Core i7/X58 Overclocking Thread

[Avan]

A and B are not steppings. The stepping on all of these chips is the same, C0. A/B stands for something else entirely...

I had a full explanation from Intel somewhere what the numbers mean, but I cannot seem to find where it is now...if I do I'll post it.

This doesn't answer the A/B thing, but here is how to read a batch number.

Example: L707A723 ---

1st letter or digit = plant code: (Malaysia)
0 = San Jose, Costa Rica
1 = Cavite, Philippines
3 = Costa Rica
6 = Chandler, Arizona
7 = Philippines
8 = Leixlip, Ireland
9 = Penang, Malaysia
L = Malaysia
Q = Malaysia
R = Manila, Philippines
Y = Leixlip, Ireland

2nd digit = Year of production: (2007)

3rd & 4th digits = week: (7th week )

5th - 8th digits= lot number: (723)

10th - 13th digits = serialization code (---)

[loonym]

I believe you're over-simplifying here, that system isn't stable.

I suppose you could call it unstable if you only want to play prime all day. My systems are tuned for Boinc WCG.

[Splave]

I suppose you could call it unstable if you only want to play prime all day. My systems are tuned for Boinc WCG.

Amen brother

[dejanh]

This doesn't answer the A/B thing, but here is how to read a batch number.

Example: L707A723 ---

1st letter or digit = plant code: (Malaysia)
0 = San Jose, Costa Rica
1 = Cavite, Philippines
3 = Costa Rica
6 = Chandler, Arizona
7 = Philippines
8 = Leixlip, Ireland
9 = Penang, Malaysia
L = Malaysia
Q = Malaysia
R = Manila, Philippines
Y = Leixlip, Ireland

2nd digit = Year of production: (2007)

3rd & 4th digits = week: (7th week )

5th - 8th digits= lot number: (723)

10th - 13th digits = serialization code (---)

I believe that was the same kind of information that I had, but I think mine was using the i7 numbers as the example. The A/B### identifies the lot. I am personally not sure that there is anything particularly different about A vs. B lot chips as my replacement from Intel which is an A lot chip is pretty good when it can do 4.2GHz with HT on using only 1.38V - 1.39V on pure air (it would likely be lower on water or better as I can right now lower volts if I toss the computer out onto the balcony ). I think it is a week 36 chip for me, but I posted this in the Batch # thread.

[The-Fox]

A and B are not steppings. The stepping on all of these chips is the same, C0. A/B stands for something else entirely...

I had a full explanation from Intel somewhere what the numbers mean, but I cannot seem to find where it is now...if I do I'll post it.

I know, I was referring to batch#.
I would really love to know what does the A/B thing stands for.

[ReverendMaynard]

I believe you're over-simplifying here, that system isn't stable.

Prime stable = prime stable.

real world stability = real world stability.

How is that over-simplifying? Seems perfectly legitimate in theory and execution. If prime actually did something other than set off premature electro-migration then I'd be all for it.

I use Linx to shut people up and it's fast

[Avan]

I know, I was referring to batch#.
I would really love to know what does the A/B thing stands for.

Yah, I think alot of us want to know that. Maybe Drwho? will pop in here and share? I know, won't happen, but would be a lot cooler if it did!

[dejanh]

Yah, I think alot of us want to know that. Maybe Drwho? will pop in here and share? I know, won't happen, but would be a lot cooler if it did!

Avan, you posted it yourself. Intel codes their chips the same across the board. Digits 5 - 8 identify the lot.

So an...

3836A950 chip would be Costa Rica, 2008, week 36, lot A950.
3836B950 chip would be Costa Rica, 2008, week 36, lot B950.

It could very well be that the two lots get filled by two different sources and it just happens that those sending chips to lot B happen to also be a bit more careful with them. But then again, some of the highest clock/lowest voltage chips have been from the (lot) A. Even my chip now, like I said before, does 4.2GHz with HT on at just 1.38V on air.

