Reprise du message précédent :

 
   
le promy arrive surement fin de semaine, p'tetre une k9a    

et encore  
 
http://www.tbreak.com/reviews/article.php?id=149
 
http://www.pureoc.com/AbitKD7RAID_1.asp

franchement je viens de repasser sur ma k3a+ ca roxx mieux je trouve c'est + simple pr monter en frekence

oui mais g une fix socket qui a largué..me les fo ttes pour la fix du prommy..peux pes prendre de risk..pour ca que j'en change  

c une rev koi ta mobo ?
ouais prend une kd7 sinon mais bon plutot nforce2 si tu te fou de setfsb

je ne sais pas si tu as eu le mp ,
jojo61@wanadoo.fr

rev 1.2    

 
ouais l est mieux que la 8k9a?    
m'etonerait  

ton mail ?

je cori skelle a aussi les mises en veille non ? je suis pas sur ?
boarf elle est preske pareil franchement

ok je v prendre l'avis de cpul..me fo 240 de fsb mini
quel est le vcore max et vdddr max sur kd7?

2.32v et 2.3 je crois un truc ds le genre
pr 240 de fsb ca va etre chaud

ouais pour ca que l'epôx est mieux    
je peux en pecho ou des epox par vpc?   a part doc

demande a cpull y connais bien les mobo
ben en angleterre y envoient ls mobo aussi

ouais je v voir mais fo que ca aille vite ..merci tanker

cpul..si tu ns regarde...  
 
donne moi ton avis plz  

hs:passe pas ton mail jojo (mp aussi )

essaye jjojo61@wanadoo.fr

8rda+ , mod vdd, mod mise en veille et hop

je suis la  
kd7 c bien pour oc
c la meilleure pour avoir la freq la + haute avec le proc
malheureusement le fsb peut etre tres haut mais comme ttes les mobos KT400 ya des corruptions a partir de 195-205  
 
dommage pour la 8rda+ au dela de 2.125v la mobo a un prob
2550 stable a 2.1v
2520 pas stable a 2.175v
2520 stable a 2.05v
 
temp ne depassant pas 40 a 2.2v
cherchez l'erreur

ajout bios 3115 pour la 8k9a2

cpul sur la kd7 il y po moyen d'enlever les corruptions avec le TRAS.
 
la je tourne a 205 en cas2.2.7.2 et a chaque demarrge de XP il me dit qu'il a restaurer les fichiers de la BDR, et la restauration a reussi....

tu f comment pour que ca le fasse automatiquement?
parce que moi ct message parlant dy system32 et je dev remettre manuellement sur un autre windows les fichiers du dossier config dns system32

quelqu'un peut m'aider a me dire si c une kd7 ou une kd7-G
http://www.generation-net.com/cata [...] ef=3900698

 
voila ce que ca donne a la plupart des demarrages :
 

 
sinon j'ai le SP1 de windows et j'ai desactiver la restauration du systeme

 
merci
j'crois que je vais garder la k3a+ pour l'instant..seul Pb pour le prommy c que g une fix socket pétée ..oh ca la fera avec 2)

   je vais me prendre une kd7 alors  

elle vaut cb neuve ?

 
moi je prendrai la 8k9a  

prend la 8rda+
sinon les perfs seront pas au rendez vous avec la 8k9a
 
je comprends pas pkoi thesites entre la kd7 et la 8k9a
 
sinon tu las monte le prommy?

parce que j'adore les ht fsb    
 
ta vu la technique pour cut les ponts?

Youhouhhh !
 
La 8RDA+, j'l'ai enfin trouvé
 
je boot, le flash 3129, 11*200, ca démarre j'install XP
 
Mici cpulloverclock pour ce magnifique topic

107? sur le net
a caen elle est à 130? mais je ne sais pas si c le modele kd7 ou le modele kd7-g  , car dans le titre c marqué kd7 et la description ca correspond à la kd7-g    
sinon elle tient bien les 2.35 au vcore ???, parce que si c que sur le papier et qu'en faite elle tient que 2v  

REVIEW BARTON 2500+ PART ONE :
 
http://www.thetechboard.com/reviews/barton.php
 

 
Barton is the next generation core to be used in the Athlon XP.  
 
