Printable View
Edit.... Nevermind.... hit up Speedlanders post
It helps out a bit more
http://www.xtremesystems.org/forums/...d.php?t=175241
There have been at least 7 requests for an excel sheet in this thread...Quote:
Originally Posted by jaredpace
Danmoody... If you dont have the inclination or otherwise, will you update your first post with proper links that someone else has made? Perhaps someone that wants to could start one, and danmoody can direct to the post from the first link...
just saying that with google docs, the spreadsheets out ther eon the web and everyone can just add their little data to each column when its done, or has enough statistics in it, you can just create a new SS summarizing all the data collectively, so you could then see like this...
q745's above 4.4ghz: 177
q740's above 4.4ghz: 101
q743's above 4.4ghz: 88
or data like average q745FP0 VID: 1.03345443
average number of faulty coretemp readings: 384
or data like average Vcore needed to obtain 4.4ghz minimum: 1.08volts cpuz
it would be a lot easier than readying 30 pages of screenshots, "lookey here!"
or, "i think i saw back on page 15..."
clean this garbage up a little, you guys have all the pertanent data, just summarize it to make it easier to soak up.
:yepp:
http://spreadsheets.google.com/ccc?k...8iPVEYLw&hl=en
Kinda need to invite people to get this going lol.
*ok, seems like Google Docs needs to have collaborators manually added...
How is it going with this one?Quote:
Originally Posted by hazardous
Mav451... you added me when i was in there... ill update the excel as well...Quote:
Originally Posted by Mav451
keep contributing, but do so with proof at anything above 4.3Ghz please.
Validations and/or ss would be good
Thanks.Quote:
Originally Posted by Ron 61
I'm not very good in electronics :P where I sould put black and where red ending of voltometer?
BTW Is it possbile that big difference betwen BIOS and Windows? or Windows (CPUz ) reading wrong?
I guess current E8400's aren't consistent. Some are super but a lot of them are gay too :( I can't get it 4.3Ghz stable, sigh.
Hello.
To be quite stable with E8500 I it is necessary to avoid the half coefficient like 9.5? With 9 is it better?
Thank you
batch etc, sig.
scaling on my is .05v for every 100 mhz above 4.0
1.2875 bios 4Ghz stable
1.3375 bios 4.1 stable (turbo, mem perf 7, 4,4,4 RAM) pic below
1.3875 bios 4.2 stable (stand, mem perf 8, 5,5,5 RAM)
1.4375 bios 4.3 orthos 1+hr, temps in 70's, did not run longer.
24/7 is going to be 4.1 to 4.15, its almost same speed benchmarks as my 4.3 because of tighter timings I can run at lower ram/fsb speed.
custom run 8K-128K small ffts
Hi,Quote:
Originally Posted by XiCe007
No the 9.5 works well.
Regards
andy
just received my E8500 #PFO Q745A554 packdate: 01/10/08, hope it is a good 1
Azerty :D ;) Same batch on Commando:
http://img.photobucket.com/albums/v6...hz128vload.jpg
I'm lost, i don't know what to think anymore...
Today i got my new E8500 with FPO# Q745A554, a batch that does well for others so i couldn't wait to get her in.
Booted fine and all but the first dissapointment came when i opened coretemp...1.1125VID which is exceptional high for an E8500 but ok.
Next thing was a problem more people have reported, one core is 33ºC the other is 48ºC, idle!!
Under load the hot core gets up to 60ºC, this must be the cpu cause my E8400 had just 38ºC idle on both cores. :(
Now onto something bigger, getting 4Ghz stable.
1.25V in windows, Prime 95 runs but just closes after a couple of minutes without an error warning or something, same goes for orthos. wtf??
I ended up giving it 1.31V in windows which droops to 1.304V to run occt stable.
This must be the very first E8500 that needs 1.3V for 4GHz.
I played with all possible voltages in bios, Vpll, GTL, termination voltage, NB, SB, you name it, all to no avail, the only thing that seems to help is Vcore.
I could run 3d mark 01 at 474x9.5=4.5Ghz at 1.52V and could do 1M close to 4.6GHz at that voltage.
