@new submission people: Great clocks and welcome.
@johnnymossville: Linx full memory is a harsh mistress. Keep it up!
Printable View
@new submission people: Great clocks and welcome.
@johnnymossville: Linx full memory is a harsh mistress. Keep it up!
Well guys i passed on lower vcore tweaking a bit :up:
http://img204.imageshack.us/img204/7...9turb0linx.png
Cheers :up:
Sergio
Nice run Sergio. :up:
My water cooling cannot handle 8 threads above 4500mhz so I started playing around with HT disabled tonight. This was my second try at 4700mhz, I think vcore could be a bit lower. On loop21 I took a screenshot and it knocked me down to 58 gflops.
Batch#3937B785 X58A UD7, 3x1gb HyperX, W7 64, watercooled
http://img263.imageshack.us/img263/5876/4725mhzlinx.png
Uploaded with ImageShack.us
http://img204.imageshack.us/img204/9...mhz25loops.png
Uploaded with ImageShack.us
Thank you TomQuote:
Originally Posted by OC Nub
Such nice runs mate i remember when you couldnt go above 4500 this is just amazing :clap::clap::clap:
Thanks Sergio, will keep pushing this chip.
I thought I'd give it a try again tonight and managed to pull off a max memory test with HT off at 4.510. My first try only made it to 12/20, but the gflops were a lot higher than this successful run. I think I'm going to tweak some more soon to see what I can do about that. I really think there's more left in this cpu as I figure out what I'm doing.
http://i39.photobucket.com/albums/e1...sign/4510b.jpg
great run :up:Quote:
Originally Posted by johnnymossville
Thanks I'm learning :), but yours is AMAZING! nice job!Quote:
Originally Posted by andressergio
4.7+ is way the heck up there.
nice start :DQuote:
Originally Posted by johnnymossville
thank you mate :up:
Cheers
Sergio
This is the highest stable OC for 24/7 that I can get out of it right now. Same as in my sig ;) Ben running this for the last few days without any issues.
http://i60.photobucket.com/albums/h2...complete-1.jpg
http://valid.canardpc.com/cache/banner/1200066.png
I was able to get this with HT disabled using a core vid of 1.50625v. Ambient air temp ~20C.
Stable enough for everyday use, but too loud to run fans full speed 24/7. I attempted with HT enabled but would always run into thermal overrun.
http://pics.hydndesigns.com/screenshot46ghz.jpg
Just for entertainment purposes, power consumption with only CPU loaded:
http://pics.hydndesigns.com/powerdraw.jpg
Very nice ;)Quote:
Originally Posted by famish99
Been trying to get 220mhz Bclk but doesn't look like 4.6Ghz is in my cards :(
I've not tryed to disable HT yet though, may have to play with that this weekend.
Do you know if your motherboard/cpu is capable of doing 220 bclk (ie lowering cpu and ddr multipliers, just pure bclk pump)?Quote:
Originally Posted by Major
Secondly, disabling HT gives you a pretty significant increase in fmax especially if you're running a moderate water loop (I feel like I could have better performance if my loop had a bit more flow). The heat increase from enabling HT was tremendous, I went from drawing 522w to 640w (take that with a grain of salt because of power supply and vrm inefficiencies).
Do you think the REXIII is a good replacement? Looking at the REXIII thread it doesn't seem any better than my UD7. I don't mean to be board specific and I have no data to back that up but just observing the REXIII thread doesn't seem to show any marked advantage over the UD7. I'm not making statements here, only that, by observing said thread, I cannot justify the necessity to change. Is there some bios settings that I'm missing? Better chip control, IMC, Vtt, V droop, ram issues, etc?Quote:
Originally Posted by zsamz_
On a different note are there any Gulftown ockers who get higher clocks with HT enabled? When I turn HT off my 4.7+ clock isn't stable anymore.
