Fun with a 8KDA3J and an A64 3200+ NewCastle...

and an R507'd Mach1 with a touch Vmods on top of the 8KDA3J. :D

Just put everything together about 30 minutes ago and I'm running my good ol' Corsair XMS 3500v1.1 with 3.2v and my CPU with 1.60v...
http://www.myfavoritehacker.com/some...81-desktop.png
Bumping up 1.65v to see what else this can do. I may have to re-seat the evaporator since I'm questioning the contact... but I'm going to finish playing first.

For those who are curious as to whether or not this board has issues going above 2.5GHz... uh... no. ;)

This bench session is being done at 3x HTT(LDT), the "auto" 11x multiplier, using the BIOS to select the Vcore, and the May 5th, 2004 BIOS(since I was too lazy to flash to the newest one yet). :D

:toast:

Very nice man :)

Omg nice! But give em some more juice and show us 3Ghz :)

I prolly order the 8kda3j on monday :D

Ok then, here's a little more juice for you. ;)

A little SuperPi action...
http://www.myfavoritehacker.com/some...SuperPi-33.png

And a little more PiFast...
http://www.myfavoritehacker.com/some...Fast-44.72.png

That's all for tonight I'm affraid. I'll hopefully get to play a little more tomorrow afternoon.

I'm going to try and run it overnight at 2.75GHz and a "reasonable" Vcore. ;)

Almost forgot the 1:1 BH-5 shot...
http://www.myfavoritehacker.com/some...258MHz-FSB.png

Memory is now at 3.3v and the Vcore tops out at 1.9v when run full-load.

Nice stuff Jeff.. seems the J does perform as well as the +..

yeah its the same layout only without the silicon image SATA controller onboard.

This week i also get my new stuff.
3400+ NC , 8KDA3J and a Mach 1 :)

Didn't get it - the board has issues to get 2.5Ghz or not? :eek:
But I really hope the J OC's and performs equally as the +...

Quote:

---

Originally posted by Mariachi
Didn't get it - the board has issues to get 2.5Ghz or not? :eek:
But I really hope the J OC's and performs equally as the +...

---

it has NO probs above 2.5g ...look at the Mhz Jeff is getting in the screenies :eek: :)

Quote:

---

Originally posted by Mariachi
Didn't get it - the board has issues to get 2.5Ghz or not? :eek:

---

A couple people are/were having problems with the board and appeared to be implying that the board may have a defect that's holding them back to ~2.5GHz... which just isn't true. ;)

You talking about the specific 8KDA3J or both boards?

The two people I'm thinking of were using the + version... I think. :confused: Btw I also have a + version and had that board running 2.6GHz with a slug of a processor. ;)

whooohooo jeff is back! :banana: :eleph: :YIPPIE:

yea

Where the hell'd I go? :lol: You're a fruity one Saaya... :toast: ;)

your back hardware/oc/benchmark wise i mean :D

hadnt heard about your new hardware for some time, the last i could remmeber was that you had some bad expririences with 2 a64s. and something about your wife telling you its ok if you spend even more money and buy another one :D

hows the flight show? :D

btw, try more vcore. even if you reach bad temps it might give you a boost, give it a try. up to what speeds is your bh5 good?

Is there a new bios that allows up to 1.85v ?

At least put 1.7v through that thing :D

yeah! what are your results like?

i can't get mine stable at 2.7 ghz at the moment...

i tried but no good.... but it is benchable....

hmm...

Quote:

---

Originally posted by LilGator
Is there a new bios that allows up to 1.85v ?

At least put 1.7v through that thing :D

---

Nope... no BIOS for 1.85v. You can tweak a little with ClockGen and get 1.75v... sometimes 1.80v though.

And the 2.83GHz PiFast is with 1.9v. :) I've tried up to 2.0v and it seems 2.86GHz is the most I can get out of this chip right now. I'm going to do a few other things to try and get 2.90GHz stable. If I can do that... I'll be extremely happy. :D

BTW, for anyone running Vcore up over 1.8v... you better cool your Vregs. Without any air blowing over them they reach 80c in a hurry when loading the CPU. :eek:

And the northbridge with the stock heatsink will get to about 55c when running 3.3v and ~260MHz "FSB". If you move just a little air over it the temperatures go down to ~50c.

I'm trying to run 2.81GHz full-load overnight to see if I can get her 100% stable there. That's at "only" 1.9v but so far she's doing alright. I had to drop my memory down to ~240MHz since anything higher will error out anywhere from 30mins to 2hrs into a burn.