[Avan]

Avan, you posted it yourself. Intel codes their chips the same across the board. Digits 5 - 8 identify the lot.

So an...

3836A950 chip would be Costa Rica, 2008, week 36, lot A950.
3836B950 chip would be Costa Rica, 2008, week 36, lot B950.

It could very well be that the two lots get filled by two different sources and it just happens that those sending chips to lot B happen to also be a bit more careful with them. But then again, some of the highest clock/lowest voltage chips have been from the (lot) A. Even my chip now, like I said before, does 4.2GHz with HT on at just 1.38V on air.

I guess I was just hoping for an official answer the difference between the two versions of the lots. It was pretty clear at first the B chips were the clear winners, but now the line between which lot version OCs better is beginning to blur. I just find it odd the B lots are a lot rarer than the A lots. Ah well, maybe we'll learn why some day.

[Truckchase!]

Prime stable = prime stable.

real world stability = real world stability.

How is that over-simplifying? Seems perfectly legitimate in theory and execution. If prime actually did something other than set off premature electro-migration then I'd be all for it.

I use Linx to shut people up and it's fast

You and a couple of the other guys are turning the term stable into something intangible that you can bend to your desires. Having no scientific standards doesn't make you more logical than those who insist upon them. Your standards are fine for you, but in a scientific community there is need to set a benchmark by which claims can be validated. Your comment of "to shut people up" is in itself quite revealing.

I don't want you to change your standards, I just want you to stop trying to lower others'. Thank you.

[JWMc]

I believe you're over-simplifying here, that system isn't stable.

Which was my point: a properly run LinX x 20 isn't enough to ensure 24/7 full load stability. I am still increasing VTT chasing stability with triple channel under Prime blend. I am up to 1.4350 VTT at the moment. The system made it past an hour at the previous two VTT voltage settings in the bios and blue screened.

I am happy that I am getting LinX x 20 and 11 hour Prime blend stability at only 1.2750 VTT when using just dual channel memory. So far my Prime blend max temps have not increased at all from increasing VTT or adding the third memory module.

[ReverendMaynard]

You and you cohorts are turning the term stable into something intangible that you can bend to your desires. Having no scientific standards doesn't make you more logical than those who insist upon them. Your standards are fine for you, but in a scientific community there is need to set a benchmark by which claims can be validated. Your comment of "to shut people up" is in itself quite revealing.

I don't want you to change your standards, I just want you to stop trying to lower others'. Thank you.

Who said Prime is a standard or the holy grail of stability especially when 100 people will tell you 100 different time frames that they call it stable. Then you have small fft, large fft and blend. Which one is the right one for i7? In reality, your system needs to be stable enough to run the applications that you run. Making anything a standard is impossible because of the rapid advancement in technology is 100x faster than the advancement of these worthless stability programs.

For anyone to argue against real world stability, that is revealing.

My "cohorts" as you coin them are unfortunately few and far between.

[Truckchase!]

For anyone to argue against real world stability, that is revealing.

Define real world.

[T_Flight]

Gaming, Folding, Contiuous use of everyday programs without serious errors, OS Running Smoothly without corruption...you know, everyday use. Real programs that people actually use.

[T_Flight]

I went for it today, and got mine up to 4GHz Benchmark Stable. It was a piece of cake. Not gonna do overnight stability on it due to heat issues. I will be taking care of the heat with Watercooling so there is no real need to test for stability yet. Still playing with things, and will be for a good while. Below is the screen.

[ReverendMaynard]

Gaming, Folding, Contiuous use of everyday programs without serious errors, OS Running Smoothly without corruption...you know, everyday use. Real programs that people actually use.

Exactly.

[Splave]

looks great T, what is your bios voltage? like 1.35?

[ReverendMaynard]

Impressive T, love Fritz Chess. You ever try Arena? Now that is a grea real world stability test

[jcool]

Which one is the right one for i7?

