A64 OC Data

Status
Not open for further replies.
3700 ran at 2700+ w/ epox 8kda3+ 225x12 vc2.7/vd2.8/chipset2.75.
watercooled w/ swiftech6000 and black ice xtreme2radiator+ 2x delta fans1212ehe.
cpu load at 37C.
2x513 corsair3500 1:1.
2x1200jb wd w/ 3ware7210 raid0.
With 2 psus Antec550 f/mobo,pump ,2 hds,6800U at 475core/1250mem. and antec480trublu f/ 2 radiator fans,4 case fans, cdrw drive.
 
^^^
try using the return key on occasion so your oc specs are at least readable ;)

and i need to try this again now that i'm in a really cold room, but here are some screenshots that i took when i was messing around with my 3200+ a while ago.. dunno why i thought to post them now instead of before though :rolleyes: just one thing to note, i can't really vouch for the stability of these speeds at said voltage cause.. well i didn't have enough time to thoughly test them, was just going for the top end of the chip (which appears to be 2.6ghz without any mods)
cpu-z.PNG
 
out of the 2800+

I set the memory down to 133, voltage up to 1.55 (1.6 actual) HT to 3x and 9x270!!

Woot - 2.43 GHZ out of an original 1.8GHZ. All this on air. (Zalman 7000 Cau)

If only the Chaintech VNF3-250 gave me correct temp readings, I would leave it there -

Right now it says 72C under full load 58C under idle but then it says 55C under idle at stock ;( -

either way it chewed through Super Pi, Prime95 all the tests..if only I could get the lil quirks worked out with the VNF3-250 I would be a happy pappy.
 
KV8 Pro v1.1
Clawhammer 3400+ @ 2350 MHz (235*10)
Vcore 1.55 volts
Memory 196 MHz (3 4 4 8) 2.8 volts
HTT * 3
Stock cooler/fan
Antec Lanboy case with 350 watt TrueBlue PS

I can't seem to get that extra 50 MHz out of this core.
 
groebuck said:
out of the 2800+

I set the memory down to 133, voltage up to 1.55 (1.6 actual) HT to 3x and 9x270!!

Woot - 2.43 GHZ out of an original 1.8GHZ. All this on air. (Zalman 7000 Cau)

If only the Chaintech VNF3-250 gave me correct temp readings, I would leave it there -

Right now it says 72C under full load 58C under idle but then it says 55C under idle at stock ;( -

either way it chewed through Super Pi, Prime95 all the tests..if only I could get the lil quirks worked out with the VNF3-250 I would be a happy pappy.
Click to expand...

Nice overclock groebuck. =)
 
groebuck said:
how ya been - you gave up the asus!! ;) - dang an fx too!
Click to expand...

Pretty good man, how about yourself? Yeah, I finally let the Asus P4S533 based rig go, lol. I gave it to my Sister. It upgraded her from a her old Celeron 333 to the P4, I overclocked it to 2.1 GHz with the same 1.6a Northwood from back in the day. So it's still going!

Yah, I finally went AMD. I built the Box back in June when the 939 stuff launched. It's been a very solid platform.

Take it easy groebuck, good to see you around. =)
 
where is everyone getting the beta bios for the vnf3-250? i googled it and serched the H and coulldnt find anything.
 
blvdKing said:
KV8 Pro v1.1
Clawhammer 3400+ @ 2350 MHz (235*10)
Vcore 1.55 volts
Memory 196 MHz (3 4 4 8) 2.8 volts
HTT * 3
Stock cooler/fan
Antec Lanboy case with 350 watt TrueBlue PS

I can't seem to get that extra 50 MHz out of this core.
Click to expand...

wow, ummm.. why do you have the timings so slow? and i think if you put the voltage up to 1.6v, it'll do 2.4 stable. i'll use my cpu as an example. i can get it to 2.35ghz with stock volage, then to 2.45ghz with 1.55v, then 2.5ghz with 1.6v, but it takes 1.8v to keep on the hairy edge of stabily at 2.6ghz. there gets to a point where voltage has diminshing returns, but since you're still at 1.55v, you can probably get at least another 200mhz out of it if you want to raise the voltage a lot.
 
