-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
CPU stepping!
- Thread starter skulls
- Start date
- Joined
- Jul 14, 2001
- Location
- Edmonton, Alberta
AIXHB has been up to now the best piece of waffer available from the T-bred line of CPUs.
This AIXIB would prove to be an even newer and potentially even better waffer than the one above stated.
I can speak for everyone and say.. there's only ONE way to find out!
OVERCLOCK THAT THING!
OC-Master
This AIXIB would prove to be an even newer and potentially even better waffer than the one above stated.
I can speak for everyone and say.. there's only ONE way to find out!
OVERCLOCK THAT THING!
OC-Master
- Joined
- Apr 14, 2003
- Location
- Brooklyn
OC-Master said:
- Joined
- Jan 22, 2003
Weeks, sSpecs and steppings are practically useless. This is because not every CPU from every stepping and week is overclocked, so it's impossible to know the best overall. People only know which ones seem to overclock the best. Overclocking is unpredictable because a CPU from a popular stepping/sSpec/week might overclock poorly, and likewise a CPU from an unpopular stepping/sSpec/week might do very well. Less than 5% of CPUs are actually overclocked, so no one can argue that one type overclocks more than another.
Another factor is user error. A certain CPU might be capable of 3.3GHz on air-cooling, but if the user isn't a very knowledgeable overclocker, or one or more of his components are sub-standard, he may only reach 3GHz. This user then decides to go on some forum and write about his "crappy stepping that only reaches 3GHz," when the fact is that the CPU is capable of more. This is why no stepping will guarantee a good overclock, or indicate the speed you will get. Every CPU is different, so the only way to find out how far you can overclock is to try it and see.
As time goes on, manufacturing techniques improve, which is why AMD and Intel are able to increase clock speeds. The engineering improvements from one stepping to a newer one is reason for the increase in overclocking potential. Some overclockers believe that steppings have some hidden meaning for overclocking, but that is just a myth.
Steppings have no affect on overclocking - they simply signify that some type of manufacturing refinement was made. Newer equals better in the overclocking world, but some ignorant overclockers like to argue that older steppings do better than newer ones. They base these theories off of only a few results, and fail to understand that truth that intelligent overclockers already know.
Some people even argue that some older production weeks are better than newer ones for overclocking. False! Newer CPUs based on the same stepping always have more potential than older ones due to the constant rate of improving technology. Weeks themselves do not indicate overclocking potential. I see so many false arguments like "week 31 is better then week 35, but worse than week 24." All you need to do to get the most overclocking potential is to simply buy a new CPU. If you can find a place to get a guaranteed week, great, that way you can simply choose the newest CPU possible. Just don't be fooled by myths that older weeks are better.
As a general rule, the newer the CPU, the more overclocking potential it has.
Another factor is user error. A certain CPU might be capable of 3.3GHz on air-cooling, but if the user isn't a very knowledgeable overclocker, or one or more of his components are sub-standard, he may only reach 3GHz. This user then decides to go on some forum and write about his "crappy stepping that only reaches 3GHz," when the fact is that the CPU is capable of more. This is why no stepping will guarantee a good overclock, or indicate the speed you will get. Every CPU is different, so the only way to find out how far you can overclock is to try it and see.
As time goes on, manufacturing techniques improve, which is why AMD and Intel are able to increase clock speeds. The engineering improvements from one stepping to a newer one is reason for the increase in overclocking potential. Some overclockers believe that steppings have some hidden meaning for overclocking, but that is just a myth.
Steppings have no affect on overclocking - they simply signify that some type of manufacturing refinement was made. Newer equals better in the overclocking world, but some ignorant overclockers like to argue that older steppings do better than newer ones. They base these theories off of only a few results, and fail to understand that truth that intelligent overclockers already know.
Some people even argue that some older production weeks are better than newer ones for overclocking. False! Newer CPUs based on the same stepping always have more potential than older ones due to the constant rate of improving technology. Weeks themselves do not indicate overclocking potential. I see so many false arguments like "week 31 is better then week 35, but worse than week 24." All you need to do to get the most overclocking potential is to simply buy a new CPU. If you can find a place to get a guaranteed week, great, that way you can simply choose the newest CPU possible. Just don't be fooled by myths that older weeks are better.
As a general rule, the newer the CPU, the more overclocking potential it has.
- Joined
- Apr 14, 2003
- Location
- Brooklyn
Akira283-IGN said:
what about the new 1700+ on newegg that are tbred As? =d heeh
- Joined
- Aug 8, 2001
- Location
- New Orleans
skulls said:
yeah
- Joined
- Jan 22, 2003
cherryp00t said:
If it's newer, it's better.