I really suspect my motherboard, it either hates wolfdales or wolfdales just hate me. :(
Gonna have to break out the wallet again for a new mobo i'm afraid.
I'll be glad when i've had enough of this. :rolleyes:
Damn.
Maybe for prime at 4ghz its a bit much however close to 4.6 with only 1.52v for 1M isnt too shabby. Me first E8400 wouldnt even boot at 4ghz with 1.375v set in bios (approx 1.32v load) so sold that one. E8400 #2 does max 1M with 1.45 bios (1.4 load) at 4460mhz and with 1.55v bios (1.5 load) 4596mhz. The E8500 (same batch Zeus) 1.45 bios (1.4 load) 4400mhz and 1.55 set (1.5 load) 4530mhz...
I think you might be right. Something is fishy, I've cut way back on ocing these until a new bios or two comes out.Quote:
Originally Posted by Zeus
Has anyone ruled out x38 being better than P35 for wolfdale? Tested the same chip on both chipsets? I'd like to see that.
Yeah, it's the same as mine. It needs 1.32v or 4.0ghz and 1.48v for 4.3ghz occt stable. It likes vcore... Throw some 1.64v at it and run some SPI1m at 4900mhz ;)Quote:
Originally Posted by Zeus
don't judge your chip on core temps these chips do not accurately report the temps to core temps. Some report too low and some too high. I wouldn't base anything on those temps both my cores in my second E8400 read 40c no matter what voltage at idle ,then under load one core jumps ten degrees and runs hotter than my old 65nm Quad Xeon 3210 with 8 Mb of on chip memory and used 105watts , which is not possible being that a e8400 45nm dual core with 6mb on chip memory.Quote:
Originally Posted by Zeus
Any pics? If on air 4900mhz with only 1.64v would be rather nice..Quote:
Originally Posted by AriciU
If you are not comfortable using a digital multimeter then don't do it. Its possible that there is a small difference between what the bios and windows is reporting but they are usually pretty close. For now my advice is to update to the 16.b04 beta bios and go by what uguru reports for voltages. Meanwhile use google to read up on how to use digital multimeter.Quote:
Originally Posted by Blackwhite
Now I know what to do with multimeter - but I'm a little scared :PQuote:
Originally Posted by Ron 61
I've set 1.485v in BIOS, then I restarted comuter and I've check vcore in BIOS again - was 1.39V - so I think my Mobo droping voltage to this level.
BTW - I've got IP35 without Pro - so the newest BIOS is ver. 15 and it is no difference on it.
All the screens are in my thread in the xtreme overclocking section... Not sure on the if it was 1.64, close to that anyway... 4905 was stable on cool air with the case out and 485x was with case in room, 24c ambient.Quote:
Originally Posted by loc.o
I changed from XfX 780i to my Asus P5k dlx, i need 1.344v @ prime with p5k against 1.336v on xfx, all at 4.2ghz. Just started and my batch is not so good same crap as Pytonorm Q748A142;)
update later with screen.
Well any comments yet on the chip?Quote:
Originally Posted by mc2k
Its been 3 hours you should have had it all tweaked out by now:D
Thanks for the link, I dont speak Greek but I work with many that do so I should be able to get the info I need.Quote:
Originally Posted by Stelios
I run 478x9@ 1.385V ( Actual 1.368 )
I run prime and OCCT, I use 3DMark and Everest to keep an eye on performance
Temps are great
26c idle
53c load
found this on another board.. would be GREAT if everybody would enter results here
http://spreadsheets.google.com/ccc?k...8iPVEYLw&hl=en
Haha by other board, you mean back one page (also in my sig)?
Probably should update the 1st post, assuming danmoody comes back sometime soon.
Your temp too high? hah, youre the one loading at 55/56 while im over 60C almost and im the one using a fuzion compared to your tdx :rolleyes:Quote:
Originally Posted by BooStFeD
using bios 1226. ill remount the sucker couple of times, if it doesnt yield lower temps ill order a pa120.3 and add it to my loop :D
Why my cpu is so hot :(
My cooler is TT big typhoon , Is it normal ?
http://img517.imageshack.us/img517/9098/61879240lu7.jpg
Same problem here, but not that hot.