Blend, definitely
Forget about the rest on i7, just a waste of CPU cycles.

[T_Flight]

Can't remember the exact number, but I input 1.35. It always ends up being 1.35xxx volts. It's the RIIE so whatever it goes to when you manually input 1.35 and hit enter. That's stock cooling now. Oh, and most of us already know this, but watch the dang heat with Stock Cooling. On stock cooling it's a completely diffeernt beast than on good air or water. Don't go too out of control on voltage on stock cooling and run programs that load for long periods of time without watching it and being right there with it. Some may throttle. Mine didn't on that run, but it was hair from it. They will throttle at 90.

I am NOT gonna do that all the time though. I might on water, but that's basically a "Showtime" run. I would not fold with that right now. I'll get water on it. I'm slow sometimes on stuff, but I will get watercooling on that system...that's a promise.

[T_Flight]

Impressive T, love Fritz Chess. You ever try Arena? Now that is a grea real world stability test

Yeah, but I'm almost scared to try it right now with stock cooling at that speed. I've already done 3.8GHz runs. Maybe after a couple of Jack Daniel's shots later I might get a wild hair and go for it. If that thing gets to 89 I'm gonna have to back out of it. I wanna be careful I don't hurt it until I get water on it. This CPU feels like a good one.

[ReverendMaynard]

Can't remember the exact number, but I input 1.35. It always ends up being 1.35xxx volts. It's the RIIE so whatever it goes to when you mnually input 1.35 and hit enter. That's stock cooling now. Oh, and most of us already know this, but watch the dang heat with Stock Cooling. On stock cooling it's a completely diffeernt beat than on good air or water. Don't go too out of control on voltage on stock cooling and run programs that load for long periods of time without watching it and being right there with it. Some may throttle. Mine didn;t on that run, but it was hair from it. They will throttle at 90.

I am NOT gonna do that all the time though. I might on water, but that's basically a "Showtime" run. I would not fold with that right now. I'll get water on it. I'm slow sometimes on stuff, but I will get watercooling on that system...that's a promise.

For stock that is still impressive. Safe to say that you're pushing the limit on it though, but there is a ton of potential with your setup.

[T_Flight]

hehe Yeah, I know it's pushing it. Has anybody ran 4.2GHz on stock air yet? I'm just curious. If they did what were the temps? I don't think I can bench any higher than this on stock cooling.

I do think the chip has more in it with better cooling though. I would expect it to since it's a 965. I know this much, for what it cost, I could've bought the same weight in gold so I would hope it does. hehe

[Truckchase!]

Define real world.

Gaming, Folding, Contiuous use of everyday programs without serious errors, OS Running Smoothly without corruption...you know, everyday use. Real programs that people actually use.

Exactly.

And with that I bow out of this divergent conversation confident that this statement alone has proven my point. Being a scientifically minded person (or so I'd like to think) I cannot accept this criteria as proof, as it is based more on coincidental data and emotional response than it is of a reproducible test scenario designed for our purpose.

That having been said, it's obvious that you folks who do clearly understand that the criteria is undefined, and you are comfortable with that. I cannot fault your choice, as it is your own to make.

Hoping that I have in the least raised food for thought, I apologize for the divergence.

[T_Flight]

I also agree with you too. I do use Stability testing. I just haven't yet on my new one. There are kinda 2 schools of thought on it. When I'm just messing with a system I don't worry about it, but once I really get a system setup for everyday use and rely on it I do use all kinds of different programs. I've used Prime95 overnight in the past as my staple for stability. Now, I would use Intel Burn Test, and maybe OCCT and Chess also along with games and Folding and pretty much everything.

I wouldn't ever run a system for mission critical stuff unless I was sure it was stable. None of them can be 100% stable. It's just not possible. Errors happen that we never see or know about even after one's been tested, but if they've been tested with alot of different programs it's a pretty good bet the system is stable *enough*. Obviously you can't gurantee that a machine will work 100% error free 100% of the time though.