Sign in with
Sign up | Sign in
Your question

Whats the difference?

Last response: in CPUs
Share
April 21, 2002 5:52:07 AM

I am very confused concerning the differences between the different Athlon XP chips (1700+ 1800+ 1900+...) I have read several posts that suggest that they are the exact same chip, however, they are locked to their respective multipliers. If this is true, then wouldn't an unlocked 1700+ with the same multiplier as a locked 2000+ perform exactly the same (with repect to heat, speed, stability)? If so, why are people apprehensive about OCing athlon xp's? Wouldnt it be logical to buy a cheap 1800+ and OC it to 2000+ levels, only adjusting the multiplier (not changing FSB speed), and resultantly, end up with a 2000+ CPU?
I might be totally off...Any guidance is appreciated.
Thanx

More about : whats difference

April 21, 2002 5:57:53 AM

good logic for the most part. i used to think the same thing.
they are basically the same thing...but because of imperfections, the each chip can do only a certain speed stable. at least stable to their specifications.
when they make the chips at the factory, they test them to see what speed they run at...and then they lock them. sometimes there is variance, and there is overhead which can be taken up by overclocking...and sure, you can buy a cheaper cpu, and overclock it, but sometimes you will have to increase the voltage, and that creates more heat. and, since it varies from cpu to cpu, it is almost like a crap shoot on how much of an overclock you will get from a cpu.
but my friend has his xp1700+ clocked the same as a 2000+...and has no problems...i dont exactly remember what he overclocked, the fsb, or the multiplier...but he has a "2000+" now.


-DAvid

-Live, Learn, then build your own computer!-
a b à CPUs
April 21, 2002 8:31:00 PM

Basically, most of the slower CPU's are very similar, but the 2000+ is getting up toward the end of what that core will do at 1666MHz, so it's probably a better processor than the lower speed units. A near perfect core tops out around 1800MHz with standard cooling, anything less than near perfect will be assigned a lower frequency.

What's the frequency, Kenneth?
Related resources
April 21, 2002 9:01:33 PM

Crash, they have aircooled 1900mhz axps now, I would say a near perfect core tops out at 1900mhz myself.


www.overclockers.com

check cpu database.


MOST overclocked axps can hit 1800 with a good aircooler.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 21, 2002 9:18:18 PM

I see two 1900+ XPs at 1980mhz aircooled. I wouldn't have believed it.

<b>I have so many cookies I now have a FAT problem!</b>
April 21, 2002 9:26:48 PM

That would require quite a fancy aircooling, but needless to say, if the OC was big, such as from 1800+ to PR ~2350+, then this is definitly a nice OC, and if someone had done it here, they probably may be able to have completed FatBurger's challenge. Not sure though, since it requires the stock speed and its price as well as total price of processor, and if it runs stable at that speed, as well as with all that, that it outperforms Fatty's OC monster. If it does, and it costs less too, then indeed the theories roaming around Athlons being hopeless at high speeds or that their dead limit is 2GHZ (such as Raystonn's claim) are utter false, and that they CAN, with some love, give you the power house. I wonder if Barton with 512K L2, would lower the max speed due to another core addition, if not then it would be a very nice OCer plus the new performance. I reckon seeing 3d Studio Max highly benefitting from more cache, so I don't see why Athlons would not benefit as much as Northwoods.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
a b à CPUs
April 21, 2002 11:05:32 PM

I'm speaking of what AMD rates them at!!!!!
When AMD marks a CPU it has to be able to run at that speed with a STOCK cooler at 100F ambient temperature. I'm guessing at their ambient temp requirement, but at any rate, it has to work in an average system with stock cooling.

What's the frequency, Kenneth?
April 21, 2002 11:18:16 PM

Crash, thats true, however amd has in the past and can, change the stock hsf to handle the higher load, the core is at 1.73 right now stock, and it most likely will hit 1.8 STOCK, add to this the 2 1.97ghz aircooled axps on overclockers.com(yowza) and you have a core with an average top aircooled speed of ~1800, not 1.66.


:-)

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 21, 2002 11:45:33 PM

I'm afraid I completely agree with Crash, Raystonn, et al. The Athlon XP core is already at or near it's limit. You just won't see an Athlon XP at 2000mhz (2400-2500+ PR rating) not stock clocked. If it takes a rare overclock to get to that point then you certainly won't see it in production models.

I don't know anything about 3d Studio Max but in your opinion which Athlon XP matches a Northwood 2.2 3d Studio Max benchmarks.


<b>I have so many cookies I now have a FAT problem!</b>
April 22, 2002 3:12:22 AM

An AXP at PR2100 is more than enough to equal or outrun a 2.2GHZ. Its FPU remains unmatched, and it's not by upping the MHZ for P4s that will show it being strong in 3d Studio Max rendering, no siree bob. Perhaps SSE2, but until they strenghen the FPU, it will remain beaten by Athlons.
BTW there are already demos of a PR2800 (2.2GHZ) CHIP by AMD.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
a b à CPUs
April 22, 2002 3:22:51 AM

Quote:
A near perfect core tops out around 1800MHz with standard cooling, anything less than near perfect will be assigned a lower frequency.