Coretemp gives 61C for both cores, but my Tuniq Tower is cold. Everest/PC Probe gives 46C (I know that these are CPU temps instead of Core temps).
But is Coretemp wrong? (using 0.96.1)
Also Core 0 is much better then Core 1. Core 0 can handle 4ghz with 1.24v, while Core 1 need much more(over 1.3 :|). How can this be?
BTW I have an advice, don't stress with gromacs core(orthos). It's useless.
I tested it for 2 hours, and with no problem. But then I started Small FFTs and it went wrong from the 10th second :|.
:)great E8200 .. 4ghz on air?Quote:
Originally Posted by vera
fpo\batch?
Just got a 8400 from Mwave, 745A551. Put it in my Microfly with a P5E -VM HDMI and TR120 UL. Using onboard video to start. Need to update bios but hate to on an Asus mb. Temps are way off. Using Vista 64-bit so no coretemp.
Edit: updated bios but temps are still way off.
http://img264.imageshack.us/img264/6...034vzj8.th.jpg
http://img264.imageshack.us/img264/8...occtzg1.th.jpg
http://img183.imageshack.us/img183/9...pi8moz0.th.jpg
My latest stability ;
Fpo : Q745A612
http://img230.imageshack.us/img230/4779/75538882kh0.jpg
Hey m8, u are right but unfortunately I am very busy atm, it need to wait till weekend before I got some time to test this sucker, will let u know m8:up: , Zeus & loco already posted some results of this batch, it seems not a "very spectacular" chip a good E8400 can do the same or better, btw nice chip u got their m8, if my E8500 would do the same I would be happy alrdy:up:Quote:
Originally Posted by Grnfinger
Not sure on mine re temps, seems a bit high ??
4230 470x9 1:20 1128Mhz Vcore = 1.3875v on water
http://www.pics-clunk.org.uk/wolfdale/470x9Occt.png
Benches 4230 470x9 1:20 1128Mhz Vcore = 1.3875v Benches
480x9
These are the 4g-4.23g-4.3g stable situations with my E8500 Q740A476T on watercooling.
screenshots are within 10' but it went on for a couple of hours.
and my max 32m without any tweaking,just to see what the cpu could do..
it's wrong core temps has a hard time reading these chip some are reporting too low and some too highQuote:
Originally Posted by barzi
http://individual.utoronto.ca/bzhou/...050-1.360v.jpg
Super quick non tweaked results.
First boot was stock speed only to set the BIOS.
Second boot into Windows :)
Will test for 24/7 stability soon.
I have the same temps with the same vcore, but heatsink is cool, so I'm guessing it's a sensor problem.Quote:
Originally Posted by vera
Hmm think I may have damaged the chip by pushing VTT @ 1.5 ?
I was Prime95'ing at 1.3500v 450x9 for an hour or so before I tried other settings... now I can't Prime95 for 1 min at the same settings without it failing on one core! It now needs 1.375V set BIOS.
DFI's digital pwm makes selecting proper vcore more difficult with vid + special vid % method. You can derive the same vcore using different vid + % combinations but they'll result in different levels of stability for prime95 etc.Quote:
Originally Posted by zoob
I'd also double check all your settings to see if they're all the same as before down to memory related settings too.
is your heatsink as hot? if you are used to, and if you have a good contact, may you guess the real temp by touching it, very approximatively?Quote:
Originally Posted by vera
I hope I'm in the right thread.
My system:
CPU: E8200 Q744A736 cooled by Xigmatek HDT-S1283
Board is: Asus P5K-E Wifi/AP, no mods, bios 1002
Memory: Transcend Axeram DDR800, MicronD9GMH chips
Vcore: 1.45v set in bios
Vddr: 2.0v set in bios
NB: 1.55v set in bios
http://img209.imageshack.us/img209/3782/4320145nt1.jpg
Just started working on my chip but the temps are very odd. Here is the data:
e8400
Q746A377 SLAPL
1.225v
Pack date 1/4/08
http://valid.x86-secret.com/show_oc.php?id=305730
http://img147.imageshack.us/img147/9...0mhzrk9.th.jpg
It was running P95x2 at 4203Mhz
http://img82.imageshack.us/img82/7443/p954203oj1.th.jpg
The odd thing is that it boots at 49c (Tc 29c which seems right from the 3 Core 2s I have owned) 1.19v Vcore 3002 MHz. If I run P95x2 there the temps jump to just 51c. Now, crank up the voltage and freq to 4203Mhz, 1.38v and it boots at 51c and primes at 64c after 15 minutes. This is on water, a 1/2" loop with dual rad and D5 pump that kept a Q6600 real cool at 3402Mhz, 1.41v all day long, so i know it isn't the cooling. Nothing but the CPU on that loop. Reapplied AS5 and same thing. It definitely clocks well and considering the Tj is 100c, 64c at 4.2Ghz under load isn't bad but these temps have me confused. WTF is going on? She boots at 4510Mhz 1.4V which is great. Can't run Pi yet but I haven't tweaked it much there (just installed).