Just saw this thread.....where have I been? :eek:

Brand new DFI Lanparty UT NF3-250GB. This is an overclockers dream. I'm presently running 315 FSB, 3XHTT, PC3500 corsair memeory stretched at 460 cas 2 at 3.00 V-Memm.
site1027.jpg


I went for the 3200+, but I would have been as happy with a 3000+ and saved a couple of dimes. :D
 
Good questions...

The memory I am using is two sticks of generic PC3200. I need loose timings to get the memory stable.

As for the CPU, I can't get 2.4 GHz out of the core even at 1.7 volts. At 1.7 volts, the CPU runs too hot (more than 70 C during Prime95).

The HTT is at 3x to take it out of the equation.

It doesn't make much sense that 50 MHz is out of reach, but I'm not sure what the limiting factor is. The memory is stable at 200 MHz at 3-4-4-8, there are no power dips from the PS, and HTT is at 3x on a board that supports 1 GHz HT.

Does increasing the voltage to the chipset help anything?



(cf)Eclipse said:
wow, ummm.. why do you have the timings so slow? and i think if you put the voltage up to 1.6v, it'll do 2.4 stable. i'll use my cpu as an example. i can get it to 2.35ghz with stock volage, then to 2.45ghz with 1.55v, then 2.5ghz with 1.6v, but it takes 1.8v to keep on the hairy edge of stabily at 2.6ghz. there gets to a point where voltage has diminshing returns, but since you're still at 1.55v, you can probably get at least another 200mhz out of it if you want to raise the voltage a lot.
Click to expand...
 
Setting up a system for my friend.

So far:

300fsb.jpg


2.4Ghz - 8 multi - 300mhz fsb - 1.65 vcore - 43c idle - Zalman 7000cu.

PQI Turbo 2 x 512 @ 2-2-2-5-1T @ 2.7 volts @ 200mhz

I'm running Prime on it right now, I'm going to try some 1:1 setups tomorrow, hopefully I can get 267 1:1 but I'd imagine the memory controller won't like much over 250.
 
Hey all, here is my system:

A64 3200+ Clawhammer C0
Stock HSF
Chaintech VNF3-250
OCZ Platinum PC-3500

I have been trying to OC my cpu, but the best i can do is 245x10 stable ... if i go any higher my system just doesn't boot. I had my vcore at 1.65~1.70, my vdimm at 2.9 divided, and my HTT to 3x, and i just cant get past 2.45ghz ... can anyone please help me.

Stock my CPU temp is 35C idle and 45C load
OC'd to 2.4 w/ 1.65~1.7v = 39C/50C
 
I have the CG 3000+ on a VNF3-250 Chaintech with the 7/29 BIOS. The chip does 2.4ghz (10x240) at stock voltage but needs a bump to 1.6 in the BIOS to maintain stability at 2.5ghz (250x10). Ram divider is set to 166 since I have PC3500 and that gives me 416 DDR speeds. HTT is at 4x for a speed of 1000mhz. Nice. The 7/29 BIOS, while better, does seem to report the CG stepping temps WAY too high. I am water cooled and showing an unadjusted load temp of 61c, idles at 44c. I am betting that I really should be at a true 50c or so under load based on what this exact same cooling rig did on an XP chip running at 1.8v and 2.4ghz and the fact that my digital temp probe reports 39c surface temps on the waterblock, radiator, pump, and hoses under extended load.

VNF3-250 7/29 BIOS
A64 CG 3000+ (259x10) 2.5ghz
Kingmax PC3500 166 divider = 416 DDR
HTT 4x (1000fsb)


I am happy with my 25% overclock.....very happy.
 
xgodxicex said:
Hey all, here is my system:

A64 3200+ Clawhammer C0
Stock HSF
Chaintech VNF3-250
OCZ Platinum PC-3500

I have been trying to OC my cpu, but the best i can do is 245x10 stable ... if i go any higher my system just doesn't boot. I had my vcore at 1.65~1.70, my vdimm at 2.9 divided, and my HTT to 3x, and i just cant get past 2.45ghz ... can anyone please help me.
Click to expand...


