Overclockers Australia has a preview on the new .13 micron Athlon XP with benches. Link courtesy of /.
http://www.overclockers.com.au/techstuff/a_amd_tbred/
http://www.overclockers.com.au/techstuff/a_amd_tbred/
Thoroughbred Preview with Benchmarks
- Thread starter astroview
- Start date
<< finally a decent looking preview. >>
not really. i am very curious as to how well it will overclock. but not with a vapochill system! i want to know what we can get out of these with air cooling.
The VCore of the 1900+ is 1.6V. That's not too bad, hopefully it'll stay that low through the 2200+ and up.
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
<< finally a decent looking preview. >>
not really. i am very curious as to how well it will overclock. but not with a vapochill system! i want to know what we can get out of these with air cooling.
Well he did say he saw a 4 degree Celsius drop in temp even with the vapo chill, wich is a huge drop, something like 10 Fahrenheit?
not really. i am very curious as to how well it will overclock. but not with a vapochill system! i want to know what we can get out of these with air cooling.
Well he did say he saw a 4 degree Celsius drop in temp even with the vapo chill, wich is a huge drop, something like 10 Fahrenheit?
I was more impressed with the LCD printer cable mod on that site 
The review was okay though, I still hate PR ratings, they cause confusion to this day
The review was okay though, I still hate PR ratings, they cause confusion to this day
I know this is an early yield and overclocking results shouldn't be as high as they will eventually become but the guy with a VapChill was able to get it to 2.0 GHz stabily. Is it me or should it have been more than that?
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
I know this is an early yield and overclocking results shouldn't be as high as they will eventually become but the guy with a VapChill was able to get it to 2.0 GHz stabily. Is it me or should it have been more than that?
No, all cores have their limits, and since this is a production sample, the overclocking results should not reflect in the final release unless 2ghz is the limit for the Athy, which I highly doubt.
No, all cores have their limits, and since this is a production sample, the overclocking results should not reflect in the final release unless 2ghz is the limit for the Athy, which I highly doubt.
I sure hope that AMD does not release those ThoroughBred chips at the same FSB as the current Athlon XPs. If I were AMD, I would counter Intel's 533 speed announcement with my own of T-bred chips that run at 166 FSB. The new chip should have the headroom to accomodate the higher FSB (besides, we will all o'clock them to that level anyway). So, that would explain the identical inside chip architecture, but the new name of Athlon XP+. The + refers to the higher FSB. All it takes is a BIOS change by the motherboard manufacturers and perhaps some memory ratio changes...and bingo, AMD remains competitive with Intel until the Claw is ready.
Put another way, increasing the default chip FSB speed (like Intel did with its Pentium III E and EB units and is now doing with its P4 Northwood a and b units) is the only way that AMD can possibly ramp up to the 2800+ speeds that Intel has in its pocket for the rest of this year. The fact that memory bandwidth stays the same (probably) is not as important as the chips' speed race.
Will the above prediction come true or only be a figment of my overly fertile imagination?
Put another way, increasing the default chip FSB speed (like Intel did with its Pentium III E and EB units and is now doing with its P4 Northwood a and b units) is the only way that AMD can possibly ramp up to the 2800+ speeds that Intel has in its pocket for the rest of this year. The fact that memory bandwidth stays the same (probably) is not as important as the chips' speed race.
Will the above prediction come true or only be a figment of my overly fertile imagination?
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
Deskstar-
Your not thinking straight, although a 166mhz fsb part would explain the uterly pointless release of the KT333 shipset. Who knows, maybe AMD will ramp up to 166mhz fsb. But a higher FSB means a great deal more heat so i highly doubt that will happen.
Your not thinking straight, although a 166mhz fsb part would explain the uterly pointless release of the KT333 shipset. Who knows, maybe AMD will ramp up to 166mhz fsb. But a higher FSB means a great deal more heat so i highly doubt that will happen.