Aesthetically, the Barton based Athlon XP looks very much like a Thoroughbred XP. The die is slightly longer than the Thoroughbred, but this is because of the extra transistors required to allow AMD to double the size of the L2 cache from 256K to 512K.
 
The increase in performance that this additional cache presents is enough for AMD to justify tweaking it's PR (Pentium Rating) by a few MHz. For example, a 333 MHz front side bus XP2500+ with the Thoroughbred core runs at 2000 MHz. A Barton based 2500+ runs at "only" 1833 MHz.
 
This bump up in the PR numbers is another way for AMD to hit home the idea that MHz isn't everything. The tweaking of the Athlon XP's Pentium Rating has been done once before and not too long ago when AMD introduced the 333 MHz front side bus Athlon XP. A Thoroughbred based Athlon running at 2000 MHz with a 266 MHz front side bus is only considered a 2400+, instead of 2500+.
 
AMD had suggested a few boards to try the Barton CPU on. All of which were based on the VIA KT400 chipset. They suggested the Asus A7V8X, the Epox 8K9A+ and the Gigabyte GA-7VAX.
 
I had tested the chip on all three of these boards, and then some. The Epox and Gigabyte fired right up with the Barton 2500+, but failed to recognize the CPU as an Athlon XP or display the correct XP rating despite the fact that these boards all had the newest official release BIOS?s on them (the BIOS revisions available on their websites as of 1/27/03). The boards simply referred to the CPU as an ?AMD Athlon 1833 MHz. Certainly there are already Beta BIOS?s floating around that fixes this minor problem or AMD would not have suggested either of these two motherboards.
 
The Asus A7V8X failed to post at all. Using a Thoroughbred based Athlon XP 2600+; I could fire the motherboard right up. Obviously there is a newer BIOS out there for this board. If there wasn?t, I?m sure AMD would not recommend using the board with the Barton. Unfortunately, I was unable to get on Asus?s FTP site to download a newer BIOS this weekend, not because of any fault of Asus?s, but because nearly 22,000 administrators didn?t update their SQL based servers when the SQL Slammer hit this weekend.
 
A couple other boards that I had handy that I had tried the Barton 2500+ on were the KT400 based Biostar M7VIT and the nForce2 based 8RDA+. Both of these boards ran flawlessly with the new CPU, but did not recognize the CPU as an Athlon XP upon boot up, just as the Gigabyte and other Epox had.
 

 
I had noticed that the Barton CPU that I was in possession of, even though my CPU was not an engineering sample, had uncut L1 and L3 bridges. This means that the multipliers are not locked and could be changed in the BIOS. On my Epox 8RDA+, I was able to get the PC to run stable all the way up to a 177 MHz front side bus and a 12.5 multiplier using the standard core voltage of 1.65V and an off the shelf Coolermaster Heatpipe heatsink and did not appear to have any problems. This increased the speed of the CPU from 1833 to 2215.
 
 
This WCPUID screenshot shows the CPU running at 177x12.5.
This was the highest speed I could get the PC to run stable with default voltage and nothing more than a Coolermaster Heatpipe.

 
 
I had heard a rumor that the Barton featured CPU throttling similar to that of the Pentium 4. I now know this to be true. At my overclocked speeds, SiSoft Sandra CPU benchmarks on a cold boot would return scores significantly higher than either of my other CPU's (both my 266 MHz FSB 2400+ and my 333 MHz FSB 2600+ Thoroughbreds), but once the system was worked up to the point where the CPU should have run significantly warmer (like running Comanche 4, for example), the benchmark scores started to plummet and reveal numbers significantly LOWER than even the 2400+. Fortunately, my Epox 8RDA+ has the ability to disable CPU throttling and although my overclocking was limited; I was able to obtain improved benchmark scores. Unfortunately, the temperature at which the CPU is throttled on my Epox board can not be set. I can only set the amount of throttling.
 
 
I was able to boot into Windows at 2227.3 with a 178 MHz FSB and a 12.5 multiplier,
but the PC would not run very stable and with throttling enabled in the BIOS, the PC would run slower than even a 2400+.
 