The CO chips, most are not going as high as the CG steppings, so you may just be at it's limit. Not a bad speed at all for the CO chip....
 
TheRapture said:
The CO chips, most are not going as high as the CG steppings, so you may just be at it's limit. Not a bad speed at all for the CO chip....
Click to expand...

there must be a way to push it further, this sux
 
xgodxicex said:
there must be a way to push it further, this sux
Click to expand...


The 3000+'s are hitting the same speeds overclocked as any other other higher binned S754's...it just means AMD's process is very refined and they are speed binning the chips. That's why I chose to get the cheap 3000 model since it appeared that I was going to end up at the same speed, in fact, the lower multiplier can be a benefit...


You chip does not suck, it's doing more than it was rated for, just because it won't do more does not mean it sucks. Get a CG stepping and you will most likely be able to go higher.
 
TheRapture said:
The 3000+'s are hitting the same speeds overclocked as any other other higher binned S754's...it just means AMD's process is very refined and they are speed binning the chips. That's why I chose to get the cheap 3000 model since it appeared that I was going to end up at the same speed, in fact, the lower multiplier can be a benefit...


You chip does not suck, it's doing more than it was rated for, just because it won't do more does not mean it sucks. Get a CG stepping and you will most likely be able to go higher.
Click to expand...

what if i were to drop the multiplier and jack up my HTT ?
 
xgodxicex said:
what if i were to drop the multiplier and jack up my HTT ?
Click to expand...
chip limits are chip limits, now matter how you set the final speed, that speed won't change without better cooling and more voltage unless it's a board/ram limit in disguise...
 
when do the efforts and cost of o/c become counter productive? i.e. spending $300 on cooling and $200 on >DDR400 ram vs. buying faster model processor/mobo combo.
 
The different is the RAM ratio (FSB:DRAM)
As both o/c almost on par in CPU

DFI Lanparty NF3
265.jpg


Abit KV8Pro
2700.jpg
 
when did you buy em? i have two sticks of hyperx 3200, but i find it hard to get em above 220, even with slacked timings.. i'm blaming it on my mobo for now cause i got em to 235 before on my a7n8x

you've also got what.. 3.6v going through those ram sticks?
 
(cf)Eclipse said:
when did you buy em? i have two sticks of hyperx 3200, but i find it hard to get em above 220, even with slacked timings.. i'm blaming it on my mobo for now cause i got em to 235 before on my a7n8x

you've also got what.. 3.6v going through those ram sticks?
Click to expand...

3.6v
PC3500 HyperX
 
strange.. i just tested with 3dmark05 out of curiosity, i got 5237 on the cpu score, with my ram running at 200mhz and the cpu at 2.4ghz

3d05.PNG
 
Please correct me if I am wrong - doesn't AMD state that going above 2.9 volts on the vdimm increase the risk of burning out the memory controller?

I'm not condemning 3.6 vdimm, I'm just wondering.
 
blvdKing said:
Please correct me if I am wrong - doesn't AMD state that going above 2.9 volts on the vdimm increase the risk of burning out the memory controller?

I'm not condemning 3.6 vdimm, I'm just wondering.
Click to expand...
just like how running the cpu above 1.65v increases the risk of burning out the die itself? :D
 
now thats just not right..... did you ever think how it would make us feel? :(
( oh we'd still like some more benchies though :p )
 
Grahxen said:
Heres mine:

Its a pretty good overclock im proud of :)
Click to expand...

It is... It is with 1mb cache right? Still trying to break 2.6ghz with my clawhammer 3.4+ with prometeia... somehow, I think it is time I look for other CPU.

[EDIT]

Finally broke down and got 3.2 newcastle. Running like charm and prime stable etc.
2750mhz @ 1.75v, R404 Mach 1 chilled. Was hoping 2.8ghz but not gonna look again just for 50mhz now, time to save up to see now 90nm unfolds for next month or so now :)
Memory is @ 250 3-2-2-10 1T but I need to play with voltage again to figure out how I got 2.5-2-2-10 before :p (think it got to 255 with 2.5-2-2-10 but don't remember what volt it was).
 
Status
Not open for further replies.
Back
Top