i just sold my rexII and ordered the rexIII thursdayQuote:
Originally Posted by newhit
and my rexII is as good as day 1:up:
i'll find out soon if its better
the downside is that the rexIII has 1 usb header:mad:
no way i would buy a gigabut over a rexIII
if the ud7 was better i woulda had one already;) and save myself 50$
quality control @ gigabut is:down: with my luck i'd get a lemon :ROTF:
clock for clock the rexIII is faster and the way i see it the rexII is better than the ud7 and dont forget the horrid colors of a gigabut board
so with rexIII i should have a winner and clock my cpu with less volts:up:
the rexIII looks better than the rexII and doin 235-250 blk is :up:
p.s. i'm not an asus fanboy i just want what i think is best
and i still have a gigabut p45-udrp if you wondering if my boycott of gigsbut is still on :ROTF:
read the newegg reviews so many ud7s 20% are doa payin even 100$ more is better than havin a board thats gonna crap out
Sorry, did not mean to give the impression that my UD7 was better. I was wondering only, if there was something extra that I was missing (apart from the general dislike of "Gigabut"). I had a REXII that my son lifted and he is more than happy with his 930 at 4GHz 24/7, gaming, video etc.
they make good boards but quality control is :down:Quote:
Originally Posted by newhit
i am verry happy with my rexII
but i want 2050-2100mhz cas7 and 230blk+
i woulda bought a classy but then i gotta watercool the whole board
Just put it under phase..........................problem solved :)Quote:
Originally Posted by zsamz_
@newhit: Don't apologize the ud7 is awesome. Someone will always chime in and try and put you down. "if the ud7 was better i woulda had one already and save myself 50$" In the last few months the ud7 is at the top of most benchmarks, this is not an accident. Wait another few months and see which does better ud9 or re3. For boards at the very high end the differences are so few one can hardly be said so be so much better to warrant such negativity. Not to mention calling them "gigabut" sure doesn't add anything positive to the discussion.
This is max I got so far, still playing around with different settings.
Will test if stable 24/7 over the next couple of days.
http://valid.canardpc.com/show_oc.php?id=1203514
Cute little guy :up:
Welcome to ze club.
i been thinkin of gettin 1 built
was gonna build a chiller but it takes too much space
my playroom is only 8x10:(
Quote:
Originally Posted by kitfit1
just cause I wanted to part of this club:D
I am lowering voltages as we speak, namely vcore as it was very high, qpi/vtt was at 1.38v
http://i49.tinypic.com/imnt4l.png
that is with HT off, I keep getting all sorts of 3B, D1 and other BSOD's with HT no matter what.
l0ud_sil3nc3,nice OC!that W3540 looks like a great chip.have fun and keep pushing it.
Quote:
Originally Posted by radaja
It's under water now, its a weird chip for sure, it needs next to nothing voltage wise to 4.37 but anything over that and it is a pig. Also I wish I could get that speed with HT on, but oh well more testing as we speak.
I am thinking about picking up an X58A UD5 Rev 2 since they have fixed the Xeon uncore multiplier issue.
I have the exact same issue, I can run up to 4.3 at like ~1.3v, but once I pass that I'm looking at 1.4-1.5v. I can't say that it's normal, but at the very least expected heh.Quote:
Originally Posted by l0ud_sil3nc3
@ lOud_sil3nc3-I have an i7-960 that runs 4.0ghz stable on stock vcore, but cannot get it stable at 4.5ghz for the life of me HT on or off. I hate that chip!!! Guess some are just weird that way.
Inching my way past my previous best.
http://i39.photobucket.com/albums/e1...n/linX4574.jpg
i7 920 D0 @ 216x21 = 4536mhz
HT on, EIST on, VM on, everything but C1 on basically.
mobo is a x58a-ud7 with EK FC block in the cpu water loop on bios F6
Running around 1.4v cpu on bios but it's about 1.36v in cpuz.
