Asus has its own list of problems btw. :yepp: Best of luck either way. ;)Quote:
Originally Posted by TurboDiv
Printable View
Asus has its own list of problems btw. :yepp: Best of luck either way. ;)Quote:
Originally Posted by TurboDiv
I would need some luck for sure - Exchange policy from NCIX is only 15 days. And i had it for a month now, just awesome.
Great board so far
http://img508.imageshack.us/img508/9576/50sb.jpg:)
I agree ^^ looks v nice my friend :up:Quote:
Originally Posted by jacek
Look here to get the version of CPU-Z that shows the vcore.
http://www.xtremesystems.org/forums/...7&postcount=64
thank you very much for the tip. I was wondering why it shows the wrong voltage. Vcore is higher - about 1.41V, llc - level1
Looks like F6 is the latest bios. Anything new?
They removed PLL overvoltage. AVOID unless your chip can do 5 ghz without it. If it can, then apparently it has some improvements, and LLC level 2 works now, but overvolts a lot. (and level 1 should be toned down some).
Looks like it might be a new official bios, hope they bring int pll back for the beta's.... runs fine though but ye they need to sort the voltages out on this board....Quote:
Originally Posted by OC Maximus
He he, I just gone on the GB site & was like :shocked: thats me!!! famous at last lol
http://www.gigabyte.com/products/main.aspx?s=42
http://gigabytedaily.blogspot.com/20...with-p67a.html
Bit of 2500K action :D
CL7 no faster than CL8 for me, anyone finding the same?
http://i106.photobucket.com/albums/m...227107261T.png
yeh the cas latency doesn't make much difference but if you can lower the trcd it'll fly
Dont think my sticks will do it, they get in windows but need mad voltage.Quote:
Originally Posted by GLENBOY
mine are the same i have the patriot 2400 sticks and mine don't like trcd 9 either they do about half super pi and crap out but at 7-10-7-27 they are rock solid at 2100 and odd
Thats interesting to know I considered a set of those, we need a 2400 multi :D
C-N, i tried F6 but the LLC1 was the same overvolt as F6D (in cpuz).
it's different for you ?
for C7 v C8 32M performance the difference on my XP is very small.. ~0.4-0.5s @ 5ghz 1067mhz. RTL change is 2 clocks per channel
24/7-type OS with services running will vary the run time more than the difference i measured between CL for tweaked runs.
did these on F6D because i need the PLL overvolt for 4800
Linx is on vista, can't test with AVX, the temps seem pretty high (H70 with open setup / 15-18c ambient)
L042B074
2x2GB OCZ3B2000LV2G (hyper)
http://i447.photobucket.com/albums/q...21_OCZ_F6D.png
and my current 32M limit (can't get below 6m04 on 3 different multis)
1.520v bios set
http://i447.photobucket.com/albums/q...79_OCZ_F6D.png
I thought level 1 was slightly better maybe I didnt test it properly. Level 2 worked but my rig rebooted twice under load when it got a couple voltage spikes.... too risky to monkey about with.Quote:
Originally Posted by cheapseats
Ye same here, nothing in it. That run wasnt optimised I dont really do Pi but thought it would be good to compair like for like timings all be it CL7 & 8.Quote:
Originally Posted by cheapseats
Nice sticks & running well, what was the vdimm?Quote:
Originally Posted by cheapseats
yep thats what we need :up:, asus have the 2400 multi so no reason why gigabyte can't implementQuote:
Originally Posted by C-N
Playing BCLK tonight, straight from BIOS :up:
DDR2289 7.10.7 1T
http://i106.photobucket.com/albums/m...228971071T.png
DDR2000 6.9.6 1T
http://i106.photobucket.com/albums/m...R20006961T.png
Very nice C-N. 107 is excellent and bios clocking is much safer on your op then EasyTune.:up:
Nice work C-N
Got my second chip in.
2600k #2
Batch L041C106
# 0815
http://i464.photobucket.com/albums/r...HEMI/2600k.png
Maxes out at 5.34ghz and max bclk of 105.5
Looks like I'm gonna need a third one.