- Joined
- Aug 8, 2001
- Location
- New Orleans
Akira283-IGN said:
Wrong. Look at this thread. http://forum.oc-forums.com/vb/showthread.php?s=&threadid=226434
That's a week 324 which is newer than many, many other steppings that blow it out of the water. UNLESS, this is the new Thorton and the guy doesn't know it.
- Joined
- Jan 22, 2003
brennan77 said:
It's sad that you think I'm wrong when in fact all intelligent overclockers are aware of it.
Edit - it is also possible that AMD engineers fixed the problems with the A stepping, and chose to resume A production over B for various technical reasons. That doesn't prove that the newer A steppings overclock just like the older A steppings.
Last edited:
- Joined
- Aug 8, 2001
- Location
- New Orleans
Even though you know that I have in fact read your PASTED paragraph about 20 times, you still feel the need to let everyone see it again?!
Why not come up with some actual evidences to argue your point. For the record I largely AGREE with you. I disagree with your very poor method of getting that information across. If in fact there is enough evidence to support your claim in the way you state it, everyone here would agree. If you haven't noticed this is pretty much the most well rounded place to gather information on overclocking. People will agree is this is in fact true. According to your standard though, you are the only intelligent overclocker around.
So, with that said, let's see some real evidence. I've been thinking about this lately and I've come up with a lot of evidence to support your claim though not in the absolute rule you have presented. It seems that there may be exceptions that should be noted. My point is that you should interact with everyone here on a rational level and you may get help in validating your point.
Why not come up with some actual evidences to argue your point. For the record I largely AGREE with you. I disagree with your very poor method of getting that information across. If in fact there is enough evidence to support your claim in the way you state it, everyone here would agree. If you haven't noticed this is pretty much the most well rounded place to gather information on overclocking. People will agree is this is in fact true. According to your standard though, you are the only intelligent overclocker around.
So, with that said, let's see some real evidence. I've been thinking about this lately and I've come up with a lot of evidence to support your claim though not in the absolute rule you have presented. It seems that there may be exceptions that should be noted. My point is that you should interact with everyone here on a rational level and you may get help in validating your point.
- Joined
- Jan 9, 2003
- Location
- London, UK
Overclock the damn thing already?
- Joined
- Jan 22, 2003
I actually didn't mean to paste it again, I thought I was in another thread.
I e-mailed AMD for stepping information this morning, and I will report what they reply with. You don't need evidence to understand this argument, though, it's mainly based on logic. The argument explains that there's no thorough, concrete evidence that supports judging overclocking ability by weeks and steppings. Logically then, one shouldn't believe in something with no basis. The argument also needs no evidence that time allows manufacturing improvements, because, to assume otherwise is to claim that manufacturing techniques actually degrade. I think we all know that that's not true.
brennan77 said:
I e-mailed AMD for stepping information this morning, and I will report what they reply with. You don't need evidence to understand this argument, though, it's mainly based on logic. The argument explains that there's no thorough, concrete evidence that supports judging overclocking ability by weeks and steppings. Logically then, one shouldn't believe in something with no basis. The argument also needs no evidence that time allows manufacturing improvements, because, to assume otherwise is to claim that manufacturing techniques actually degrade. I think we all know that that's not true.
Last edited:
- Joined
- Oct 16, 2002
- Location
- Madison Heights, MI
well yeah newer steppings are better and i have always been skeptical of .... get week xxxx or week xxxx because it clocks higher.
The stepping is a very good indicator of how well the chip will do.
I see no reason for someone to go after a AQUCA when they could get a AQXEA(these are barton steppings).
The stepping is a very good indicator of how well the chip will do.
I see no reason for someone to go after a AQUCA when they could get a AQXEA(these are barton steppings).
- Joined
- Jul 14, 2001
- Location
- Edmonton, Alberta
Severian said:
LOL, nicely put!
Saying that steppings do not effect the maximum attainable overclock is FALSE.
lets use the palamino cores. Its like saying a AGOGA could overclock better than a AROIA or AGOIA when EVERYONE knows it doesnt stand a chance. The AROIA had way more head room than the AGOGA.
But, people argued saying that there was very if no difference between the AGOIA and AROIA steppings which was kinda true.
But all in all, the newer steppings 95% of the time always mean more potential. Better quality waffers always lead to higher stable frequencies.
OC-Master
- Joined
- Apr 14, 2003
- Location
- Brooklyn
this guy has been stiring up alot of poop
- Joined
- Aug 12, 2003
- Location
- West Virginia
Personally, I read enough posts and looked at enough sigs to determine that the cpu in my sig should perform pretty well. It did, obviously. All I have to do is look at more sigs and read more posts to find that there are many, many cores out there that have NEWER week codes than mine but can't o/c as high. And many of those cores are being o/c'ed by EXPERIENCED people.
