i5-2500K/2600K batch number and o/c results

[C-N]

So, try to get batch with the smallest last three number as possible, correct?

Was going to ask the same question, I remember older chips having wafer number (apparently) & it was thought better to get a lower number. I have had very low number chips & found them no better than mid no chips so didn’t go much on it. SB now has a number printed on the board again maybe this is the wafer number & for fun I would like to know which number is from the centre of the wafer.... is it like 0 or 500 in a 1000 scale for arguments sake???

[Falkentyne]

Well mine said 998 for the last three numbers, and it has no problem doing 5 ghz at 1.39v bios (1.44 load), and 1.25'ish at 4.5 ghz...

998 should be the worst of the worst, from that logic...

[Sam_oslo]

The wafer has an impact too, but I believe Intel can enforce a QA on these chips, pretty firmly.

These variations in OC-headroom seams to be deliberate. Intel tend to produce good chips in a short periods ( it was week 46 to 49 for i7-45nm), but then it starts going done, usually. In last round, Intel had kept some week-48-batch# in storage for over 1 year, and used them for W3565 (If I remember the number right, but anyways, those chips were meant for a new push in sales, with good OC-headrooms)-

I believe the week-of-production-number and A, B, or C is more important than that wafer-quality theory.

[Chri$ch]

2600K L042B076



actually the maximum on air

[Astennu]

OMG what a nice cpu!!! Chri$ch !!!

What kind of memory are you using?

I can only get: 933 : T2-8-10-9-22-90-10-5-5-5-26-6-3-36-36 out of my Corsair XMS3 2x4GB PC16000 (2000MHz 1.65v). (higer CPU,IO,SA,Mem or PPL voltages did not help to get T1 or other tighter timings stable.)

Or 1600 T1-7-8-7-20 but the 1866 setting is faster and lower in latency.

I also need 1.28v for a stable 4.5 :'( I got a L042B074. I hope my next 2600K will do better because this sux..... I want 1866 8-9-8 T1 and lower voltages. I cant even run 5.0 1.45v stable. its aint the worst CPU but there are a lot that do way better :P

I just took delivery of my CPU, which may be the newest in this thread. To be built with a GD65 and 4Gb XMS3 when I get home from work:

i5 -2500K
L049A796

Anyone know what the significance of the A is? Most other CPUs in this thread have a B in that place.

I wonder what mem speeds you will hit. I'm not doing verry well with the XMS3 on a GD65

Well mine said 998 for the last three numbers, and it has no problem doing 5 ghz at 1.39v bios (1.44 load), and 1.25'ish at 4.5 ghz...

998 should be the worst of the worst, from that logic...

True but there is alway's a chance that you get a chip with almost no defects from the outer circle. Your chip is indeed better then average.

[Sam_oslo]

Batch# L040B165

2600k @4.5GHz(45x100) 1.27v 1-hour Prime Small FFT

[Chri$ch]

@ Astennu

i use 2 kits of these kits

http://geizhals.at/deutschland/a606014.html

[st0ned]

Good ? :




What about :



I'm still testing, but even passed 1run with this volts... I almost had an orgasm

[Chri$ch]

why do you set LinX only to "4 Threads"? That doesn't say so something about stability.

if you run it only like this you can disable HT in Bios...

[Sam_oslo]

Good ? :

http://img255.imageshack.us/img255/7267/apetite.jpg[/IMG]


What about :

http://img109.imageshack.us/img109/2932/5100x.jpg[/IMG]

I'm still testing, but even passed 1run with this volts... I almost had an orgasm

Yeah good one, that's a really good vCore for that speed.

You where posting ~1.425v for @5Ghz before. What happen?
Have you discovered something new, that you can share. or you where pushing too much vCore without testing with less before?

EDIT:
Oh I see Chri$ch's point, he just posted while I was writing this. You are running the LinX at 50%, that doesn't mean stable.

[bro20000]

Good ? :




What about :



I'm still testing, but even passed 1run with this volts... I almost had an orgasm

Nice! Am hoping to get the same results on asus maxium IV has i do on msi p67a-gd65!Have you tryed 1.34v 5ghz!

[st0ned]

EDIT:
Oh I see Chri$ch's point, he just posted while I was writing this. You are running the LinX at 50%, that doesn't mean stable.

why do you set LinX only to "4 Threads"? That doesn't say so something about stability.

if you run it only like this you can disable HT in Bios...

Guys guys, LinX with SP1 and AVX, runing in 4 threads is harder then LinX on 8 threads. I've allready posted here ( I think ) something about it.

Watts with 8 threads will go something like 110W, with 4threads will go as high as 136W. It also shows up on temps, where I get somthing like +8ºC.

The post I was talking about, it was with a different CPU but still true.

Yeah good one, that's a really good vCore for that speed.