Why was Pll Override removed in Official Bios F6?
There is a suggestion that PLL override can cause instability in 24/7 systems. If you go look at the HWBOT forums you will see some discussion on that in the PLL and bios thread.
It will go a bit more but the chip wont run on the x55 multi with high BCLK.... well any BCLK actually 100 is it. It might go a tad more on cold but air is a no go.Quote:
Originally Posted by flesheatinvirus
5.3GHz is still bustin mate.... good game this aint it :DQuote:
Originally Posted by OC Maximus
Like 1Day said (something to do with S3 etc) but it looks like F6 will if not already be the official bios, hope we get it back in the beta's.Quote:
Originally Posted by OC Maximus
Worth a read.... :up: GB
http://gigabytedaily.blogspot.com
Oh yeah I got my notice from the egg stating the new replacement boards should be out sometime in April.
Who has switched to using the 6Gbps sata ports?
I think the 6Gbps ports are native to the chipset this time?
The standard chipset drivers enable the ports?
I think we can adjust ahci versus ide for these so this problem shouldn't be a big issue for most oc'ers?
Only if you want to use more than 2 sata ports then right?
Maybe just for folks that want to bench for pcmark i guess? :(
Good to know GB will look after us whether we get a problem or not. We'll likely be on something else by then anyway!Quote:
Originally Posted by flesheatinvirus
Don’t know about you guys but I like to tool with a board for a bit open bench then case it up & run it 24/7 for a few months before deciding to either use it to replace my work 24/7 rig or update my regular 24/7 rig. My regular home rig runs lots of drives in RAID & has all my important data on so if I'm going to keep the board I would ask for it to be reworked or replaced if necessary regardless of whether I see a problem or not.
I vote GB gives us all Areca 1880/4GB controllers for replacement!
He he ye think I would accept that :DQuote:
Originally Posted by SteveRo
Hi Steve :)Quote:
Originally Posted by SteveRo
It's native to the chipset, I've been playing a bit with the onboard ports and the main difference from previous chipsets is > 1GB throughput.
Where do I sign up :)Quote:
Originally Posted by SteveRo
Quote:
Originally Posted by C-N
nice 1145mhz :up: with the lowered vtt and vSA..
testing my gtx2 but it's much harder going than ocz @1067 7-7 or 7-8
tried it with your 0.9v vtt and 0.815v vSA setting and it now passes 32m instead of freezing loop 1
thanks for the idea ! :D
btw 1093mhz 7-7-6 on ocz was 1.82v bios set
Cool, worth posting my voltages then as thats exactly what I found right from the first day I got the board :up:Quote:
Originally Posted by cheapseats
I went right through all timing options & speeds last night & with my PSC (perhaps not the best) I cant find anything much better than anything else bandwidth, latency or Pi time wise. The only nice thing is my PSC can run 8.10.8 or 7.10.7 2133 at 1.6v & still perform very well.
I have some Hypers (not used them yet) & from what I've seen about they have the edge on PSC the only downside is they like the vdimm to do it.
I'm using all 4 HDD drives on the 4 SATA2 ports.
I'll let you know if something burns up. I don't mind...
(of course I'll do the RMA if they/newegg cross ship....)
Just got my 2nd HD6950 (1st on artifacted at stock) & flashed it with the 6970 bios from TPU & guess what it artifacts as well!!! Figured I'd try the ASUS overclock bios 890/1375 & it runs fine!!! something fishy here.
Anyway looks like I got it going ok question is how does 31600 look in 3DMark06 890/1375 5GHz DDR2133 7.10.7? Anyone got a similar setup?
You are absolutely right , anyone say Ivy. They are solid in replacement though just in case. Keep the board for now and beat it like a red headed step child for a month or two then let them give another for your trouble. My board actually works great and I have yet another on its way.:D I have two bench/play only rigs so data loss is nothing.Quote:
Originally Posted by C-N
Another UD5?Quote:
Originally Posted by flesheatinvirus
I love my board & would rather just have it re worked if that's an option.