I was hoping that the additional heat generated by the 166 bus increase would be offset by the smaller chip size of the T-bred. Otherwise, the T=bred seems of less than earthshattering specs.
ToBeMe
How did you get a 2.4b chip to run a rig on 4-25-02 ???
I didn't think that they were announced until today.
How did you get a 2.4b chip to run a rig on 4-25-02 ???
I didn't think that they were announced until today.
<< But a higher FSB means a great deal more heat
It does?
>>That's news to me also. I thought it was a small difference.
Don't forget that higher bus speeds also mean more potential for crosstalk and other corruption. It's not always easy just to crank up the speeds. One easy example is ATA33 vs ATA66/100/133. An 80 wire cable was needed to keep the signal clean. Anoth example is the SIS645 chipset. If you want to run PC2700 memory, you can only populate 2 of the 3 DIMM slots. Fill the 3rd slot and the memory clocks itself down to PC2100 levels.
Windogg
Windogg
formulavav
The supposition was the that multiplier stayed the same (11.5 x) but the that FSB was increased from 133 (XP) to 166 (XP+). The resulting increase in speed means more work by the processor in a given period of time; more work (cycles, etc) means more heat, unless some physical change occurs in the processor. In the case of the current XP, the processor is a 0.18 micron die versus the newer unreleased XP+, that is a 0.13 micron die. The smaller die should allow the distance, time, work and temperature of the chip to be reduced compared to a similarly clocked XP. Assuming the heat can be effectively removed, then the XP+ might be able to run at a higher FSB (166) than the current XP (133) without any increase in temperature. Now, I am certain that I have violated every rule of Electrical Engineering in the above explanation. But, in general, that is the story.
But, I have received reasonably reliable information that my hope that the XP+ will be released with 166 FSB is not based in fact and is nothing other than my own hope. So, although it would be nice, I must throw cold water onto the thought, and drift back into reality of the chip world where nothing is known for certain until 3 months after a product is released.
The supposition was the that multiplier stayed the same (11.5 x) but the that FSB was increased from 133 (XP) to 166 (XP+). The resulting increase in speed means more work by the processor in a given period of time; more work (cycles, etc) means more heat, unless some physical change occurs in the processor. In the case of the current XP, the processor is a 0.18 micron die versus the newer unreleased XP+, that is a 0.13 micron die. The smaller die should allow the distance, time, work and temperature of the chip to be reduced compared to a similarly clocked XP. Assuming the heat can be effectively removed, then the XP+ might be able to run at a higher FSB (166) than the current XP (133) without any increase in temperature. Now, I am certain that I have violated every rule of Electrical Engineering in the above explanation. But, in general, that is the story.
But, I have received reasonably reliable information that my hope that the XP+ will be released with 166 FSB is not based in fact and is nothing other than my own hope. So, although it would be nice, I must throw cold water onto the thought, and drift back into reality of the chip world where nothing is known for certain until 3 months after a product is released.
mastertech01
Moderator Emeritus Elite Member
- Nov 13, 1999
- 11,875
- 282
- 126
Ill be happy for just cooler at default... but 4C isnt much when you are running 50 average and up to 60 with 7K cooling fans with no OCing.
well, despite this site's claim that it "ran cooler"(despite no such claim when he pumped hte voltage to 2.0), and without taking into account the much smaller contact surface, even with a very agressive cooling setup he was only able to gain 50mhz over his highest Palomino overclock. To me, with a vapo-chill setup, that isn't impressive at all for a .13 versus .18 process, even if it is early production.
Mike
Mike
this isn't supposed to be overly impressive in the first place, and it isn't...
i can't see them coming out w/ a 166mhz FSB for, what i'm guessing to be, a few more t-bred releases....new bios, etc. just for a few t-bred models?...
doubt.
i can't see them coming out w/ a 166mhz FSB for, what i'm guessing to be, a few more t-bred releases....new bios, etc. just for a few t-bred models?...
doubt.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 19K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter