well, i bought 3847A839, and it needs 1.3V to run @ 21x191 (4011Mhz)
:(
Printable View
well, i bought 3847A839, and it needs 1.3V to run @ 21x191 (4011Mhz)
:(
hello :)
umm i just got batch 3845B271 any word on how this batch is doing guys ?
have not tested much yet but so far have it stable @ 4.3 @ 1.31V with HT ON.
this was taken early in the run which then failed after 8 mins ... but its fully stable @ 1.31V :)
http://img171.imageshack.us/img171/3350/temps43.th.jpg
Got a 3845B259 tonight from Microcenter, it needs 1.37v for 4.2GHz and needs 1.39v VTT for 4GHz uncore. :( The 3849B018 I have does it at 1.34v with 1.34v VTT for 4GHz uncore, but, it acts weird with voltages. Only stable at 1.21v VTT or 1.34v VTT, anything else will blue screen. It is picky about core voltages too, they need to be within specific areas, too high or low and it'll have problems. Should have stuck with my C0, it'll do 4.3.Quote:
Originally Posted by MACMAC
:down:
Batch 3845A935 is very good:up:
http://img33.imageshack.us/img33/9179/xe4000lin.jpg
I've tested two week 45s, B027 and B026, both were very good OCers. The B027 is the highest ranked 920 at hwbot at 5383.7 Mhz. Also it ran prime95 continuously at 4267 Mhz and 1.184 Vcore, on air.Quote:
Originally Posted by no_n|ck
http://valid.canardpc.com/show_oc.php?id=593562
The B026 chip matched it OC wise on air, I never tested it subzero cooling.
amazing 4267@1.184
I am rock solid stable at 4.4Ghz at 1.36V with hyper threading enabled very impressive stepping indeed. I am on water so my temps are very manageable I never get above 70C best chip I have ever owned...
guys, is this considered bad =(
http://img6.imageshack.us/img6/7065/asfs.png
1.3V in bios
About middle of the pack. Mine is about the same :PQuote:
Originally Posted by Drake22
Just got a second 920 for a HTPC /general purpose family computer. Here's how it OCs on dry ice:
http://valid.canardpc.com/show_oc.php?id=596380
This is good for 7th place on HWbot, max OC .
my DICE pot is a hardware store copper end cap and a copper tube to match, cost $8.00
http://img13.imageshack.us/img13/2346/034nxh.jpg
also, I ran 3DM06, 3D game simulations only (I haven't played a CPU rendered shooter game since 1996, so I don't bench that)
http://img29.imageshack.us/img29/465/3d065111.jpg
batch is 3849B016. Not as strong as my first place 3845B027 performance (5383.7 Mhz), still reaching the upper 5200s is pretty good.
3845B223 not a golden one but more than average i think...
@ 4.2 Ghz rock stable under 1.225V
HT ON, turbo ON, speed step ON
All other Voltages on standard and memory @ 1.65 (dominator GT)
pic shows with SPeedstep on :( but just watch the BClcK
http://i40.tinypic.com/11cgrr5.jpg
I'd say these numbers are considerably above average....nice job!
got mine today.
Batch is 3907A338 produced in Costa Rica.
I just did a quick 0C with 200x21=4200 @ 1,28v which seemed to work fine.
Now im currently testing it with CPU's normal voltages so far i'm @ 3570 MHz with 1,16v set as default voltage by bios.
I make some screens later i'mma just to find out what a D0 can do i just recently hit the 1366 with a C0.
i wonder why is my processor using 1.2 volts when under load, and 1.264 when idle?
VDroop I guess.Quote:
Originally Posted by Drake22
I grabbed a 3945B260 but not sure how OCable it is. Just not sure if I should open and try it or sell it for a marginal profit.
Which of these batch may be the best? :confused:
3852a277
3849a824
3851a286
3850a319
3850b177
3908A419
3845B258
I have the 3849a822 myself which clocks pretty good (1.31v 4.3ghz prime and linx stable HT off)
It seems the 3845B027 is the best of all... it makes me think that most week 45 cpu's clock good... so if you take the last one, that 3845b258 then I think you have a reasonable chance of a good overclocker.
week 49 isn't too shabby either but I had no choice lol... :D
well.. couldn't drop the vcore lower than 1.3 :(
i've still got 10 days left of 30 days trade in thing, do you guys think i should try luck with other batch or stick with current?