There seems to be the usual amount of confusion about temperatures so here's another way to directly access the data from within the Intel CPU.
Start up CPU-Z and go to the About tab and click on the Registers Dump (.txt) button.
Save this file, open it and scroll down to the model specific registers for each of your cores.
http://img82.imageshack.us/img82/1610/cpuzdumpmx1.png
MSR 0x0000019C contains the only Intel documented temperature data for these processors. By looking in here you are reading the digital thermal sensors (DTS) directly. Bits [22..16] are located in the third and fourth digit. In my example that is 0x3F hexadecimal which equals 63 decimal.
If you are mathematically challenged then fire up Calc or head to Google and type in:
0x3F in decimal
and it will do the conversion for you.
This shows that my Core0, at idle, is 63 degrees away from TjMax or its safe maximum operating temperature.
I am using an older E6400 which has a TjMax = 85C so CoreTemp reads the DTS data and then calculates that my absolute temperature is 63C degrees away from 85C so it reports an idle temperature of 22C (85C - 63C). CoreTemp is using TjMax = 105C for the new 8x00 series processors.
Intel neither publicly documents that these on chip sensors are accurate for low idle temperatures or provides any public documentation stating what TjMax actually is. It only documents that these sensors are used for thermal throttling and ultimately for thermal shut down if needed.
When your processor heats up the DTS will decrease. When the DTS approaches zero you will begin to slow down due to thermal throttling kicking in but most people are miles away from that ever happening. As long as you always have some headroom as reported by the DTS and your computer is stable at full Prime load then there is no reason to be too worried about your absolute temperature. It is just a number and you are operating within the Intel temperature spec for these chips. At 4+ GHz you might be a little over on that spec but that's a secret that Intel doesn't need to know. :D
You can also start CoreTemp and go into the Options->Settings menu item and turn on Show Delta to Tjunction max temp.
This setting also reads the DTS data directly and should exactly correspond with the above DTS data that CPU-Z prints out.
unclewebb, that was very informative. Thanks a lot.
So what program is best to use for the temps? Speedfan?Quote:
Originally Posted by trt740
Can someone give me a link where I can download Prime, but for dual core? Like this picture.Quote:
Originally Posted by zoob
On this page:
http://www.mersenne.org/freesoft.htm
I got the Prime95 for single core...
Thanks
@ unclewebb If intel does not provide any public documentation stating what TjMax actually is, then from where is it derived?
Okay, been pulling a few of my hairs out over this lately.
It seems I didn't get a good one out of the batch. I think my CPU starts needing lots more power over 3.8ghz. At 426x9 ~3.8 I can complete orthos and prime for at least an hour on 1.35v. Once I start heading for the 4ghz line it starts needing VCore to reach stability. To get stable at 446x9~4014 I need 1.44 volts as reported by the hardware monitors, I haven't checked with a DMM yet.
I can boot into windows fine, run my applications, but as soon as I goto Prime it, it makes it about 10 seconds and fails. I have also tried adjusting my CPU VTT and GTLREF settings up and down 1% at a time from 55%-75%. No dice..
Perhaps I've got a bum chip here? How is everyone seeing vdroop affecting their OCs at higher frequencies? I am considering running to Rad-Shack tomorrow to pickup a variable resistor and try to get vdroop under control.
Part of me wonders if it will OC better with a proper BIOS from abit for the wolfdale CPUs?