HD6950 is a different animal with the ASUS OC Bios.... just maxed out Catalyst
3DMark06
http://i106.photobucket.com/albums/m...k069501475.png
3DMark Vanatge
http://i106.photobucket.com/albums/m...arkVantage.png
Yes another UD5 from the first board I had to rma. I bought another before even sending that one out, Yes I am crazy. They would give you that option though the turn around time would greatly increase. I am going to wait until they finalize design change then hit em for a fresh board keeping the one coming in for research. Very nice core clock on that 6950, 06 score looks good as well.
No prob I got over 10 other boards/rigs & I'll probably have another few by then anyway so wont miss it for a few weeks.Quote:
Originally Posted by flesheatinvirus
Great thats all I'm after... as long as its inline with everyone else I'm a happy man :up:Quote:
Originally Posted by flesheatinvirus
Exactly, I have so many different systems here its simply retarded. There is another 2600K in here somewhere unopened I need to test. 3DM06 is actually perfect for quick testing Sandy chips being very cpu intensive. 30K plus is a serious score though crank it up to 5400 or more and watch what happens. That 6950 should scale rather well.
I wonder how BIOS support is gonna be over the next couple months. I'd like continued support, but I think they could also use the lag to get their UEFI toether for the next rev.
No doubt, just one more thing to focus on.
nothing changes with bios support, we're still going to keep going the same way before intel made those announcementsQuote:
Originally Posted by mikead_99
Dinos is gigabyte gonna bring out EFI yet like asus boards? I know the giga board are partially EFI with 3tb drives allowed etc but are they going to make the full switch??
Hello every one. My questions is this should I buy this version UD5 or the UD7 version?
Both great boards.. Officially i like the UD5/Ud4 better but seeming as Intel has issues from the callback i'd go the UD7 for the 6x extra sata III portsQuote:
Originally Posted by truehighroller
Thanks I appreciate it, I was thinking the power phases are better too and maybe I'll get lucky and get a better binned chipset for the extra money, maybe..
Lovin' my UD5. Don't see any need for me to step it up to UD7 unless they start getting better bios support since they are flagship at the moment.
Regained my 5770 Gold today in 3D01, w00t.
Now where is my x57 capable 2600k, lmao.
http://hwbot.org/community/submissio...0_108141_marks
Good to hear, thanks for the news Dinos.Quote:
Originally Posted by dinos22
The giga power phases are way over rated for me & the average joe so dont worry about them. These new boards & SB are not like the old boards where better chipsets got you more fsb so don’t worry about that either. What I would say is the difference in cost between UD5 & 7 is small imo & if you need the sata ports that would be enough of an excuse for me to spend the extra money.Quote:
Originally Posted by truehighroller
I love my UD5, its a flippin awesome board it easily maxes my CPU, RAM & VGA out effortlessly.
Well, I have 3 320Gb drives in Raid 0 so I'm thinking I probably should from what you're saying about the issue. Do the six extra ports on the UD7 not get affected by the bug?Quote:
Originally Posted by C-N
Got some Gskills on the way. My non-compliant Hypers will be taking a back seat for now.
Thanks for the info flesheatinvirus ;)
UD7 otw:D
:fight: :DQuote:
Originally Posted by Dumo
They both great boards that clock alike so imo the only thing that can split them are features, if you need them you need them if you dont you dont....
is turbo working properly on UD5 boards? I posted in the UD4 thread about the UD4 not working right, just wondering here... for example, 46x/47x/48x/48x (4/3/2/1) and it never goes past 46 even with power draw unlocked.
so I'm not sure about the UD5 or UD7 but the UD4 seems real buggy, and people are complaining about a lot of quirks.
and what about individual core clock control? I heard some other brand boards can control each core individually... like one weak core you could have at 4.9 and the rest at 5.0 if that one core is holding you back. or am i misinformed?