REVIEW BARTON 2500+ PART TWO :
 
http://www.thetechboard.com/reviews/barton-2.php
 
My test machine is an Epox 8RDA+, 512MB of Samsung based PC2700 (just one stick), a Coolermaster Heatpipe heat sink, a GeForce2 Ti 64MB video card, a Maxtor ATA133 40GB drive and Windows ME.
 
The benchmarks below were obtained using an OEM AMD Athlon 2400+ CPU with a 266 MHz FSB and a 256K cache Thoroughbred core running at 2000 MHz, an OEM AMD Athlon 2600+ CPU with a 333 MHz FSB and a 256K cache Thoroughbred core running at 2083 MHz and, of course, an OEM AMD Athlon 2500+ CPU with a 333 MHz FSB and 512K cache Barton core running at 1833 MHz. The benchmarks also feature the Barton 2500+ running at 2215 MHz, which was obtained by changing the front side bus from 166 MHz to 177 MHz and by changing the multiplier from 11 to 12.5.
 
Despite AMD?s adjustment of the Pentium Rating for the Barton, the Barton does not even beat out a 266 MHz FSB Thoroughbred 2400+ in some benchmarks. I found that benchmarks where raw clock speed plays the greatest part of pulling off the big numbers still beat out my Barton 2500+.
 

 
As you can see, SiSoft Sandra?s CPU Arithmetic and Multimedia Benchmarks were better with the aforementioned 2400+ running at 2000 MHz, then the Barton 2500+ running at 1833 MHz. The extra cache did help Barton pull off better scores than a 2400+ in SiSoft?s Memory Bandwidth Benchmark, however. In PCMark, CPU scores were low for the Barton, but Memory scores were high. And in a ?real life? benchmark, like Novalogic?s Comanche 4 Demo, the Barton pulled off both higher FPS and higher Tris per second.
 

 
Once we get to benchmarking the memory bandwidth of the system, we find where the extra L2 cache benefits the PC. Cache helps your PC's performance by storing information that the processor may require often. Since cache runs on the CPU at the speed of the CPU, the retrieval of this information is much quicker than if the CPU had to resort to asking the slower system RAM (like your DIMM's, for example) to give up the information the CPU needs or, if the information is not available in the system RAM, go fetch it from an even slower source (hard drive, CDROM, etc... you get the idea).
 

 
So cache helps memory benchmarks, but is an increase of L2 cache on the CPU enough to justify a 100 point increase in the XP Performance Rating? As you can see below, PC Mark scores for the 2500+ at 1833 MHz are low when benchmarking the CPU, but are definitely better in PCMark's Memory tests.
 

 
 
Where AMD extrapolates an XP Performance Rating is not from it's CPU benchmark or it's memory benchmark, but from the real world results of the two. In essence, you could say that the end result is the average of the two. In the Comanche 4 Benchmarks below, you can see that the 2500+, even running at it's default speed of 1833 MHz, even beats out the 2600+ running at 2083 MHz in FPS and average triangles per second.
 

 
What really impressed me is the speed and stability I was getting out of the overclocked Barton. The Barton is hardly the fastest chip on the market, but if a 25% overclock with default voltage and a $35 fan, I can see these little guys flying off the shelf.
 

 
 
Update 02/01/03: For those who want to are asking why I used Sandra 2002.... I do so because I can. Seriously, though. I plan on getting 2003. I blame the newborn baby for all of my shortcomings.
Footnote: The information about what motherboards AMD recommends was second hand information. I do not have a contact person at AMD.
 

  j'ai tourné avec le vapo a fond hier, eau à 16° avec un 120 panaflo et 11°avec un ebm beaucoup plus puissant !!
  1350 le xp  
oups edit:pas 1350 !!! 2350  

pas mal!!!!!    
t un dieu        
 
jai un ebm + panaflo + papst

1350MHz?  

au bout d'une heure l'alim vapo s'est mise en protect j'ai du mettre le pana pour le refroidir   les 2 ventils en 12v ça faisait un boucan   + ptits blemes de fuites suite a la derniere fois ou j'avais essayé de détruire la boiboite  

ah ok 2350
 
je prefere
voltages a fond en +  

2350 c pas terrible