New WR pifast A64 NewCastle

[YoupY]

After a day of tweaking happyness we GoT it

40.59



Making the Ministry of Overclocking Crew the fastest A64 Newcastle as far as we know

The Crew

The Doctor
Fewture
Epsilon
YoupY

And with as guest

Kakaroto

All this started this morning @ 11.00 when the crew gathered. We finished the Dry Ice container at and started to setup the system. In the afternoon the system was ready for benching.

Mobo : Epox 8kda3+
Mem : OCZ EB PC3700
proc : A64 NC 3400+

Too bad we couldn't get any decent results with this setup. After a few tries the Windows install got corrupt and we had to reinstall, which failed upto three times. We got really sick of it and changed the mobo for a DFI NF3-250 Ultra.

In the late afternoon we made progress having 2 systems benching. The system mentioned above and the system of Fewture with

AXP-M 2600+
Abit NF7-s
Twinmos BH5

To our regrets the system of Few didn't live up to our expectations, also troubles of the Windows install. So we decided to give it another go with a DFI infinity tomorrow. So we hope to post more results of it tomorrow...

After a few succesfull runs with the A64 setup we finally reached 40.59. Giving us the fastest Athlon 64 Newcastle ranking with only 0.4 seconds to go for the fastest non FX proc

The pics (bad quality )













There have been made more pics of this day which will hopefully be posted soon

[bh2k]

wow, some very impressive work there

[Syzzer]

Hooooly sh*t, way to go!

Nice score!

Gotta have more records for us Dutch people!

[Kanavit]

is that dry ice setup? I'd like to see some super pi 1mb runs if possible. maybe some 3d runs.

anyways, what a rad looking lan party you guys got going over there. newcastles cg overclock very well.

[sapphire]

Sweet work dude's
Finally a dutch record again

[PMM]

0.04 off CodeRed with his 3400+

you need to push further...

[YoupY]

0.04 off CodeRed with his 3400+

you need to push further...

If I'm not mistaken CodeRed uses a 3400+ with 1 MB cache making it a claw. Only he and Kunaak stand in the way of the fastest non FX proc.

[CodeRed]

If I'm not mistaken CodeRed uses a 3400+ with 1 MB cache making it a claw. Only he and Kunaak stand in the way of the fastest non FX proc.

Yeah my chip was a 1Mb CH

Superb results there guys


If your going to bench again, PM me and I'll send you some "tools" to get that 0.04 sec

EDIT: actually need a bit more for S754 WR ... cal930 has done a 40.47 sec. See here
EDIT: never mind just saw kunaak's score

[The Doctor]

Were gonna bench again tommorrow. Think we should be able to get a score lower then 40.19s.

[WeStSiDePLaYa]

nice work,


but uhhh, is that a container of vaseline i see? in a room full of guys?

[eva2000]

Yeah my chip was a 1Mb CH

Superb results there guys


If your going to bench again, PM me and I'll send you some "tools" to get that 0.04 sec

EDIT: actually need a bit more for S754 WR ... cal930 has done a 40.47 sec. See here
EDIT: never mind just saw kunaak's score

what tools may they be ?

[CodeRed]

what tools may they be ?

Oh nothing too special, just a small gain from this one ... around 0.05 sec, but sometimes every last ounce is needed.

[Zeus]

What a bunch of geeks.
Well done guys!

[macci]

Very nice work there!
Its done at 283x11 1:1 3-2-2 w/ EB RAM? Pretty damn fast

almost a sec. faster than my old 3000+ @3127MHz

In case your pushing it again and happen to reach the hexus pifast goal too easy then here's another challenge if your interested

[CodeRed]

Very nice work there!
Its done at 283x11 1:1 3-2-2 w/ EB RAM? Pretty damn fast

almost a sec. faster than my old 3000+ @3127MHz

In case your pushing it again and happen to reach the hexus pifast goal too easy then here's another challenge if your interested

Yeah, I think if they hit the pifast goal then a lean XP install should see them hit 05:10'ish in superpi 8M. Is that the Newcastle WR?

[Mrki]

cool guys

[YoupY]

Very nice work there!
Its done at 283x11 1:1 3-2-2 w/ EB RAM? Pretty damn fast

almost a sec. faster than my old 3000+ @3127MHz

In case your pushing it again and happen to reach the hexus pifast goal too easy then here's another challenge if your interested

This EB's do even faster. We have seen them @ 292 @ 3.0-2-2-5-10-16-1T, but with 10x multi then won't do much more then 295, which gives us too little cpu power.

SuperPi 8M is a nice challenge too We'll see if we can do that too

[Mrki]

This EB's do even faster. We have seen them @ 292 @ 3.0-2-2-5-10-16-1T,

What? 292 @ 3-2-2-5??? VDimm?

[Crow]

excellent score

[macci]

What? 292 @ 3-2-2-5??? VDimm?

I'd guess there running in the 3.2V range.

mine ran 3-2-2 at 294.8 w/ 3.3Vish

[The Doctor]

I'd guess there running in the 3.2V range.

mine ran 3-2-2 at 294.8 w/ 3.3Vish

Thats correct it did 292MHz @ 3.2v

Some more pics:

Fewture doing..... something:


The container:


Me preparing the isolation for the backside:


Container on the epox board (board was not used for the pifast run):


A look inside the container:


what we saw after removing the container (too soon ):


Setup overview

[kakaroto]

Looks like fewture is lapping his zalman chipsetsink :banana4:

[bh2k]

Smoking is bad for you :-p

[Epsilon]

Smoking is bad for you :-p

He was there and he is right fewture was lapping a zalman heatsink :P

[YoupY]

Very nice work there!
Its done at 283x11 1:1 3-2-2 w/ EB RAM? Pretty damn fast

almost a sec. faster than my old 3000+ @3127MHz

In case your pushing it again and happen to reach the hexus pifast goal too easy then here's another challenge if your interested

The challenge is intresting, but we had some troubles with the DFI today. For some reason it wouldn't boot. After that we tried with an Epox 8kda3+ but we failed getting a windows install. Setup crashed several times and we decided to take an old GigaByte K8N Pro out of the closet where it was collecting dust.

We managed to get some decent results but not the kind we hoped.

Temps :



Max cpu speed :



SuperPi



Pifast



For some stupid reason I forgot to change the Comand Rate to 1T so all these benches are with 2T I'm not sure if the nf3-150 chipset likes 1T but I can rip my head off for not trying at least

We have made a go for the 8M SuperPi but @ 3100 Mhz ond the cpu and 283 HTT sync on the mem we should come out near 5.40 if it didn't crash during iteration 17.

We hope to get more lucky next time