I'm asking, because I may upgrade to the UD5 when It comes time to replace my UD4. that is, if microcenter will still let me get a refund in april or exchange and pay the difference for a UD5. (don't want to wait 2 months without a computer...)
Thanks :up:
Can’t 100% say as I only used turbo a couple times just to see what it did. It works in as much as the auto ratio gives you the multi you set so yes if you set 50 & leave it all auto you get 50. I know the power control thingy’s work as well because if you set them to say 100w the cpu will throttle the multi back under load as its not getting enough juice to run.Quote:
Originally Posted by LeetMiniWheat
I don’t think 46/47/48 etc work, I think it does lower down in the 30's but I think from memory you cant run one core at 50 & others lower...
First 3 posts updated with useful info & results, When it fills up a bit more I'll tidy up & just show your best score & limit the results to top 10 :up:
Get your backups posted :D
If I've missed anything or if you have some tips that may help others say so & I'll update/fix it when I get a sec.
======
Got my rig all cased up & CPU under water now, not very impressed with temps tbh but will try a few re-seats tonight & see if It can do better.
Got the 6950 up to 970 x 1500 as well... just for benching when I get round to it.
Guys I got a curly one over here, My 2500K is solid @ x55 x 1000 bios bclk yet 1001 wont boot, locks the board up & my Vertex has lost the ability to be detected by bios 3 times as a result. Its like the bios has no table to allow a 2500K to go faster!!
Has anyone got there 2500K past x55 1000?
http://i106.photobucket.com/albums/m...s/54876MHz.png
http://valid.canardpc.com/cache/banner/1643110.png
That is interesting, could be just that chips limit with bclk though I would think that is rather uncommon. What current bios you running? I have seen that with my force drive a couple times a while back where it would read the drive in crazy code then after rebooting with optimal settings come up normal.
Yes its funny how it stops... never seen anything like it, if the chip is maxed & wont boot anything more so be it but as it run's stable at 1000 & wont boot 1001 suggests another issue to me. Another interesting thing is if I boot at 1000bclk get into windows & up bclk even .01 (one click) in ET the multi drops back to x50 from x55!!! I have tryed lower multi & higher bclk & higher multi & lower bclk same result each way.Quote:
Originally Posted by flesheatinvirus
I am on F6d & this chip can make 107+ bclck on lower multi's
That just answered my other questions. Solid bclk with lower multi's is good, f6d is good as well but having the multi drop down on bclk increase within windows is certainly not good. With my current chip at high multi near limit will certainly limit my bclk increase within windows but it is usually a hard lock resulting in a reboot with no change in multi setting if bios untouched. Will this condition occur with any bios?
My chip still boots from bios to windows @ 5460MHz with 107x51 & still runs 32M on the 2133 multi it just needs 1.28v PCH. A lockup/BSOD etc would be kind of what I would expect if the chip was maxed but something else is happening in bios that makes the SSD spaz out.
I'm going to grab another chip & see how it max's out & send this one to a friend to see if he can go past my max.
http://i106.photobucket.com/albums/m...573107BCLK.jpg
There you go, two heads or more are better then one. That is interesting and I have never seen that condition before though I did experience some very strange things occur with my ssd when first using bios f6a that I simply can not explain. No issues what so ever now with f6c, chip is still solid at 5650mhz.
Just noticed this on HWBot... could be interesting to some
http://hwbot.org/article/newsflash/1...abyte_p67a_ud7
There's a f6e sitting on the tweaktown forums....
HWBot is down?
Can't access hwbot, either.
Anyway, anyone try that f6e bios for the UD5? No feedback on it...and it didn't seem like it was out before F6 final came out?
F6e - 26.Jan 11
Theres no PLL overvoltage in f6e.
Is that confirmed or were you asking the question?Quote:
Originally Posted by Hoss331
If it was a statement have you tried booting past where your chip usually stops just incase they have automated internal pll & hidden the option?
====
HWBot working for me right now.
Im saying it doesnt have it because it doesnt have it, and yes I tried for high clocks.
Thanks, I'll skip that one then.