You where posting ~1.425v for @5Ghz before. What happen?
Have you discovered something new, that you can share. or you where pushing too much vCore without testing with less before?

I changed CPU. Both have the same batch but this second one has been proving it self a better one.

Nice! Am hoping to get the same results on asus maxium IV has i do on msi p67a-gd65!Have you tryed 1.34v 5ghz!

Thanks, will do. But with 5.1 down the line I'm putting my efforts for this mark now^^

[Sam_oslo]

Guys guys, LinX with SP1 and AVX, runing in 4 threads is harder then LinX on 8 threads. I've allready posted here ( I think ) something about it.

Watts with 8 threads will go something like 110W, with 4threads will go as high as 136W. It also shows up on temps, where I get somthing like +8ºC.

The post I was talking about, it was with a different CPU but still true.



I changed CPU. Both have the same batch but this second one has been proving it self a better one.

If 4threads-LinX is harder than 8threads, then something is really wrong with LinX.
I don't know what's going on with LinX, but it was the same stuff with 980x too, until they fixed it later.
Did you try Prime Small FFT with 8threads? That should give a better indication, then.

Maybe your new CPU is better, and hopefully it is, but you wouldn't know before testing it with Prime.

[bro20000]

4.95Ghz 1056Mhz 6-10-6-27 1T Linx stable!




Before Ram slot one died!

[aytacaksel]

sandy bridge die size: 20.5mm x 9.9mm
sandy bridge wafer: 300mm

[Chri$ch]

@ st0ned

i don´t see a difference..

LinX 4 Threads = ~215-235W
LinX 8 Threads = ~225-238W
Prime 26.4 Custom = ~224-234W

And yes, i have SP1 installed and get ~120G/Flops @ 4.5GHz with only 4 Threads

[Astennu]

@ Astennu

i use 2 kits of these kits

http://geizhals.at/deutschland/a606014.html

Thanks. So it seems 4 modules work better then the 2x4gb setups. It was different in the past...

@ st0ned

i don´t see a difference..

LinX 4 Threads = ~215-235W
LinX 8 Threads = ~225-238W
Prime 26.4 Custom = ~224-234W

And yes, i have SP1 installed and get ~120G/Flops @ 4.5GHz with only 4 Threads

Its all useless.... I was prime stable (custom 960K and i stopped it myself) for more then a hour. And when playing SC2 it still crashed on me. Its hard to really stress test SB.

[Lanek]

2600K L042B076



actually the maximum on air

Look like im lucky, same Batch here ..

[jcniest5]

I wonder if Chrish's OCing is also possible on the P8P67 Pro also or is the Maximus IV Extreme really that good in OCing any CPU?

[jcniest5]

sandy bridge die size: 20.5mm x 9.9mm
sandy bridge wafer: 300mm

I still don't understand how the wafer works in term of better or worse OCing. Someone please explan the picture...

[Chri$ch]

Thanks. So it seems 4 modules work better then the 2x4gb setups. It was different in the past...



Its all useless.... I was prime stable (custom 960K and i stopped it myself) for more then a hour. And when playing SC2 it still crashed on me. Its hard to really stress test SB.

have you tested only the 960K? If yes, this doesn't say something about stability.

start prime (26.4 x64) and make a custom run and not only with 960K!

[roadie]

I just took delivery of my CPU, which may be the newest in this thread. To be built with a GD65 and 4Gb XMS3 when I get home from work:

i5 -2500K
L049A796

Anyone know what the significance of the A is? Most other CPUs in this thread have a B in that place.

/Edit/ 45x100 set in BIOS seems stable at 1.25V. CPU-Z shows 1.225V idle and 1.264V under Prime95 load. MSI Control Center has 1.2150V displayed. All other voltages on Auto with Vdroop control on low.

Over 2 hours Prime 95 stable at 4.6Ghz. 1.255V set in BIOS. The voltage issue above still in effect.

[Lanek]

have you tested only the 960K? If yes, this doesn't say something about stability.

start prime (26.4 x64) and make a custom run and not only with 960K!

Can i ask you what was your QPI on 5.6ghz ? I m new to this Intel new cpu's and really i m a bit lost with the QPI voltage, with I7950 or 980, the qpi was around 1.355 for my ram with a 4.4ghz oc, and with it, it boot on 1.05V and stable @5ghz ( same ram used ) .. really low... i don't know how it react on increase it. or not.

[aussie-revhead]

Guys, are there many of you getting the fabled 50x or higher out of your cpus? Chri$ch has a beauty but how many more out there are so good? Is there a good batch to watch for? Do we have a thread here discussing this? My D1 ES is good for 47x so while its not the worst its far from the best too.

Any help would be great.

[flesheatinvirus]

My L040B208 can do 5600 also pretty easy on water.