current:
3847A839
191x21; 4011 MHz
HT on
1.3000 V in bios
1.264 V idle in windows
1.2000 under load
40 C temp idle (Titan Fenrir, fans 45%)
75 C under load with Linx
stable 50 linx passes
if i drop vcore just one lower, it fails linx after 20 seconds lol
mobo: ga ex58 ud5; f7 bios
would you say guys i should send it back and try the other batch, or this is just fine for 24/7 use and i should chill out? :D
also, what do you think about mobo? (i could get rampage II extreme for 75$ more)
thanx for replies =)
sorry for the 3845b258 not in the list are these:Quote:
Originally Posted by Garrett
3852a277
3849a824
3851a286
3850a319
3850b177
3908A419
Anyone have a 3849B040? I just got it from microcenter, no board or anything else yet, but I would like to know what to expect(so i'm not too surprised when it clocks like crap :p:)
Edit: seems like someone has one that is the same time-made cpu as me... but is the number after the batch the CPU number during production or something? In other words, what part of this batch actually defines which "family" the chip is in.
Mine's not a golden one, also would love to have an EVGA Classified :rolleyes:
-Win7 64 bit-
http://valid.canardpc.com/show_oc.php?id=598679
4616 MHz 3DMark 06: 27279 Marks
http://img239.imageshack.us/img239/1...6mhz750mhz.jpg
4669 MHz Super Pi: 8,674 Sec
http://img513.imageshack.us/img513/6...tras124669.jpg
http://img513.imageshack.us/img513/772/resim026.jpg
You can do that on Air with a W3520 with less Volts - 220x21.
CPU score of almost 8400 but using XP Pro 32-bit.
http://img301.imageshack.us/img301/7...20x2018091.jpg
^ Holy crap:shocked: nice one... how did you manage that with under 1.3 volts?
I would love to see prime95 load temps on that.
I have batch 3849B056 at Microcenter, currently stable at 3.9 GHz 1.25v full load
I have not tried any further or lower vcore,
Any one have the same batch ??
I have a 3849B041. It does 4.2 at 1.25v (according to CPUz) and 4.4 at 1.35v, all full stable. 4.4 runs 75c though on water so I backed it back down to 4.2 where it runs 55-60c.Quote:
Originally Posted by rpg711
Someone want to help me pick which chip to keep?
:D:rofl:
I smell cherries everywhere :up:
Both of those 920 batches have seen excellent results!Quote:
Originally Posted by MACMAC
You know who to call when you want to sell one of those...your fellow Canuck :up:
I got a 3851a282 chip and....I suggest you stay away from that one :pQuote:
Originally Posted by darkman
4ghz requires 1.4 in bios (1.344 stressed according to cpu-z)
Xeon 3845A935 is very good:Quote:
Originally Posted by MACMAC
http://img149.imageshack.us/my.php?image=xe4gpr4.jpg
http://www.xtremesystems.org/forums/...&postcount=904
edit
Ive got a 920 3845A chip coming next week....Im hoping for 4.4 w/ less than 1.4v.
you guys are your batch numbers make me lol
If you can get your hands on several 3844A717 SLBEJ (D0) packing date 2/20/09...would you guys do it? I believe those are the very first D0 batch that went through the transition from C0 to D0, the earlier 3844A are of C0 and the later 3844A became D0.
Of course it depends on the price and what your need for 920 D0 is.
I have a 920 D0 Batch 3845B260 SLBEJ, seems to do ok.
http://farm3.static.flickr.com/2442/...dab3e7f4_b.jpg
Scratch the info... 1.4 is if you use x20Quote:
Originally Posted by Dahmer
x21 right now requires 1.28 stressed for 4.0 GHz
Does anybody know what is the official Intel's batch # sequency?
What would be the newest, 385xBxxx or 39xxAxxx??
Picked up a 3844A717 this morning, threw it in the classified and run 20 passes linx at 4.4Ghz with 1.375 vcore, not a bad chip I guess. :shrug:Quote:
Originally Posted by ben805
mine is similar....4.40Ghz only needed 1.368vcore in LinX, but Prime95 crashed in 6hr with 1.387v and needed 1.40v for 10+hr run. 3844A717 runs much cooler than the 3849A that I had.Quote:
Originally Posted by road-runner
The 39 is newer. It indicates the CPU is made in '09. The 38 is made back in '08 week 50 something.Quote:
Originally Posted by WCRF
Yeap, I saw a bunch of them at my local MC.Quote:
Originally Posted by road-runner
Thank you very much for your reply, jcniest5...:up:Quote:
Originally Posted by jcniest5
Hi guys, i can take one of these batch: 3842B037 - 3848A657 - 3851A258. can you tell me any information about them? Which is the best?
Thank's
I guess the 3842 is a C0 so you'd better skip that one, week 48 and 51 hmmmm... not sure about those... tricky choice...
I googled the last two steppings and the 3851 will do 4.2ghz on a DFI X58 board...
