A64 3000+ steppings

[Rancidelephant]

Hi all, i have 2 A64 3000+ here. which stepping will be the better overclocker.

ADA3000DIK4BI
CBBID 0504DPHW
1193829A50184

or

ADA3000DIK4BI
CBBID 0504DPLW
1194209B5018D

Thanks in advance

[Poki]

I vote for ADA3000DIK4BI CBBID 0504DPHW 1193829A50184.
I can't wait to find out... What do we win if we're right?

[kouch]

I say either one, since they are both CBBID, they are both likely so suck. Sorry if I sould a little pessimistic, but after 5 CBBID's not going a tick over 2.5 at any volts, I have come to the conclusion that they are the suckiest things that every sucked as far as overclocking.

[skematix]

glad i got the 0447 TPFW :) does 2700mhz with 1.5vcore...but won't go farther oh wells im happy enough....maybe its my mobo that can't handle

[Spyrious]

I got a cbbid 0504 dphw with the dfi nf4 ut d but cant go over 215htt for some reason.

[TEDY]

I got a cbbid 0504 dphw with the dfi nf4 ut d but cant go over 215htt for some reason.

try new bios.

[NeoGeo333]

Got a cbbid 0503 and the thing is not stable above 2200mhz no matter what volt.

[$0m#0n#]

Got a cbbid 0503 and the thing is not stable above 2200mhz no matter what volt.
rofl 2.2Ghz? :eek: the worst oc ever saw.

[Rancidelephant]

well im only at 2365mhz :p: not the worst but well on the way :(

[TEDY]

I'm so desperate finding some decent CBBHD AMD64 3000+...

[Evil_Spork]

i suggest you try them first, then you can tell us which is better.

no one can know what a cpu will do without testing. we need less threads asking what a comp will do OCed and more OCing!

[XiN]

I have a bad CBBID Winchester (CBBID 0503 EPMW) :(

Best result 275x9@1.6V, spi 8MB

http://img178.exs.cx/img178/3475/275x9spi8mb8cn.th.jpg (http://img178.exs.cx/my.php?loc=img178&image=275x9spi8mb8cn.jpg)

But it can't finish 16MB spi

[MaxxxRacer]

I am getting kinda scared about this. Im getting a new winnie 3k in a few days and i sure hope it isnt a dog like these are.

But with all of this i still stand by my statement that stepping dont mean carp. I know ppl including myself that got chips that were of a BAD stepping and mine oc'ed to 2.8ghz (from 2ghz) . Its really luck of the draw. It only becomes anything but luck of hte draw when the chip manufacturere decides to experiment with the process to refine it and screw up. Sadly, the amount of samples of amd chips that XS takes in isnt enough of a census to get this kind of information. we are just to small to broad statements about these chips. Now if we took in 1000 chips a week, then we could start making broad generalizations. but we dont.

[Rudzer]

Got a cbbid 0503 and the thing is not stable above 2200mhz no matter what volt.

My A64+ 3000+, only seems to go up to 2.2Ghz also, plus can't get +260Mhz FSB no matter what timmings, volts (DFI lanparty nforce4 ultra & samsung tccd)...nice isn't it?

http://www.rudzer.com/pics/stepping.3000.939.jpg

[TEDY]

rudzer i know how you feel.

[justwOo]

I've got a 3000+ 0447RPAW CBBHD thats priming right now at 305x9 on stock cooling and 1.6v :D

[Rudzer]

rudzer i know how you feel.

The worse is that if I managed to get rid of this cpu, the chances of the new one I end up buying is good are low to say the least :(

[Spyrious]

Ok got it to work at 250htt.My geils cant work with bank interleave enabled.
I have another problem now.With clockgen in windows i can get it to 270 semistable (superpi 8M ,sandra ,...) ,but it cant boot into windows when 270htt is selected in bios?

[rzonealley]

well i guess all CBBID cant reach high clock... mine so far only max out at 2.43(270x9)..on 1.65vcore.

[SIOUX]

I just got this 3200+ CBBID 0503EPMW to play with. Not so good, but better than most ID´s i guess. I cant give it mo0re volt than 1.6 for some reason. i kan use the "10% over VID" option and 1,45v on the CPU vcore (NEO2) But if i put vcore to anything higher than 1.45 it just stay at 1.6v nomatter what. but very strange. i had 4 winnies, and never seen this before. but here are some results with the vcore available

[rzonealley]

I just got this 3200+ CBBID 0503EPMW to play with. Not so good, but better than most ID´s i guess. I cant give it mo0re volt than 1.6 for some reason. i kan use the "10% over VID" option and 1,45v on the CPU vcore (NEO2) But if i put vcore to anything higher than 1.45 it just stay at 1.6v nomatter what. but very strange. i had 4 winnies, and never seen this before. but here are some results with the vcore available


exactly same batch as mine.. but urs is totally better...