Your chip is ok, I think so.I had the same - problem is with our Boards and the BIOS.Quote:
Originally Posted by HaTaX
I have to set in BIOS 1.485vcore , MCH 1.45V and VTT 1.31V and then I can run stable on 4 GHz.Offset betwen BIOS and Windows is very big, and Real vcore is 1.39V (1.368V after vdrop).Dont touch GTLREF settings it is waste of time.
very wrong...they are an important factor for stability..:)Quote:
Originally Posted by Blackwhite
Yeah, I noticed the large drop in requested voltage VS what is actually being supplied. That being said, to get to 4014mhz I have to set the voltage in my BIOS to 1.555...
Blackwhite, what is the VID on your chip? Mine is 1.0375 read in coretemp.
OMG So it is not just me then? I thought there was something really wrong because I also have to set BIOS voltage very high to get much less REAL voltage.. it seems..Quote:
Originally Posted by Blackwhite
I have to set it at 1.465 BIOS (I think) but you say it is an offset between BIOS and windows? strange because, even in the BIOS PC Health Status it reads the same as Windows
Set at about 1.465v it reads..
1.32v - PC Health Status
1.352 - CPU-z
1.35 - Everest
I have the IP35-E and I am reading a few people mention something like this with these new chips on the Abit IP35 Series I think? but nothing as extensive as what I have here.
It SEEMS (and I pray this is correct or I am feeding my chip a very unhealthy ammount of vcore) I get 1.352 REAL 1.32v LOAD
http://img301.imageshack.us/img301/4710/41002iw9.th.jpg
http://img228.imageshack.us/img228/3...0ghzck1.th.jpg
But I just do not know, I guess I would have a total BUM chip if it needed 1.425+ to run at 4.05Ghz
EDIT: Urr I have just been looking at my old sig with my E6750 - E6750 @ 3600Mhz 1.49v (BIOS) 1.44v (Load) - 1.49 BIOS and 1.44 Under load? thats bugger all vdroop compared to whats happening now.. unless the sensors are buggered and I am killing my chip with 1.465vcore.
Huh it is strange with my CPU:Quote:
Originally Posted by HaTaX
Without OC my VID is 1.125, after OC 1.0375 :)
- coretemp 0.96.1
My chip is Q744A698 SLAPL
ArCEIM after setting 1.485v in BIOS , CPUz, EQ, Everest showing me 1.4v - my vdrop is 0.032V and I need this value (1.368v) to run 4 GHz
I've checked something - I've set 1.485v in BIOS, then I restarted my computer, and I've checked in BIOS ("PC Health Status") voltage was 1.39V, in Windows 1.4v - so definitly it is very big offset.
Definitly we need new BIOS from Abit.On my E6600 everything was fine - after set 1.325v in BIOS , windows red 1.28v, vdrop was 0.016v.
giorgos th. maybe it is wrong, but I've tried change this value, and was even worse - maybe I did something wrong :)
Here is my settings:
Need to be careful with GTL voltages, but generally if you're failing Prime or Orthos right away, GTL can help with that and you don't need to apply additional VCore. There's alot of threads out there on it but I found this post to be a good point of reference as it did increase my stability VS what I was "trying" to get done:
http://www.xtremesystems.org/forums/...91#post2716991
Anyway, I've noticed that I've got a good amount of VDroop going on here. .04 max between load and idle. I know that doesn't sound huge, but typically I can see my volts trail off just before orthos fails. Think the mod from johnny bravo would help out in my case? I may be able to minimize the VCore required to hit 4ghz by doing that, maybe I'm just a little nervous about degradation on these chips.... They haven't been out that long yet and I'd rather not kill this one so early on.
it's your motherboard's fault m8 , P35 based ABIT mobos have such a vdroop.Quote:
Originally Posted by HaTaX
I have a good one ... how HIGH is your o/c guys with EIST enabled ? i am interested about it.