If I, a relative noob to o/cing can push this t-bred to 2400mhz+, why can't more experienced overclockers do the same with newer chips?
It's ridiculous to believe that newer is always better. Especially if the manufactured item has gone through no revisions. You seem to think that manufacturing techniques change daily. I'm sorry to be the one to break this to you, but re-tooling a manufacturing facility every week can get somewhat expensive.
It's actually more likely that batches of materials vary in quality during the manufacturing process. I am certain that strict quality assurance is used when accepting materials for processing, but there must be a variance. No two loads of silicon will be identical. Even something as simple as a technician flipping a switch a few seconds later than he should could cause enormous difference.
It's the same in every manufacturing facility. There is a reason people search the country for just that one certain 350 engine for their restored corvette. It performs better than all the others of the same year and design because something that particular week was unique for some reason. The manufacturer didn't even know it because the difference was so small. But the final product was affected in an unforseeable way.
To believe that newer is better simply because it IS newer, is foolish optimism.
<edit>Now, overclock the damn thing already!
If I, a relative noob to o/cing can push this t-bred to 2400mhz+, why can't more experienced overclockers do the same with newer chips?
It's ridiculous to believe that newer is always better. Especially if the manufactured item has gone through no revisions. You seem to think that manufacturing techniques change daily. I'm sorry to be the one to break this to you, but re-tooling a manufacturing facility every week can get somewhat expensive.
It's actually more likely that batches of materials vary in quality during the manufacturing process. I am certain that strict quality assurance is used when accepting materials for processing, but there must be a variance. No two loads of silicon will be identical. Even something as simple as a technician flipping a switch a few seconds later than he should could cause enormous difference.
It's the same in every manufacturing facility. There is a reason people search the country for just that one certain 350 engine for their restored corvette. It performs better than all the others of the same year and design because something that particular week was unique for some reason. The manufacturer didn't even know it because the difference was so small. But the final product was affected in an unforseeable way.
To believe that newer is better simply because it IS newer, is foolish optimism.
<edit>Now, overclock the damn thing already!
- Joined
- Jul 22, 2003
- Location
- Odessa Tx
Who is to say that a new revision is even better? Sometimes things go wrong. All the simulations and theories and experiments and things can show that a new design will kick the old one into next millenium but they make it and whadoyouknow... it doesn't do better, it's actually worse. The best way to find a good overclocking chip is just like Cuda said, look at what is giving the best results for other people and get the same thing. If you wanna spend some extra money on an un-proven stepping and further the community then I applaud you but otherwise the experiences of others is the best way to find a good chip.
- Joined
- Apr 14, 2003
- Location
- Brooklyn
yea i've gotten a 1700+ AUIGA from newegg with like a date code of 0332..
i have a 1700+ dlt3c JIUHB from scv 0319.. guess which one performs better ???
i have a 1700+ dlt3c JIUHB from scv 0319.. guess which one performs better ???
- Joined
- Jan 22, 2003
Cuda said:
It's unfortunate that you wish to remain ignorant and deny the truth. If you read carefully, I never said that the manufacturing techniques improve daily. It can take weeks, even months for significant improvements to be made. That doesn't change the fact that they're constantly improving, not degrading.
That's exactly my point: every CPU is unique, so one should not believe myths that one week or steppings is always better while another is always worse. Just buy a new CPU and overclock it. Simple, eh?
Then by your terrible logic, it's MORE foolish to assume that one can determine which weeks and steppings are better than others by analyzing less than 5% of all the CPUs in the world. Some of these overclocks aren't even done properly or to their fullest extent, which makes it even more foolish.
You need to actually listen here. Not every CPU from every stepping and week is overclocked, so it's impossible to know the best overall. People only know which ones seem to overclock the best. Overclocking is unpredictable because a CPU from a popular stepping/sSpec/week might overclock poorly, and likewise a CPU from an unpopular stepping/sSpec/week might do very well. Less than 5% of CPUs are actually overclocked, so no one can argue that one type overclocks more than another.
Another factor is user error. A certain CPU might be capable of 3.3GHz on air-cooling, but if the user isn't a very knowledgeable overclocker, or one or more of his components are sub-standard, he may only reach 3GHz. This user then decides to go on some forum and write about his "crappy stepping that only reaches 3GHz," when the fact is that the CPU is capable of more. This is why no stepping will guarantee a good overclock, or indicate the speed you will get. Every CPU is different, so the only way to find out how far you can overclock is to try it and see.
The overclocks that various people get are not sufficient in any way to prove that one CPU is a "better bet" than another.
Similar threads