Curious to know - did you have a 3849A866?Quote:
Originally Posted by ben805
I saw both at MC - the 3844A717 and the 3849A866 and after doing some research, it seemed that the 3849A866 was able to clock higher with lower voltages than the 3844A
thinking of upgrading to I7 920 from my wolfy e8400
but this PQ5 mobo has been a POS and I have been living on a single 2gb stick for a yr now :(
overall the system isn't bad but what mobo and ram combo is reasonably affordable and should hold a 4ghz stable OC. ???
Buy an older used 920 or will the new stepping be faster???
Yes, it is!!! :up:Quote:
Originally Posted by angelreaper
http://img200.imageshack.us/img200/3479/superpi4776.jpg
I just got a 920, and it's batch 3919A509. So far I'm just being conservative with my Asus P6T Deluxe v2 and running at 160 x 21 @ 1.248v (per CPU-Z). So far running at 73c under 8 threads of Prime95. So far so good! Reminds me of the old Intel Celery days....50% OC!
Just got a 3908A419 hopefully tonight I will see how it performs :)
My CO decided to kick the bucket... so Intel sent me a 3911A207... anyone have any insight as to how this batch may perform?
Buying a core i7 920 for the 1st time, I know very little about this micro-architecture at this stage...
What's the best stepping/batch atm that has the most 'headroom' for OC'ing & subsequently is the most rock-solid/reliable if one doesn't?
I realise there probably isn't one solid answer to that Qn but opinions are most welcome! ;-P
The week 384xB are good clockers, as are quite a few of the 391xB batches. There are always random good and bad ones though. Check out the Batches thread in this forum for a better idea.
Thanks KUP for the advice!
Can get batch 3850a795, would just like to find out if its any good ? :)
Lov Voltage i7 920 D0 / HT ON - Benchmark Test
Test System
i7 920 D0 / Vapochill Ls
Asus P6T Deluxe
Asus 4890 / 1050 / 1100 Mhz
1KW PSU
Cpuz Validate: 4920 mhz 1.4 Vcore
http://valid.canardpc.com/show_oc.php?id=668407
4861 MHZ 1.4 Vcore - SP1M
http://img44.imageshack.us/img44/542...4vcoresp1m.jpg
4803 mhz 1.4 Vcore 3D03 77.6K
4890 1050 / 1100 Mhz
http://img39.imageshack.us/img39/367...mhz14vcore.jpg
4828 MHZ 1.39 Vcore SP1M - HWMONITOR ON
http://img36.imageshack.us/img36/770...1m139vcore.jpg
1.39 Vcore 4702 MHZ VANTAGE PASS
2X ASUS 4890 / 980 / 1100 MHZ
Vantage P score: 22875 Validate: http://service.futuremark.com/compare?3dmv=1346737
http://img188.imageshack.us/img188/7...s4700mhz13.jpg
[QUOTE=EXPANDER;3989919]Lov Voltage i7 920 D0 / HT ON - Benchmark Test
Test System
i7 920 D0 / Vapochill Ls
Asus P6T Deluxe
Asus 4890 / 1050 / 1100 Mhz
1KW PSU
Cpuz Validate: 4920 mhz 1.4 Vcore
http://valid.canardpc.com/show_oc.php?id=668407
4861 MHZ 1.4 Vcore - SP1M
http://img44.imageshack.us/img44/542...4vcoresp1m.jpg
Very impressive, would you care to reveal your p6t bios settings for that :D
What's the best actual D0 batch for OC ?
Thanks.
^^ What kind of test is that? You only used 512MB of memory.
Quickest :D
Stepping# 3845B258
Cpu Cooling = Ultra 120 Copper Lapped + Washer Mod
http://i19.photobucket.com/albums/b1.../4GhzStabe.png
Any good??????
That's very good :D WHat's that addon you're using? Looks sweet :D
Its windows 7 Ultimate with the CPU monitor Gadget :)Quote:
Originally Posted by koksy
wow?!!?!?!?! wtf 4ghz 1.18 volts and priming away...:eek::eek::eek:Quote:
Originally Posted by roller11
nice, thanks, I'll have a look at it :)Quote:
Originally Posted by 4Qman
Its kinda like cheating to get high clocks I say, same as pple still using 3dmarks 03 instead of 06 or Vantage, they dont want to risk crashing.Quote:
Originally Posted by SimpleTECH
That seems like a very good cpu, imo. But why run LinX at all, if your not interested in testing stability ?
And you should think about updating your bios, your mb is throttling allready.
use the whole memory, 20 times and you'll be able to say It's "a little" stable.
HT is off too.