I've just started with this board... there are a few heatsinks I want to place and probably another fan even. I may try to do something "water" with the NB too... I have an old chipset block that should be usable. I just need to get some time to do all this stuff... :rolleyes:

I've got the + version. I used the supplied heatsinks on the voltage thingy chips and replaced the NB heatsink with an old Pentium 5x5 sink with a nice fan. Got my motherboard voltmodded too. I reached 2750Mhz with my Clawhammer with 1.85V though a little bit unstable,I had the LDT on 4x. I think you should try to get that stable too. For now its the psu thats holding me back :( , the voltages fluctuate way 2 much about 0.1V on both Cpu and Mem. But hey what else do you expect with a no-name 300W :D . System won't even boot with a Vcore of 1.9V :mad: . When I get a new psu I hope to break 2.8Ghz.

btw:
I've got an AMD 64 3200+ 1Mb Cache XPMW

Nice! :up:

Jeff what do you think of the 3200+ NewCastle? I'm thinking of buying the same pair myself. There is a OEM 3200+ Clawhammer available but I think its C0 stepping. And I want the most chance of more mhz tbh. Also the clawhammer is from a diff retailer.

My memory is OCZ PC3500 2x256mb Bh5 ram, think this'll work ok?

Any pointers on how to setup the board ok?

Also would the above combo smoke a 3.0C + Ai7?

hey jeff!

good to see your results...

i'm having some problems with mine :(

i can't get mine stable above 2.6 or so even with mach 1 (unmodded) cooling...

2.65 gave me mad windows corruption last night when i woke up this morning...

I've tried dropping the HTT down to 2x even but that didn't seem to help with anything....

i did the vcore mod and set the vcore to 1.7 bios and then tweaked it up to 1.85...

even at 1.85, 2.7 isn't stable... it errors out in about 18 minutes...

any secrets or tips to get more out ofmy chip? i'd really like to keep this chip instead of trying for a new one...

also....

1 stick or two sticks of ram? i'm using 2 sticks of bh-5

i'm also using bios 5/14...

also, what do you think would be the safest max vcore to use for 24/7 use... 1.9? 2.0?

any suggestions?

@The Runner: If you're looking for MHz then I'd go with the Newcastle. I'm having a lot of fun with it right now and I wouldn't hessitate in buying another one(over the Clawhammer). If you can get a great stepping Clawhammer... I'd say go for it, but I struck out with NewEgg two times in a row. That's why I decided to try a Newcastle and I'm glad I did.

BH-5 is working great for my setup. I'd recommend the Vdimm mod to feed it with 3.1v to 3.3v though. I'm at 259MHz@3.3v with 2-2-2-6 timings. (just as a side note... running a stick AData 282MHz@3.1v with 3-4-4-8 timings performed just a hair less than my BH-5)

@shimmishim: What core is your chip? Also, what is your memory at? I had a Clawhammer that was stuck to 2.55GHz with 1.7v(didn't have the Vmod done then). This NewCastle ran semi-stable at 2.7GHz@1.7v. I really think the lack of onboard cache helps.

Also, maximum safe voltage? I have no idea. ;) I'm running 1.90v-1.92v 24/7 full-load right now. It doesn't "feel" lilke this is too much but only time will tell.

Oh yeah... I'm only running one stick of 256Mb. I don't do much else beside run a few benchmarks and do distributed computing projects so any more than that isn't need(for me at least).

[edit]
Forgot to add that I'm now using the July 9th BIOS and it seems at least as stable as the May 5th BIOS.
[/edit]

i have a newcastle CG revision 3200+

week 13 of '04

:)

yeah... i'm maxed out at 2.6 stable ( 9 x 289)... with only 1.6 in bios or 1.75 under 100% load with the vcore mod.

this is the same as my C0 3200+ clawhammer i had way back and also my C0 3000+ newcastle... but these chips were on a shuttle an50r...

so... hmmm...

maybe i'm maxed out...

maybe i should try out another one...

edit:

no matter how many volts i put into it (as much as 1.9) it won't do 2.7 stable (18 minutes max)....

@shimmishim: Try running that speed with your memory at lower speeds. I can NOT run BH-5 259MHz 1:1 2-2-2-6 24/7 full-load with the 1T enabled. If I set it to 2T it runs a bit longer but it still will crash out(eventually). I haven't looked into a way around this yet since I just don't have the time this week. But this weekend or next week I want to try and get above 260MHz stable enough to run DC projects overnight. I have a feeling it may be the NB overheating or something along those lines.

I did discover that my AData can run 260MHz 1:1 3-4-4-8 24/7 full-load though. I can even set the memory to 3.0v and the chipset to 1.65v and it runs fine. If I try to run 280MHz 24/7 full-load... it acts alot like my BH-5. That's why I think it may just be something that needs cooling and/or a litte more voltage. :shrug: ;)

hey, can you list your settings?

i've tried with the 11x mult and the appropriate HTT and i still couldn't get above 2.6 stable...

it's benchable at 2.7 but i'd really like it to be stable at 2.8 :)

i really think i'm going to have to try another one out! oh well...

hey Jeff, have you vmodded the mobo yet? cause this 1.7vcore max is really holding me back with my 3000+/8kda3+/vapo ls.... i can get it to boot into windows at 280x10 but not prime stable

need either those new bios allowing upto 1.85v or a vmod..

I heard they were planning to release a new BIOS to allow vcore adjustment to 1.85v

But the 2-phase power still worries me.