Watercooled setup:
apogee gtz
Indigo Xtreme
2x mcr320 with 9 gentle typhoon 1850rpm (3 on 1, 6 in push/pull on 2nd).
1x mcp655 + EK custom top
distilled water
Max temps are about 73c on 1st core.
There's also a similar loop on the 5970 (ek fc block, both rads are only 3 fans each, same pump/top combo).
http://img257.imageshack.us/img257/5...zlinx20.th.jpg
I'm considering pushing to 21x220 and see if I can do it with less than 1.4v... I'll keep you guys up-to-date ;)
Congrats antiacid - nice clock for such low vcore - also very nice temps!
Great result :up:Quote:
Originally Posted by antiacid
@ ALL You can really see some impressive chips around here... ;)
Did you try any of F7 beta bioses for UD7? I'm wondering would there be any difference in terms of Vcore?Quote:
Originally Posted by antiacid
BTW what LLC level do you use?
Congr once again for nice OC :yepp::up:
Using 980x - was on F4 went to an F7 beta and then went back to F4 - I saw no improvement in my oc, actually f7 beta was a bit more finicy.
I like the wider v upswing in level 2 vdroop control in F4 vs the f7 beta i tried.
I'm at llc 1 and I didn't try any bioses past F6. I wonder if there would be voltage gains but I doubt it.Quote:
Originally Posted by Gorki
Right now I'm trying 220x21 and I'm playing around with voltages to make it stable for linx 20 pass. It passes between 3-6 times before BSOD'ing at 1.408 in cpu-z. I'm at 1.36 qpi/vtt and thinking of going to 1.38v.
Wondering if that'll help my qpi which is now 3960mhz (vs 3880ish at 216).
Also, I'm going to start using llc 2 because I was up to about 1.44 in bios and it still showed only 1.39-1.408 in cpu-z... It's just ridiculous.
On to testing!
http://img37.imageshack.us/img37/522/4620mhzlinx20.jpg
220x21 @ 1.41v :) (i7 920 3849B018)
Actually, this time around I ran with llc 2 therefore it shoots up to 1.42-1.44v in cpu-z during the linx 20 pass test.
I ended up getting my volts a bit higher too on the qpi/vtt: 1.38
Note that the temperatures are now 83-84c at max, which to me is unacceptable. I'm eventually going to re-mount when I get around to getting the case painted and that might help (the remount, not the paint lol) but until then, it looks like 4600+ is just too hot for my use.
some more volts of importance on my ud7:
pll: 1.8v
qpi/vtt: 1.38v
vcore: 1.425v (in bios)
dram: 1.64v
IOH core: 1.16v
Impressive 920 :up:
Got myself another chip to join the club with, just need to tweak it (testing IMC right now)
Looking good mate! nice chip!Quote:
Originally Posted by antiacid
85c is nothing to worry about! on air I saw temps in the 90c plus range!!! HAHA
IF my mb would do 220mhz Bclk I'm sure I could get 4.6hgz out of my 920 but it would take upwards of 1.52vcore I'm sure.
your telling me I am at 4416, and completely stable, but dam 4.5 seems like no matter what I do, (besides turning of HT) always fails IBT.Quote:
Originally Posted by OC Nub
^^ I am stuck at just under 4.4ghz on this 960, cannot get 4500 stable. Have put the chip away several times out of frustration but keep reinstalling it to try again,I am about to give up on it. I will stick to lower priced CPU's, both my 920's will do 4500 stable with enough vcore and good cooling.
Back in the day when I had money :DQuote:
Originally Posted by OC Nub
I used to buy the mid priced chips, mainly P4 and AMD socket 939 but seldom do they OC better than the base speeds. Only when you pay big buck for the special editions do you see any improvment in OC
well I finally got 4.5 with HT on and cas 6
voltage and temps are high and I am in the process of lowering vcore and vtt
http://i49.tinypic.com/2uqiz3l.png
the reason for only 4GB and dual channel is I had to RMA some memory in my HTPC to Corsair so I had to rob a stick for it.