[DildePer]

I recieved a CBBID 0505BPAW today.

Returned it, besides that - 10pins were bent.

Never ever CompuMail.dk again. But they seem to have the newest chips.

[ians2k]

I was able to do 2475(275x9) on my 3000+ Winchester(0503 CBBID) at 1.7v.
I have my cpu watercooled(37c idle, 45c load) but not sure if it's safe to run it at that speed 24/7. Going to spend sometime this weekend tweaking it.

[SIOUX]

I was able to puch it to 9*300, 3d stable, by putting Ras-to-cas to 4. With my FX the same g.skill go 308 2,5-3-3-10. also with my 3000+ they can do 308 2,5-3-3-10. but with this 3200+ they need 2,5-4-3-10. so i blame it on the mem controller. My neo2 cant run with divider, so its hard to see where the CPU realy maxes out.but i guess 2700mhz is it, course even with ram at 2,5-4-4-10 i crash over the 2700. soon i will go DFI. i already got the board, just in need of a 850XTPE:)

[Rudzer]

I just got this 3200+ CBBID 0503EPMW to play with. Not so good, but better than most ID´s i guess. I cant give it mo0re volt than 1.6 for some reason. i kan use the "10% over VID" option and 1,45v on the CPU vcore (NEO2) But if i put vcore to anything higher than 1.45 it just stay at 1.6v nomatter what. but very strange. i had 4 winnies, and never seen this before. but here are some results with the vcore available

consider yourself very lucky!

[SIOUX]

consider yourself very lucky!



thats hard when my last winnie (3000+) did 2800 mhz 1.6v :) but anyway this chip is only for a secondary rig, so it doesnt matter so much......

[3FFeX]

rudzer i know how you feel.
..o i feel it :mad:

[scrible88]

My Newcastle 3000+ will do 2920mhz suicide screenshots and is superpi stable @ 2810mhz... on stock air.

[funkdog]

Have a CBIDD0451

right now I'm at 2.65 295x9 with 2.8vdimm 1.6vcore. watercooled

been as high as 300x9 with 2.8vdimm 1.725vcore

[HiJon89]

I guess I'm lucky, my 3200+ that arrived monday is a CBBHD 0450APFW, I've searched a ton of forums and googled it and so far I've only found 1 persons overclocking results with this chip, they got 2.6Ghz 1.5V with XP-90 and 30CFM fan, I'm hoping for more with 1.65V and a TDX, I'll post my OC results somewhere when I get the rig built :toast:

[funkdog]

tried a couple variations on my o/c..

anything above 295x9 is gonna take at least +.325vcore..

295x9 with 1.6vcore or +.225 seems to be where it's at for this chip.

Can get to windows at 310x9 with +.35vcore, but won't PI or Prime.

Better cooling and a volt mod might bring this chip higher, but prime stable (running now) and Pi runs good, at 295x9.

I've suicided at 312x9 into windows with this chip also... 2.8ghz is unreachable stabily with current volt increase on this board.

I'm happy though, temps stay under 50c loaded and for a 150 buck chip I'm able to overclock 50%.... all is good.

UPDATE: Yeah this 295x9 seems like a good sacrifice... surfing/posting/listening to mp3's/downloading/ and priming at 46c.

not bad in my book.

[Eradicator]

Add another one of these new 3000+ to the crappy oc book..

ADA3000DIK4BI
CBBID 0503GPKW

Can only pull off 2400mhz with it, my older 3000+ will do 2.75 on air?? I had never experienced a cpu that was fsb limited before like this one, even if i use a low multi like 5x, its max fsb stability is 270????

[funkdog]

Add another one of these new 3000+ to the crappy oc book..

ADA3000DIK4BI
CBBID 0503GPKW

Can only pull off 2400mhz with it, my older 3000+ will do 2.75 on air?? I had never experienced a cpu that was fsb limited before like this one, even if i use a low multi like 5x, its max fsb stability is 270????


wanna sell your older one?

[Eradicator]

LOL... Nope... Sorry.... It does 3050 under my custom phase 24/7.. My first DFI Ultra-D crapped out on me and just waiting for its replacement to arrive....

[rzonealley]

Add another one of these new 3000+ to the crappy oc book..