What is your settings GTL voltages?Quote:
Originally Posted by HaTaX
I've got standard 63%, after 67% my computer don't want to start.VVT is 1.31v - 1.2v real
BTW - I heard (it is nothing sure) Intel will give new steppings, because it is something wrong with this one.
http://img149.imageshack.us/img149/6...5004400ts1.jpg
Q743A797
Pack date : 01/17/08
My 24/7
Idle 1.43vCore, Loaded = 1.37-1.39vCore
http://img528.imageshack.us/img528/9...39vcoretu7.jpg
http://img127.imageshack.us/img127/4...39vcoread0.png
Max PI on Water.
Set 1.7550 in BIOS for 1.67 vCore IDLE. :(. vCore/vDroop Mod coming soon! :).
http://img122.imageshack.us/img122/7581/9406kd4.png
Chip goes under DI soon :yes:
Unclewebb-Quote:
Originally Posted by unclewebb
Thanks for the info. Very well written and understandable. It is the idle temps that I was concerned about as the peak temps seem low for the stress I was putting the chip under. I'd come to the same conclusion that the chip is a keeper since I had a long way to go before i got anywhere near Tj. I'll now not worry so much about the idle temps thanks to you. :clap:
The gtl options available for asus p5k-e ?Quote:
Originally Posted by Blackwhite
If you flash to bios 1002 they are but i have a feeling they don't work.Quote:
Originally Posted by vera
Changing these settings does nothing for stability on my board. :(
So for better stability I shoul set GTL higher or lower?
It's really hot :p:
http://img508.imageshack.us/img508/2971/43862480ws1.jpg
same on me :confused:Quote:
Originally Posted by Zeus
hmmm... I got my chip in the mail but haven't opened it.. some of these results aren't all the encouraging.. I'm using an ultra 120 extreme with a Q6600 right now that's prime stable at 3.6 GHZ with 1.33 volts in windows, with load temps under 60 degrees, and I don't know if I'm willing to trade those extra cores for only 400 or 500 mhz, and with high volts and temps just to get there. I was counting on 4.3 -4.5 prime stable, and although 4.3 is possible although unlikely, it looks like 4.5 is quite unlikely. 4.0-4.1 isn't enough for me to switch I don't think.
what to do...
4.3 is OCCT stable with 1.48v (real) at a max of 64C using a TR Ultra 120 (not extreme) + scythe fan and AS5. I've tested with orthos for about 1 hour and it was fine. FPO was Q745
I don't know wtf is going.
This 8500 can't finish superpi 1M at over 4570MHz, not even with 1.57V.
I really like to know what's crap, the cpu or the board? :shrug:
Asus made a mess by putting GTL voltage setting bios that does nothing.
AriciU, does changing GTL voltage do anything for you?
Nope. Only things modified for my runs were NB volts at 1.55, fsb term at 1.40, cpu volts at 1.68 (bios), rest was all auto... except ram of course.
I'm currently thinking about getting a 8200 for a htpc build to exchange my T7400.
My plan is to clock it around 3 ghz and undervolt as far as i can.
So can anyone with an average cpu check how low he can get the vcore @ 3 ghz ?
And do you guys think it's worth the change ? I mean regarding power consumption and performance (mainly for HD decoding).
E8400
Batch Q746A377
Ran like a dream on my DS3P. This was OCCT stable. 500x8 1.35v bios
http://img126.imageshack.us/img126/5427/superpikp6.jpg
http://img119.imageshack.us/img119/5...perpi32lb3.jpg
Now I have it in my Asus p5e and I can't get this result back.
I got a 3D Mark 06 run on the p5e
http://service.futuremark.com/compare?3dm06=4984060
Considering putting it back in the P35.
Any one running a 8400 in a p5e???? Can you PM me your settings if you are plz.
OK I have my E8500 running at 4200Mhz, 525x8, water cooled, Fuzion 5.5mm, Laing 18W and a Swiftech 320 with Silent Eagle 2000@ 7v. I Also have a 8800Ultra in the loop, with a full cover EK.
Using 1.45v vcore set in BIOS + 1.30vtt, it idles at:
Ambient temp 24c
core0 33c
core1 41c
Orthos small FFT's Load:
core0 54c
core1 60c
Are those temps OK? Its weird how the cores have such a difference in temps at idle and load, but no matter how I mount the block, it is the same. I am using the D-Tek pro mount system too...... I have an Abit X38 mobo and CPU-Z shows my Vcore to be 1.192v, whats up with that?