I still can't believe the DDR3 2000 c8 sticks I have can almost do DDR3 1600 6-6-6-18
congrats mate :up:Quote:
Originally Posted by l0ud_sil3nc3
Well guys
After some days of tweaking and finding out references on the MCH i can say im fully stable on a P45 board at 4500 / 2000 1T
http://img227.imageshack.us/img227/6...cr1pl8linx.png
Very happy with it
Cheers :toast:
Sergio
thats some nice stuff sergio now becnh :P
Damn you. 4500 mhz at 1.392v? On a 1.2000 VID chip? Mine is 1.1500 and takes almost 1.35 at 4.0 GHZ. What's going on? Maybe I'll just have to dedicate an entire day to tweak it below 1.3v where it should be at for its VID. You've probably seen my frustration in the other thread which I finally got stable from feeding it lots of volts but something just doesn't feel right.....
The only thing I could think of is my hard drive...it's the only thing I haven't upgraded in a long time...would this be causing the issue? It's an old Seagate Barracuda 160GB that I bought like 3 years ago and haven't bothered to upgrade cause I'm such a neat freak....I just suspect it and it probably isn't causing me the problem but you never know....I'm getting desperate now :mad:
A HDD shouldn't effect your overclock, but if you want some peace of mind, you can always disconnect and do some quick tests? Anyway, the VID of a CPU barely says anything about how it will clock. The most important thing to look for is it's batch #Quote:
Originally Posted by Dark-Energy
im at 1,376 stable but for such pressure on the MCH and fast mem i need to notch vcore, same happened with my UD3P and Gskills PI1200Quote:
Originally Posted by Dark-Energy
u can see here
http://img204.imageshack.us/i/45001600pl9turb0linx.png/
i also been through many horrible chips my friend and never got lucky until i start buying for know good OC batches and tested for some nice friends here its all a hazzle
dont go desperate if its bad just sell it and spot a good batch nr
Cheers mate all the luck for ya :up:
thanks man will do :DQuote:
Originally Posted by MrDucky
How come I only have 49 GFlops at 4.4 Ghz??? Not enough voltage?
Try selecting "all memory" when running LinxQuote:
Originally Posted by KaRLiToS
Has nothing to do with voltages!
ok I'll try that
Well the batch on mine is L932B343 Pack date 9/16/09 but I found this thread here...
http://translate.google.ca/translate...%26channel%3Ds
From March of last year (2009) with someone of the same batch and got 4 GHZ at 1.175 volts...but that is way before the pack date on mine....seriously what? I'm confused...or maybe I am misreading it cause it's translated
1/3 hp kayl single stage, 3845A935
http://www.photohost.biz/files/5elw80cireopv6lyk9xm.jpg
How are temps at that speed/voltage?
best batch from W3520 nice score :DQuote:
Originally Posted by Anth Seebel
Yup, nice result. Especially with relatively still sensible Vcore :)Quote:
Originally Posted by siyah
i want one so bad.anyone got one for sale hit me up:DQuote:
Originally Posted by siyah
Does raising PCIe Frequency to 110 Mhz will help me lower the CPU temps, I can't reach 4.5 Ghz, I have good watercooling kit, (check sig) and my stable voltage at 4.3 Ghz is 1.31 v (IN BIOS)
Thanks guys
Hi, I think PCIe Frequency don't matter at your problem ?? .... you need better batch (cpu)
Karlitos: most probably you'll be able to go higher with vcore ~ 1.4-1.45, but performance gain will be minimal in % while power used/heat dissipated will increase 1.5x times. I've often seen people going upto 4.5-4.6 for benchmarks/finding limits, and then rolling back to whatever frequency are more sensible in power usage for running 24/7 (usually 4.0-4.3 with very little voltage bump).
