 setting.
Enjoy your new chip.:up:
I'm using water and just by using auto I was able to get 4492 with idle temps under 30.Quote:
Originally Posted by kurumi_Japan
prime 95 workers would crash out... needs tweaking but its boot stable on auto lol which I thought was impressive
http://i48.photobucket.com/albums/f2.../Capture-3.jpg
Quote:
Originally Posted by Xel_NaGa
I think ASUS auto setting is very dangerous because of it will push up voltage too much.:down:
If I start making 4.2GHz overclocking setting, I set as follow voltage & MEM setting.
CPU core V:1.25 to 1.35V
QPI V:1.25 or 1.3V
CPU PLL V:1.80V
DRAM V:1.60 or 1.65V
IOH & ICH core V:1.10V
IOH & ICH PCI-E V:1.50V
BCLK:168MHz
CPU multi:x25
PCI-E Freq:101MHz
DRAM Freq:1680MHz
DRAM timing:7-8-7-20-60-1T
Uncore Freq:2520 to 3360MHz
QPI Link:6048 or 7392MHz
If I can not pass LinX or Prime95, I push up CPU core 0.025V & QPI core 0.0125V.
The first, check your Uncore limit, it will accept high uncore setting or not.
I think uncore setting is key of 980X overclocking.
PSYCHO a.k.a. kurumi
Quote:
Originally Posted by donmarkoni
Quote:
Originally Posted by Diabolo80
Did you guys just get lucky with good quality IOH's or are you using some cooling other than air to reach higher than 226 BCLK with out slow mode?
Which BIOS are you guys using as well?
Donmarkoni shows 0011, not sure what that is :shrug:
I'm having some issues as of last night, I got a weird start up issue that took me a while to trouble shoot and finally nothing short of moving the jumper pin to CLR CMOS AND removing the battery would get the board to turn on...
and this is using settings I've been stable with for a long time.
1005 bios.
CPU cooling by SS (-40c), IOH cooling is standart + 80mm Fan, I use 1005 biosQuote:
Originally Posted by fl0w3n
238bclk, true+delta, room ttemprature over 20, slow mod, 2 cores,dual chanell, my worst 920 sample, bios 0096 (thx nick), validation
http://www.pohrani.com/f/w/cU/4cJ4Q7b/238.jpg
Quote:
Originally Posted by fl0w3n
http://img80.imageshack.us/img80/637...bclk250mhz.jpg
Just Air cooling with 80mm fan on IOH chip.
My favorite BIOS is 9124.:)
My 980X has BCLK 250MHz wall.:(
So maybe my REX3 can pass over BCLK 250MHz w/o slow mode with good chip.
BCLK limit depend on CPU & IOH quality.
So you had better to check other mobo, the first.
If other mobo can pass more BCLK speed, IOH is not so good.
But same BCLK wall, it's CPU limit.
If you want to more BCLK speed, please try as follow setting.
IOH Vcore: 1.40v - 1.46V
ICH Vcore: 1.15v - 1.20V
IOH & ICH PCI-E V: 1.60V
(Please be care of NB & SB chip temp.)
B'regs.:up:
Great thread, thanks guy's.
Does anyone know if it's possible to connect an 8pin 12v from another psu into the second slot? without my REXIII blowing up:D My psu only has one of these connectors.
Quote:
Originally Posted by DFI pit bull
How about change molex to EPS 12V like as follow cable?
http://img52.imageshack.us/img52/5725/molex2eps.jpg
http://www.microcenter.com/single_pr...uct_id=0339351