So you might have forgotten what I originally said, you're forgiven.

What's the frequency, Kenneth?
April 22, 2002 3:39:02 AM

Cool.

I know some early Athlon XP's worked in a dual configuration. Is this still true? Can Athlon XP 2100+'s be used on a dual mobo?

<b>I have so many cookies I now have a FAT problem!</b>
April 22, 2002 3:28:03 PM

Quote:

In reply to:
--------------------------------------------------------------------------------

A near perfect core tops out around 1800MHz with standard cooling, anything less than near perfect will be assigned a lower frequency.



--------------------------------------------------------------------------------

So you might have forgotten what I originally said, you're forgiven.


What I meant kemosabe is that a "near perfect" core would top out closer to 1.9ghz , whilst an average core runs around 1.8.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
a b à CPUs
April 22, 2002 4:25:23 PM

I'm under the impression that the rare few that will go that fast with stock coolers in normal environments are getting marked as 2100+ versions, clocked at 1733MHz.

What's the frequency, Kenneth?
April 22, 2002 4:36:16 PM

Hmm, the 2200+ (1.8GHz) is coming out on stock cooling! With a bit extra cooling and voltage, I'm sure 1900MHz or 2GHz would be possible.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 5:50:06 PM

How can one test the potential of any given xp chip. I have an 1800+ coming in the mail and was wondering if there is any way to test that specific chips potential.
April 22, 2002 8:43:28 PM

Running TOAST maybe?
Any recent chip usually CAN do 100MHZ OC if properly configured.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
a b à CPUs
April 23, 2002 1:53:21 AM

Still falls just at the edge of my estimated 1800MHz for near perfect chips with stock cooling in standard environments. AMD is milking the cash cow on this one the way Intel sat back and milked their Willies.

What's the frequency, Kenneth?
April 23, 2002 1:54:34 AM

Quote:


Still falls just at the edge of my estimated 1800MHz for near perfect chips with stock cooling in standard environments. AMD is milking the cash cow on this one the way Intel milked their Willies.


Hehe, you're right.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 23, 2002 11:28:44 AM

Whats the sick facination with stock cooling crach ;-).

There is a difference between aircooling and stock cooling.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
a b à CPUs
April 23, 2002 12:20:43 PM

Basically the original post ask if there was a difference between speed grades, and my answer is that the ones rated at the highest speeds are most likely the few that pass all the test at those speeds. And part of that test is that they can't produce so much heat as to overheat using a retail boxed cooler at a predetermined ambient temp. You can add an Alpha or Swifty, AMD can't.

What's the frequency, Kenneth?
April 23, 2002 11:34:12 PM

Actually crash, the die arent tested with any heatsink, if amds chips cross a thermal threshold which requires a better hsf to run, they will upgrade the stock hsf, if im making sense.


Amd decides what heat is acceptable for stability, and as processor speeds increase, that amount increases as well.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
a b à CPUs
April 23, 2002 11:49:58 PM

If the die crosses the thermal threshold for the standard boxed heatsink, they will set it at a lower speed where it no longer has problems.

What's the frequency, Kenneth?
April 23, 2002 11:58:20 PM

I think what Matisaro is trying to say is that AMD can just release a new HSF with the retail package if they want to release a Palamino at 2GHz.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 24, 2002 12:09:59 AM

Crash, not if they planned on ...... exactly amd_man!


The hsf thermal standard is not a set amount, it is flexable, when amd has cores which will run stably at a higher speed they wish to release, but will require slightly beefer aircooling they will change the retail hsf spec to meet compliance, they will not release slower chips with older hsfs if they WANT to have a faster cpu out.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
a b à CPUs
April 24, 2002 3:09:19 AM

AMD can't just start packaging Swifties or makeing them a requirement for OEM chips. That would be rediculous. A chip has to pass all the standards at the speed it's set for. If it can't pass at 1800MHz maybe it will at 1733 or 1666. That's how things get graded. Only those that actually pass ALL specs at 1800MHz will get marked as 2200+, they can't go revising standards for one CPU. You have to understand that if only 3% of their processors make the grade at 1800MHz, that's fine, because they only need 3% of their CPU's to be 2200+ models.

What's the frequency, Kenneth?
April 24, 2002 3:15:14 AM

Quote:
AMD can't just start packaging Swifties or makeing them a requirement for OEM chips. That would be rediculous.



Actually they can, and do.

If you think the 2100+ axp comes with the same heatsink as the tbird 750 you are sorely mistaken.
Quote:
Only those that actually pass ALL specs at 1800MHz will get marked as 2200+, they can't go revising standards for one CPU.


When amd decides to release a faster grade chip, and that chips thermal output is above the minimum for the current stock heatsink, they DO upgrade the heatsink. In effect revising the standards, unless you wish to assert that the stock hsf which came with the tbird 750 is the same as the one which came with the 2100+.


Quote:
You have to understand that if only 3% of their processors make the grade at 1800MHz, that's fine, because they only need 3% of their CPU's to be 2200+ models.