What are the improvments please ?Quote:
Originally Posted by stasio
^^^^ It was the chip, new one runs nice :D
http://i106.photobucket.com/albums/m...0MHz81081T.png
http://valid.canardpc.com/cache/banner/1648781.png
If you don't mind me asking, where did you get a new one at?
Got it hereQuote:
Originally Posted by truehighroller
http://www.cclonline.com/product-inf...69&tid=gsearch
Nice one C-N, looks like a keeper.
are you FN kidding me man, nice 2500K now :rolleyes::mad::D:rofl::up:
Awesome chip perfect for 5g 24/7, its got to break in voltage wise but atm she runs linx 5g at 1.32 bios! Low 60's
Nice to meet you all. Hot stuff going on here!
Just want to share the pain with some of you, I have that cold boot problem as well(some times it goes three times straight one after another). First I thought it was my HX650, after roaming in forums searching for answers (and unplugged my HDDs, HDD fans, USB port PCI) I almost bought a new AX650, but fortunately I found TurboDiv's posts LoL.
I have another PCI device left, a wi-fi antenna thing, so I might take that thing off and put back my HDD & co.
As Ram, I have RipjawsX set @1600CL8-8-8-24 1,56V.
I'll let you know how it goes..
:welcome:Quote:
Originally Posted by kadombing
Sorry to hear you are having problems, the only way I can make my board cold boot is to have the memory to tight or have PCH too low for the bclk attempted... sure it can be forced to boot after a few attempts & when you get it going it runs all benches but its a pig to start.... maybe this is a bug maybe not :shrug:
Nice chip, C-N. What batch #?
Thank you :)Quote:
Originally Posted by Jupiler
L049A796 SN: 2067
I think its just random luck with these chips from what I see voltage & speed vary wildly irrespective of batch.
Just added a small bit of info to the top of post 2.... http://www.xtremesystems.org/forums/...30&postcount=2
I have one of these PCI Diagnostic cards cost me £1.50 from Hong Kong & it works a treat.
TurboDiv's assumption was right I guess. I emptied all my PCI slot, stock clock, RAM tweak from the readable 1333 CL9 to its native 1600 CL8, everything else are stock settings. I run stressing and testing, no errors, only the normal-sometimes-twice reboot when I turn it on or on restarting. But as soon as I put the PCI card back and turn it on, I get error message saying there is system instability due to overclocking voltages and stuff.. and BIOS get revert to factory settings, except the RAM voltages(my setting is 1.56V and stock is 1.5V, and after error message is 1.526V or something).Quote:
Originally Posted by C-N
BTW, any of you experiencing RAM detection on CPU-Z of 10700 while BIOS(or Sandra) says 12800?
Very interesting.... something for the GB boys to look into I think, Although I have used both a PCI video card & my PCI diagnostic card without fault.Quote:
Originally Posted by kadombing
My PC-16000 shows up as 8500 in CPU-Z... a lot of my RAM shows this, think its just the JDEC spec rather than the overclocked speed.
what PCI devices are being used and in what configuration
specific models please and how it affects the board so it can be easily replicated?
Dinos: Soundblaster X-fi Fatal1ty.
@Dinos
Netgear RangeMax WPN311 Wireless PCI Adapter
Belkin USB 2.0 Hi-Speed 5-Port PCI Card
Stable system, turned system off to put PCI devices on, turned system on, wham! BIOS setting reverted due to unstable voltage/overclock message came out ( happened more than once)
Beside of that, can anyone tell me why CPU-Z shows x 37 multiplier while I set it up to x 40?
Yes, I disabled C3/C6, Thermal Monitor, C1E, EIST. Yes, I ran Prime95 and IntelBurner, but it still showing x 37. (core temp shows x 33 lol)
What did I missed?
Try clearing the CMOS. Also make sure the turbo options are disabled (so all cpu's use the same multiplier, and the independent multiplier settings should all be greened out).
Thanks. I did, but it was the same. So I tried to change all the cores' multiplier one by one before blued(or greened) them out by disabling the Turbo, and it was the same.