Seems to be stable though, no matter what tests/games I throw at it :)
Not a good 1M time, but low voltage for this clock :) (load ~1.39v)
this is on Gigabyte GA-G33M-DS2R :D
http://www.abload.de/img/01raol.png
Quote:
Originally Posted by Chri$ch
E8200 and this G33 chipset seem to run in perfect harmony.
Great result, wish my P5K-E and 8500 liked each other more. :(
same problem with DFI DK P35...
max stable FSB is ~505-510Mhz, max FSB 1M 536Mhz.
On Gigabyte G33 run prime with 550FSB (4.4Ghz) ;)
I don't understand, how can a 4.6Ghz get such a poor PI time? I mean, how is it possible? thats barely better than mine on 4.1Ghz - sexual clock speed on such a low voltage but I dont understand the performance, looking at that I would say I would grab that board the G33M but, hell, need explanation first :DQuote:
Originally Posted by Chri$ch
Here is what i managed to squeeze out of good old air cooling (TRUE).
Chip is Q746A482
9.391s - Intel Core 2 Duo E8400 @ 4815MHz 535x9 1.66v - Aircooled (TRUE) - AxeRam 800 @ 1070 Mhz 5-5-5-15 2.2V
http://www.imgx.org/thumbs/large/11573_tzwlm/4815.JPG
5-5-5-15 timigs, low subtimings (performance level) + Gigabyte board (G33 chipset?) :DQuote:
Originally Posted by ArCElM
Funky: Good work. Maybe someone around here should start a 9 second SuperPi club. :D
It' impressive , p35 boards (mine asus p5k-e) sucks for 45nm , not boot 540fsb with my e8200 Q745Quote:
Originally Posted by Chri$ch
Just got a second e8400 and I'm wondering if I should bother testing it or wait to see which chipset / bios is actually ready for it. I'm thinking P45 lol!
I know p35 and x38 are working but it still seems flaky.
Just how much better would P45 be?
I mean, if it's KT133 >> KT266A bs all over again, I'm gonna be sad.
I have a feeling the secret is in the termination voltage in conjuction with GTL voltage.
These 45nm cpu seem to need other values than Conroe.
The "cheaper" Asus boards have a fixed value for this which is optimized for 65nm cpu's.
Bios 1002 with 45nm cpu support have both these settings in bios but they actually do bugger all.
I now flashed to bios 1004 and it looks like i gained some stability, i can run prime at a tad lower voltage now, Asus probably changed the default GTL voltage value.
I'm gonna play with these settings soon to see if these work now with this bios.
Maybe i'm talking out of my backside but changing GTL value on the DFI P35 helps stability a great deal.
HI guys, i hit a brick wall with my e8400, keep gettin the "FATAL ERROR: Rounding was 0.5, expected less than 0.4" message in prime past 4.3 ghz. aiming at 4.5 but no vcore will satisfy the beast. my assumption so far is that these chips dont need that much vcore to hit sky high but its other settings like cpu ppl, fsbt, and those gtl thing but have no idea what to set.
anyone running this chip on maximus formula?
so far am at 500fsb, stable @ 8.5 multi but when i change to 9 nuthing seems to work
@ 500x 8.5 am 24hrs prime stable blend
1.425vcore
1.5v cpu ppl (bios set)
1.54 fsbt(1.43 actual)
2.3vdim @ 5-5-5-5-3-35-6-3 @ 1200( passed 12 prime + 24hrs blend)
1.65nb
1.18sb
1.55 sb2
any tips on gettin it stable with multi 9 .
rig
maximus formula se (907)
e8400(Q745a372, 12/20/07, 1.1125vid)
2gb crucial ballistix tracer 8500
2x 80gb seagate sata raid 0
2x saphire 2900gt CF(disabled when priming)
Swiftech 220 compact kit(4x scythe 1500rpm fans on rad)
northbridge and cpu loop.
load temps at 1.425(nb39-41, cpu 54-58, cores 58-62 both same, sb 35)
idle 30cpu, 38cores, nb 38, sb 32)
@ 9x 500 tried 1.45, 1.47. 1.48.1.5 and 1.525(load at 63cpu) and nothing seems to work. any tweakin guru's please tip me. thanks.