I am fully aware of how processor rating works. What you need to understand is that the thermal limits for approval they choose are not set in stone, and they have and will again change the stock heatsink to allow for faster cpu releases.




:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 24, 2002 3:17:58 AM

Yep, classic speed grading trick. Intel and AMD have been doing it for years! Chips of many speeds come off the same production line and then they are graded based on testing...

Jesus saves, but Mario scores!!!
a b à CPUs
April 24, 2002 3:18:19 AM

yes, sure, they could release a 65DB cooler, but do you really think that would be good business practice? Or perhaps use a $40 copper cooler, would that be good for profitability? Or perhaps use a cooler so large that it only fits 25% of boards? Or perhaps use a peltier? None of these options are worthy of consideration for a CPU that just might end up on somebody's work desk all day long! Once in a while AMD Meltdown and Fugger make sense when it comes to deriding the extremes some people will go to in order to defend AMD, AMD is smarter than to use such schemes even if some users find them acceptable, because the majority does not, and the majority rules the market.

What's the frequency, Kenneth?
a b à CPUs
April 24, 2002 3:19:49 AM

At least somebody understands!


What's the frequency, Kenneth?
a b à CPUs
April 24, 2002 3:21:33 AM

I can just see it now, thousands of customers saying "I don't buy AMD, they're too noisy". I don't think AMD is that stupid. They will leave it to overclockers to tolerate such auditory abuse.

What's the frequency, Kenneth?
April 24, 2002 3:31:57 AM

A: do you think the coolers with the p4's are cheap?
B: better cooling does NOT always mean louder cooling.
C: do you deny the fact that amd heatsinks have changed over the course of the tbirds evolution?
D: do you think a copper cooler would cost 30 bucks if one was guarenteed to be sold with every amd(mass production=cheaper)

If you do not deny c, your point is incorrect.

Amd can and does change its heatsinks due to increasing thermal maximums of their cpus. This is a fact, cpu grading has nothing to do with this point I am making.


Quote:
Once in a while AMD Meltdown and Fugger make sense when it comes to deriding the extremes some people will go to in order to defend AMD, AMD is smarter than to use such schemes even if some users find them acceptable, because the majority does not, and the majority rules the market.

??? Im not defending amd, I am stating a fact, that comment came off almost as an attack imo crash, and if that is the case you should take a step back.


YOu claimed amd never changes its thermal specs, and all cpus are judged by a set heatsink standard, I PROVED you incorrect about that, the amd line has used several different kinds of heatsinks in its lifespan, thats all there is to the discussion.


If you think my comments on the truth of the matter are a "defence of amd" you should re read them, the comments I am making apply to intel as well, did intel not just move to the zalman cooler? The stock cooler on all cpus CHANGES AS THE CPU GETS FASTER, and thats all I EVER SAID in this discussion.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 24, 2002 3:32:47 AM

Quote:
I can just see it now, thousands of customers saying "I don't buy AMD, they're too noisy". I don't think AMD is that stupid. They will leave it to overclockers to tolerate such auditory abuse.



MY cooler cost 29 bucks, is whisper quiet, and my cpu tops out at 104f@2000++, you do not need a loud cooler to have a good one.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 24, 2002 4:59:06 AM

My Volcano 6cu (copper core) costs $12 retail and cools my cpu 7 degrees Celsius better than stock without additional noise. If AMD needs to upgrade their HSF (and I'm not saying they'll have to) they can do it at very little additional cost. I believe that the TBred and Barton will have lower heat output due to the .13 micron process and lower voltage. This means that the stock HSF will actually function better, even at higher clock speeds.

If the thought I thought I thought had been the thought I thought, I wouldn't have thought so much.
a b à CPUs
April 24, 2002 5:09:44 AM

We were of course speaking of the current core, the Thoroughbred will add legs to it.

What's the frequency, Kenneth?
a b à CPUs
April 24, 2002 5:11:13 AM

So you argue agaist CPU scalling based on quality controll?

What's the frequency, Kenneth?
a b à CPUs
April 24, 2002 5:15:26 AM

Quote:
YOu claimed amd never changes its thermal specs, and all cpus are judged by a set heatsink standard, I PROVED you incorrect about that, the amd line has used several different kinds of heatsinks in its lifespan, thats all there is to the discussion.

Since I have not made such claims, you have proven nothing. I'm simply stating that there is an extreme likelyhood that the processors with the highest speed ratings are the ones that pass the most rigorous testing. Refer to my first post before we get sidetracked any further.

What's the frequency, Kenneth?
April 24, 2002 12:33:24 PM

Quote:
So you argue agaist CPU scalling based on quality controll?



No, not at all, I think we have a misunderstanding friend.


My statement is thus, and thus only.


AMD rates their cpus based on testing as you described, but amd is not adverse to upgrading the hsf when a plateu is reached(and they have in the past).


If amd decides to release a faster chip, and their top bin cpus constantly overheat using the stock hsf spec, they will upgrade the stock hsf to ensure stability and saleability.




:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
!