So I tried the same thing, plus enabling Real Time Ratio in OS.. And it worked! :D Sorry, first time build and first time overclocking :shakes:
https://mail.google.com/mail/?ui=2&i...f_gk1yna2d0&zw
VERY very odd.Quote:
Originally Posted by kadombing
What bios version are you using?
Usually a cmos clear with the power plug disabled fixes weird stuff like that.
Though I flashed f6d in MS DOS with spiflash.
I disabled real time ratio in OS, always kept it disabled, and I always had everything else disabled (except bi directional prochot), and I never had a problem with the multiplier lowering.
I also have all power saving stuff (besides screensavers) disabled in windows.
Oh, I didn't reset the cmos. I'm using F6. Actually I too disabled everything else, but somehow it didn't worked with real time ratio disabled. Couldn't be bother to reset cmos at that time.Quote:
Originally Posted by Falkentyne
Another thing is that EasyTunes wouldn't let me change ratio from the OS because all the arrows were messed up like a graphical error. But after rebooting(and making a different profile) it worked.
Hey guy, sorry to hear you got that nasty bug like me. For the record, I also have a Creative X-Fi Xtrememusic on one of the PCI slots. I was extremely annoyed with all of this cold boot issue and (knock on wood) i think i might have cracked it for the last 2 days. I don't want to bring anyone's hopes up but it appears to be extremely related to the CPU clock + PCI card + the power line of the house going to your system + the quality of the Active PFC on your power supply unit. ( Did i just mention like 90% of the components, lol)
So what i did first was bought a voltage regulator that connects my rig + monitor only and as soon i fired it up it registered that my line was overvolting and the unit was correcting it. Next i lowered my CPU down to 4.6 ghz and left the voltage a little bit higher - same as for my 4.9 ghz stable clocks. Ram was set to 2133mhz Cl7 as with its specs. At these settings my machine was having problems cold booting before w/o the voltage regulator but now it seems to be running fine for the last 2 days w/o any cold booting issues. At the same time i noticed a bit higher values for the 3,3v line from ET6 with the voltage regulator which suggests that the Active PFC of my power supply was not coping properly regulating the voltage from the house line. Also i noticed that BCLK exactly at 100mhz was making things more stable than to let the mobo default it at 99.8mhz.
First i would suggest to try and lower your CPU speed down to like 4.4 ghz with memories at the right specs and slowly build it up multi by multi until it starts cold booting or if it is cold booting w/o any CPU overclocks try and get a voltage regulator - it is always better to have one anyways (~30$). Right now if i increase the CPU multiplier to 47 it has issues cold booting from the first time but it is stable once it boots to windows, up to 49 multi with 2133mhz ram. This is all still testing things but so far i had like 5 successful cold boots w/o any problems with 2133mhz ram (XMP - profile1).
DEFINITELY reset the cmos. I found that the board was giving me 8x PCIE at 100 mhz (auto) and 99.8 and below, but was giving me 16x at 100.1 or 100.2. When I reset the cmos, auto gave me 16x PCIE again.Quote:
Originally Posted by kadombing
Is there a bios available for the ud5 that extends the multi range past 57?
I'm still here, trying to figure out these Gskill's and working on picking up a 3rd 2600k ..............................
Not that I've seen, perhaps PM Dinos & ask as he got one for the UD4 & 7.Quote:
Originally Posted by Hoss331
Hi bud, thought you had them all worked out where u stuck maybe we can helpQuote:
Originally Posted by OC Maximus
Post 2 updated with "Voltage Read Points"
http://www.xtremesystems.org/forums/...30&postcount=2
Goes without saying dont poke about if you not happy doing it!!!!
I find all but the VCore to be believable but as Sin said the choke voltage hasnt gone through any caps at this read point so :shrug:
If anyone has measured the VCore on the back of the board like we did with X58 & finds it more believable please let me know.
Any reason the UD5 is lacking a bios update when UD3P,UD4 and UD7 has new improved bioses ?