Lol, I put mine up on Amazon but you won't like the price. :rofl:Quote:
Originally Posted by radaja
cooler: idle -42C, load -37C, real temp 3.46: -6C idle , 34C load ..Quote:
Originally Posted by tool_462
Quote:
Originally Posted by siyah
Quote:
Originally Posted by churchy
thankyou for your kind comments :DQuote:
Originally Posted by radaja
Hey zalbard, when will the list be back up on the first post ?
hi everyone i just wanted to know a noob question
i know cpu's have millions opon millions of transistors but if one of them stoped working will the cpu become unstable under overclocking
or give some sort of problem :) thanks
If you had a cpu with only one transistor busted, I would pay you amazing money for it hahaha. Defectivity rates for silicon are fairly high especially for such a large die on small geometry. Typically 20% of the die is set aside for what are called repair arrays, so as long as you don't have too many bad transistors on a part, it's usually repairable. This isn't a user process though, most parts are repaired at the factory via fuse burn, but it's possible some motherboards are able to do soft repairs via scan chain. Even so, most of the time you're going to have some stuff busted and you'll almost never notice until you hit some corner case.Quote:
Originally Posted by rubber5
wow i didnot know that thanksQuote:
Originally Posted by famish99
its just i really wanted to know becuse i know its rare for cpu's to go down
but the million dollar question is how meany transistors will have to go bad before the cpu is classed as dead?
Lol quite the opposite, cpu's are much more broken than you think. It doesn't take many transistors really, some logic is not repairable, a defect in those usually means a scrapped or binned chip (it may behave at a lower frequency/temp combination). As for million dollar... I'd argue it's multi billion dollar question lol, test time makes up for 1/3rd of the cost of a chip. You try to test as much of the chip in as little time as possible, but test escapes are unavoidable.Quote:
Originally Posted by rubber5
yip your right about the billion doller question lolQuote:
Originally Posted by famish99
thanks for that i wish i had the balls to push my cpu to 4.5 ghz i have an i7 920 d0 1.125volts bios @3.5ghz and it can go to about 70c max but about 65c when crunching
i dont want to push my luck cant really replace it at the mo lol :)
The OP is almost re-done, sorry about the delay. Will probably finish it by Sunday. Monday latest.Quote:
Originally Posted by kitfit1
There is an emergency situation at work atm. :(
ahhh i can not remember the rules at 1st page
1. Must use all available ram.Quote:
Originally Posted by PatRaceTin
2. 20 loops of LinX passed.
3. 4.5GHz
:shrug:
I brought the requirement part of the OP back up. I'm almost done fixing the list. :up:
i hope this follow the rules
normal water cooling
DRAM Voltage is 1.94
http://img823.imageshack.us/img823/948/45linx20.th.png
@PatRaceTin - very nice and congrats! room temp of 29c probably isn't helping is it - still you stayed under 101 so as i understand it - no temp throttling - good job!
linx 4.5ghz HT off 2150 C8 1.62v eleet set , load read 1.617v
10 runs.. another 10 to go
http://i287.photobucket.com/albums/l...g/18686597.jpg
20 runs complete
http://i287.photobucket.com/albums/l...g/97c635f8.jpg
Check the OP, folks! :D
If you spot any errors (and I bet there's a ton of them, notepad FTL, and I had to re-do the whole thing from the scratch), please let me know! :up:
My W3520 is 3848A352, not 3849B018.
Also, could you add my max 4c\8t run from here, think its nice to see both 4c\4t and 4c\8t max in the list.
Thanks, and keep up the good work :)
My batch # is 3845A935. Thanks zalbard for the new list. Very organized. :up:
3003B466
TRUE + Ultra Kaze 3000
RAM very relaxed (~DDR3-1500), 1.5v vdimm, rest auto
Attachment 105315
@Gautam - very nice! - 4.5 linx stable on air with good temps!
When is the next die shrink? 5.0 linx stable on air for that?