Did you try to find the max fsb first?
You might 've hit an fsb wall, more people report trouble to get over 500fsb on the maximus.
Try to find the fsb max with multi 6x first.
Adjusting gtls manually helps with stability in my case with the DFI, but it's guess work. I really wonder how it will be possible to have any auto setting for that if they expect to support both 45nm and 65nm. And if it has to be manually set, what should the values be for a given clock and vcore. And does that change from board to board, chipset to chipset?Quote:
Originally Posted by Zeus
I don't think anyone has these answers yet. I'd like to see a board & bios designed exclusively for 45nm and compare to what's currently available.
I'm gonna watch this thread like a hawk regarding those GTL values...lol make another spreadsheet for that?
BIOS/OC/GTL? Or something.
But then we also have X38 vs. P35 difference, woohoo ;D
nama: I realize that Ballistix memory can run at DDR2-1200 and that your memory has been proven to be able to run at that speed but it's still a good idea to reduce the stress on your memory when going for maximum CPU MHz. Try running your memory at 1:1 and see if that helps any. There's no guarantee that all Wolfdales are going to hit 4.5 GHz and be 100% stable.
:) Raise gtl lower vtt then conroe:D .Quote:
Originally Posted by Mav451
I think it was Praz earlier in this thread that mentioned
CPU VTT 1.2
GTL 60/59/60
on a DFI X38
Quote:
Originally Posted by Zeus
i can go all the way to 550fsb. and its stable at 500x8.5 just i want to run multi 9. its just setting i can get right that why i need someone with gtl,ppl knowledge coz alot of guys seem to have success when they set the manually. am working on finding the right conditions but if someone else know it will save me time
I don't think anyone has all the answers either. But the art of overclocking is to be able to find the answers yourself. With the 65nm procs, clocking was very very easy. If you go back to the NF4's, getting a good clock was all down to setting up your ram properly, which a lot of peeps could never do really. With the 45nm procs, us clockers have got to use our brains and clock logicaly. Just dialing in a clock and it works, has gone out of the window. Old school trial and error, is where it's at now.Quote:
Originally Posted by mrcape
thanks and noted. at these speeds 500fsb plus i noticed my memory doesnt like high timings. it reboots when stressing it. but works fine at tighter timings. will work on 1:1 and see if it helps any. will post my findingsQuote:
Originally Posted by unclewebb
just got an 8400, q746a, pack date 1/10, from newegg nj warehouse. ordered yesterday. will stoke it up and report oc in a few hours. from googling, this (q746a) seems like a good batch.
Hmm my e8400 is getting hot! Coretemp gives 63C load, with 1.264vcore. This is on 4005mhz (445x9). Am i doing something wrong, or is this just a hot one? Or coretemp is wrong?
This is all with a Tuniq Tower 120, with stockfan on low.
I don't want to get into discussing the art of anything, but I'll agree that it takes a lot of testing and tuning when something so new is out. I've been doing it going back through NF2, and I always like to leverage forums for info to balance my own findings against.Quote:
Originally Posted by kitfit1
A comprehensive data set on these setups could only help, especially entries for the same chip on multiple boards.
i confirmed there is no degration in my chip, i just lower the FSB termination voltage to 1.2v(from 1.5v) and Northbridge to 1.4v(from 1.55v) and i just passed 30 mins OCCT at 4.32ghz with 1.4v (as before). Also, i am running orthos at this moment, will report back results later.
Yeah I added a column for GTL data in the database. That way we're not entering data into two separate databases. But I have a feeling people's settings will be changing as they figure out what works...so hopefully everyone can update their entries.Quote:
Originally Posted by mrcape
I think the CoreTemps with the E8400 are meaningless. It doesn't matter if it reads 45C or 75C under load, as long as the computer keeps running, all is well. Intel has commented that only the cpu temp is important and they recommend 61.2C as the maximum cpu temp with the E8400. The cpu will automatically throttle-down if temps exceed the maximum.Quote:
Hmm my e8400 is getting hot! Coretemp gives 63C load, with 1.264vcore. This is on 4005mhz (445x9). Am i doing something wrong, or is this just a hot one? Or coretemp is wrong?