Yes those two core i7 920 batch are wellknown to be great overclocker particularly the 3845B Batch. Yes I was just doing a casual test on them and thought to share with viewers in the forum, more prolonged and stressful test will follow once I could manage time. Could you please explain that throttoling bit, how and why? I have downgraded my bios recently to 1504 for test reason. But would like to know your suggestion. :)Quote:
Originally Posted by Sniper
And who told you that!!?? Stability is a term depends on what you think is stable, days after days priming and linx stability doesnt assure stability in other applications. :yepp:Quote:
Originally Posted by lordmagnum
Lol eagle eye... though I do keep HT on when im @ similar to your speed and voltage for 24/7 settings. :DQuote:
Originally Posted by Brad8988
You can read all about throttling in the P6T Dlx thread, and the special bios from Asus. Basically the motherboard lovers your cpu clock when temp/stress get high enough. This is not what most people here want, as it gives you less performance.
And, the reason why people react to a screenshot of LinX with low mem is quite understandable. When we browse theese threads and see different screenshots, we automatically assumes quite a lot just by glancing at the pictures.
A shot of just cpu-z makes us assume this is max of what the cpu can do, maybe stable enough to boot windows, but not even stable enough to run superpi 1m.
A shot of superpi 1m, not that much more stable than cpu-z, but some.
And the list goes on, with superpi 32m, hyperpi etc.
At the bottom of this list is LinX and Prime, showing them usually means your at a quite stable oc. The longer they have run, the better proof of stability.
Showing LinX with 756/1024 mb of ram, its just a let down. Most people don’t use LinX that way (cause it makes no sense), and have no feel for how far off from a stable oc you are - hence no real info.
And, your right. No screenshot of Prime or LinX can guarantee stability, but it’s an accepted standard for comparing and sharing info about steppings and oc on the net, and most people can get valuable info from it.
To much assumption I would say, as I have already mentiond...casual test...beside for experienced eyes easy to spot it without assumption those are just initial glimpse, heres a little more mature one about an hour which is dedicated to those who wants to see max memory in action... I could go on like this for days but as I did ran test before for days now saving my electricity would be much more sensible thing to do ;)
This is the first chip my 24/7 linx stable settings
http://img11.imageshack.us/img11/683...zzzzzzzznh.png
Is there a good thread about Xeon W35x batches on this forum?
Dunka :-)
And the batch was 3849B not 3845B the last one should be ablt to do a little better, will update later if possible. Thanks. ;)Quote:
Originally Posted by imo
Now that look nice, why didnt you post that in the first place ;)
anyone?
Quote:
Originally Posted by jalyst
Nice results so far imo. Why not include i7 Turbo in your screen shots and take your screen shots while your CPU is at full load?
http://www.fileden.com/files/2008/3/3/1794507/Turbo.zip
Many P6T based boards have a problem with turbo throttling at the speeds you're hitting when air cooled. If your multiplier is not 21.000 at full load on all threads then users need to see this so they can understand what these boards are capable of and what these boards are not capable of.
There's nothing worse than buying new hardware based on what you read in a forum and then discover that it can't deliver the goods. If you do find that your computer is turbo throttling then do some further testing and let us know how high you can get, voltage and MHz, before this becomes an issue. If this is an issue with your board then Asus needs to be made aware of it so they can consider releasing a special bios for your board too.
First post.
FPO/Batch#: 3852A327
http://img200.imageshack.us/img200/4570/tempsn.jpg
And it did pass that ~30min test, temps did not increase from that. Just don't have picture about it.
EDIT: Hmm, I did drop those voltages... Using laptop atm, I'll check if I can find pictures.
Maybe the ELEET screen can help you:Quote:
Originally Posted by jalyst
http://img43.imageshack.us/img43/9831/prime4500.jpg
Welcome to XS! :welcome:Quote:
Originally Posted by Tat3
And nice results, your chip can probably go a little bit further though! :up:
Guess not :(
Quote:
Originally Posted by jalyst
Quote:
Originally Posted by jalyst
There is only one thread about batches - this one. Quite a few Xeons W3520 are listed here.Quote:
Originally Posted by jalyst
Thanks.Quote:
Originally Posted by zalbard
Final volts:
CPU: 1,36V (that 1,32v must have been with some power saving feature on, dunno other explanation)
QPI: 1,26V (got system temperature down)
Getting 4,4GHz stable would need quite much more voltage, thinking about it for benchmarking.
Ram is running at 1,6v, should try 1,65v and a bit tighter timings :)
Just picked up a second i7 920...
FPO/BATCH 3919 A 755 ... :shrug:
Thanks Zalbard, adios.
Quote:
Originally Posted by zalbard