ADA3000DIK4BI
CBBID 0503GPKW

Can only pull off 2400mhz with it, my older 3000+ will do 2.75 on air?? I had never experienced a cpu that was fsb limited before like this one, even if i use a low multi like 5x, its max fsb stability is 270????

same story as mine max at 2.4.

[Moonman]

I have three 3000+ winchesters. And here are results.

both CBBHD 0448 SPMW are doing 2.7Ghz prime stable with 1.44 V (my Neo2 overvolts a bit) with XP-90 + 120 mm papst at 1300 rpm. Ambient temp. around 20 degrees C.
2.86 Ghz is max bench stable.

CBBID 0501 maxes out at 2.68 Ghz bench stable. Prime stable at 2.39 Ghz with stock volts and XP-90 + papst.

[TEDY]

omg moonman :))))))

[rzonealley]

CBBID 0501 maxes out at 2.68 Ghz bench stable. Prime stable at 2.39 Ghz with stock volts and XP-90 + papst.

looks like the same things happen to me but mine bench stable at 2.56
but prime stable at 2.43 with xp90 on 1.64vcore.

[SIOUX]

After some work on this CBBID 0503EPMW i got to this......






i will let it prime tonight and see what happens. but i can play games non stop at this speed, so it stable enough for me. just want to see how far it will go with prime stable

[sinsinking]

SIOUX, ur CBBIB 0503 is already NOT BAD at all
even better than some good stepping average quality chip
hit 2.7ghz XD
maybe u can wait for venice if venice overclock better and sell this for the price u can get venice or -$10 or -$20 or stay with this 3200+ is enough

[rzonealley]

yeap his CBBID 0503 definitely better than those i heard using same batch...
his winnie really a gem for CBBID stepping.

[SIOUX]

dont get me wrong guys.....i like this chip alot. but liked I said earlyer, its just for my second "home use" rig :) but i cant stop playing with it hehehe:)

yes...i could not push it further than 2500 at first. i then put tras-to-cas at 4, and it was the bottleneck all the time. i did not believe that to be the problem at first, course they run about 310 with tras set at 3 on my FX. but with this winnie i need to set it to 4. again a mem controller issue.I got suspect when i tryed to put my ram under 200 mhz, and was able to pull 2800 spi1 with the CPU hehe:) but yes...its strange for a CBBID to OC this "well"

[Spyrious]

My Newcastle 3000+ will do 2920mhz suicide screenshots and is superpi stable @ 2810mhz... on stock air.

Why is your spi so high?
My 3000 winnie at 2500 (281htt ,ram 183div ,2.5-3-3-10,cpc on, bank interleave
off) does it 34secs.

[d3birth]

I got a CBBID 0504 stable at 2.4ghz 1.7v, 275htt, ram 173mhz 2.8vDimm :(

I'm gonna get some faster ram this week though. See how it turns out.

[sobol]

I got winnie CBBID 0505 APAW , total crap , max stable 2360 mhz with 1.8v air.
Eaven 1.9 v not helping for 2400 mhz, this new chips are teh bad. Oh i feel prices on ebay of winnie 3000+ from 2005 when venice will be available.

[rzonealley]

guys i noticed one thing on my winnie. the proc didnt really like divider maybe duo to the mem controller. and the proc didnt like really high voltage. i managed to hit 264fsb(max for my ram) 1:1 at only 1.575 vcore. with divider on 270fsb i have to use 1.675vcore. quite weird.

[stinart]

I got winnie CBBID 0505 APAW , total crap , max stable 2360 mhz with 1.8v air.
Eaven 1.9 v not helping for 2400 mhz, this new chips are teh bad. Oh i feel prices on ebay of winnie 3000+ from 2005 when venice will be available.


Man I have exactly your winnie, with Giga-K8ns ultra can't pass 2300 with 1.7 vcore.Someone spoke of bananas?

[demonR6]

Interesting.. my supposed crap castrated Clawhammer 3000+ is able to run 255x10 LDT 3 with mem at 166 SuperPi and Prime95 stable. When I try and run the mem at 1:1 it won't play nice though. At 1:1 I cannot go over 230x10 and I cannot drop it to x9 because it simply won't go near beyond the 230 mark. I am able to pump 1.9v vcore and keep it at 47c under load (SuperPi).. This chip seems to not mind high v. I backed down the vcore to 1.425 @ 123% and still run also was able to play Joint Ops last night without and problems for about four hours. I am ditching this chip for a diff one though, I want to get something with a more successful track record.