Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Some Crafty 16.19 results on my XP 2.44GHz

Author: Robert Hyatt

Date: 08:43:38 02/19/03

Go up one level in this thread


On February 19, 2003 at 02:25:39, Aaron Gordon wrote:

>On February 19, 2003 at 01:45:52, Matt Taylor wrote:
>
>>On February 19, 2003 at 01:33:33, Aaron Gordon wrote:
>>
>>>
>>>>AMD doesn't guarantee 2.3 GHz out of any of their chips. As I recall, you had a
>>>>particularly bad one recently...
>>>>
>>>>Anyway, if AMD doesn't define the limit of stability, who does? Some P4s
>>>>overclock to 4 GHz now, don't they?
>>>>
>>>>-Matt
>>>
>>>Yeah, I got a JIUHB 0301, The TbredBJ's (J being JIUHB) are the $50 chips that
>>>do 2.0 to 2.3GHz, I got 2.15GHz, I wasn't happy so I sent it back. TbredBA's are
>>>much better (A being AIUHB). They WILL do 2.3GHz to 2.6GHz. Get any AIUHB 0242
>>>(week 42 of 2002) or newer and you will be guaranteed to hit 2.3GHz. How do you
>>>know which chip you're getting? Well, on excaliberpc.com they guarantee you get
>>>a 2100+ AIUHB 0304 (week 4 of 2003) TbredBA. Also, yes, some P4's have been
>>>hitting 4Ghz but thats with freon. An XP 2.5GHz is compariable in speed to a
>>>P4-4GHz and you can get the XP to 2.5 on just air. With Freon they'll go up to
>>>over 3GHz.
>>>
>>>From what I've seen/heard AMD/Intel just make chips on a few lines. AIUHB's for
>>>example. Once they hit a better chip consistantly overall or add something new
>>>(like AXIA vs AYHJA, running 20% cooler) they'll add another stepping. Also I
>>>figure they both do this.. lets say AMD is producing all it's highest quality
>>>cores (as both AMD/Intel are doing constantly) and AMD chips can do 2.3-2.6GHz.
>>>This means they can guarantee any of those chips to run 2.3GHz and lower, hence
>>>you get the 2800+ (2.25GHz). I figure once they start producing chips that do a
>>>minimum of 2.4GHz, 2.5Ghz, etc (Maybe future AIUHB 0320's or perhaps some new
>>>stepping alltogether) then they'll be able to release faster chips and guarantee
>>>them at that speed. So what I'm trying to say is as long as there is a chance
>>>they won't do such & such speed, they won't mark them as such. Some chips can,
>>>however.
>>
>>That is pretty much what I said, but it is a different perspective. My point is
>>that some chips -can't-. There is no point in stating that a particular
>>processor can reach a specific nps because I can't simply go buy that processor.
>>I can buy what AMD will sell me. If somebody runs their AthlonXP at 3 GHz under
>>freon and reports their numbers, how does that help me? Likewise if somebody
>>overclocks higher than I can, how does that help me either?
>>
>>>Enricos AIUHB 0246 for example runs 2525MHz at 1.92v air-cooled without
>>>problems. My chip does 2440MHz (will be getting it up a little higher later on,
>>>but you get my point). It is possible if AMD wanted to they could produce an X
>>>number of chips and test them all (too time consuming) and only get the ones
>>>that run 2.6GHz, then they could release those in a limited quantity. Of course
>>>then everyone would get those chips instead and AMD would start losing money on
>>>the chips they made previously where they could have waited and sold an X
>>>ammount of slower chips. I'm not sure I'm explaining it right but I think you'll
>>>understand what I'm talking about.
>>
>>I do. It is still a gamble according to what you say.
>>
>>-Matt
>
>I'm just saying you will get 2.3GHz out of an AIUHB 0242 or newer. That is a
>definite. Now, getting higher is up to the chip. It looks like the average is
>around 2.4ghz or so.. but if you want an absolute guaranteed 2.3ghz then you
>can't go wrong with the AIUHB 0242 through present (latest I've seen is 0304).
>Pretty good deal for $97, and if you want a definite 2GHz a JIUHB (any date) is
>your chip, and again.. $56 makes it a steal. :)


I'm saying you will get a few random bugs by doing so.  I refuse to tolerate
inconsistent
behavior.

The vendors market chips at a particular clock frequency for a _reason_.




This page took 0.02 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.