@Mr. Zalbard - execellent work on the OP reorg. - Also - much thanks for all the time you put in managing this thread - one of my favorites!
Thanks, everyone, for submitted info! Everything should be addressed! :up:
Thanks for your kind words. :)Quote:
Originally Posted by SteveRo
Second this, once again great effort Zalbard. :up:Quote:
Originally Posted by SteveRo
Dunno if 81C max counts as a "good" temp...but it makes it through at the least.
i concur,very nice work keeping this thread awesome zalbard:up:
and very nice chip Gautam:D
update
set 216 lowered vcore ,same vdimm
http://i287.photobucket.com/albums/l...g/c863932d.jpg
Hey Zalbard my CPU batch is 3849B018 (currently missing from front page info). Good work with the thread too, nice to see someone putting in a continued effort.
Quote:
Originally Posted by Gautam
Quote:
Originally Posted by radaja
Thanks, guys! :)Quote:
Originally Posted by CryptiK
Quote:
Originally Posted by CryptiK
Updated, thank you! :up:Quote:
Originally Posted by cstkl1
kitfit1
i975
Home built rotary SS cooling
4.917ghz 4 cores 8 threads @ 1.505Vcore
http://i48.tinypic.com/fxgcib.jpg
^^
Jesus! :eek:
Very nice temps, you have a decent unit! :up:
@kitfit1 - wow! :shocked: and only at 1.5v too! what is your ss tuned for - ~300w?
I think you probably have more left to go after being only at only 1.5. :eek:
1.6v for a 45nm should be ok for an hour ot two benching I would think - nice work! :up:
Thanks guys.
This phase unit is one i built a month or so ago (it's in the phase section).
At the moment i'm still in a very refined "tuning stage". If any of you have built a rotary SS you will know what i mean.
Very trial and error, and can take a long time. I think that i'm not far off getting this one spot on. Tuning is a bugger with a rotary,
but the end results, if you get it right, more than make up for all the grief involved. I will point out that i'm about 10 to 11 rounds through 5ghz at the moment.
But i have been there before, just not at this low temps.
By the way, there is no secret to running high clocks with low temps. All you have to do is run CPU PLL at 1.2v. But you do need to be at -38c minimum to be able to run at that.
Anything higher just dosn't run the distance.
nice run kitfit
:up:
Great work...though I gotta say you must have some tough ears to be able to rough that out if that's your 24/7 setup you're using a rotary on :p:Quote:
Originally Posted by kitfit1
@kitfit1 - WTB 5.0Ghz 24/7 with HT off for the world's fastest gaming rig :up:
^^^
Lol, HT off is not a problem. I'm after 5ghz with HT on. Just a bit more difficult me thinks.
Word.Quote:
Originally Posted by kitfit1
Most games do a little better with HT off though :p:
Just bought a 920 as my 930 would not go past 4.2. I had it boot to 4.5 but forget about getting it stable. I couldn't get it to run 4.2 with HT that took to much voltage for my liking. Now I just installed this cpu yesterday and this thing just blew me away.. Here is my first stab at 4.5 with this chip. The voltage is high for HT off as I was trying with HT on and turned it off to see if I was at least close to being stable. Dont Add this to the list yet I think I can lower the volts a bit with 4c/4t and then make a pass with 4c/8t and then push higher. 1.4375 was set in bios.
Batch #3927A440
http://img696.imageshack.us/img696/1704/45004c4t.png
http://img690.imageshack.us/img690/9...004c4tdone.png
Edit - Just remembered has to be 20 pass stable which im sure this is will be.
good work zalbard :)
nice runs in this thread indeed , some low gflops runs in list tho which means cpu wasnt loaded properly due to instable settings another clue will be lower temps than a stable run..
hey kitfit1 nice run , what hp is your rotary compressor ? must be a monster to deal with that load :)
It's a 28cc Highly (Hitachi) CHZ33LC4-U. 1HP i think.Quote:
Originally Posted by Anth Seebel