Kinc Going for Round II with GTX 512MBs

Kinc, FX-60 sure is tricky. 3,4 GHz and 3,45 GHz are pretty easy to bench but 3,5 GHz locks up too damn often.

I'm currently working with new setup. Here is a score with graphics cards stock cooled and FX-60 @ 3,45 GHz.

http://www.io-media.fi/temp/orb.png

:)

Thats looking very juicy Sampsa :cool:

Thanks for the teaser.

I wonder if Pedro has any luck with his new tweak for dual core. That would surely help if we could find a way to drive these new cold bugged dual cores harder at idle like with Ricky Tweak/single core. Its all a big challenge because any good bencher knows your best scores come off a fresh boot. If you have to play around in Windows running background stuff to keep the CPU going you loose alot of points.

Quote:

---

Originally Posted by Sampsa

Kinc, FX-60 sure is tricky. 3,4 GHz and 3,45 GHz are pretty easy to bench but 3,5 GHz locks up too damn often.

I'm currently working with new setup. Here is a score with graphics cards stock cooled and FX-60 @ 3,45 GHz.

http://www.io-media.fi/temp/orb.png

:)

---

Tricky little thingy :)

What HTT, voltage and base temp (ln2 or dice?) are you runing for 3450MHz?

3514MHz is stable enough at the right base temp. It is just that three pots on ln2 when running max oc on the cards are pretty tuff to control by your self. The FX-60 temp difference in idle and load is nuts, never seen anything like it. But I'm getting the hang of it so the next session (not with GTX512) coming will be alot easier.

Quote:

---

Originally Posted by Kinc

Tricky little thingy :)

What HTT, voltage and base temp (ln2 or dice?) are you runing for 3450MHz?

3514MHz is stable enough at the right base temp. It is just that three pots on ln2 when running max oc on the cards are pretty tuff to control by your self. The FX-60 temp difference in idle and load is nuts, never seen anything like it. But I'm getting the hang of it so the next session (not with GTX512) coming will be alot easier.

---

For now I've only been using 17 x 203 MHz (5x HTT) on RDX200 and Vcore: 1,50 * 109,8%. Cooling: dry ice. If my container's base temperature drops below -50C desktop usually freezes. The sensor is a little bit off, so the real value would be around -57C or something. :)

GT2 seems to be very strong with Dual Core FX-60, but FX-57 @ almost 4 GHz is too good in GT1 :)

tuned.....great guys.....I love this game !

Quote:

---

Originally Posted by Sampsa

For now I've only been using 17 x 203 MHz (5x HTT) on RDX200 and Vcore: 1,50 * 109,8%. Cooling: dry ice. If my container's base temperature drops below -50C desktop usually freezes. The sensor is a little bit off, so the real value would be around -57C or something. :)

GT2 seems to be very strong with Dual Core FX-60, but FX-57 @ almost 4 GHz is too good in GT1 :)

---

@ -55C basetemp I could boot @ 18x200MHz. Stable in windows but it would hang after a few secs in 3dmark.

GT1 is dependent on HTT/Mem 1:1, with around 250MHz HTT @ 3500MHz a FX-60 will defenetly take a FX-57 @ 4GHz.

Have you tried 3dmark 06 with your setup yet? :)

Quote:

---

Originally Posted by Kinc

@ -55C basetemp I could boot @ 18x200MHz. Stable in windows but it would hang after a few secs in 3dmark.

GT1 is dependent on HTT/Mem 1:1, with around 250MHz HTT @ 3500MHz a FX-60 will defenetly take a FX-57 @ 4GHz.

Have you tried 3dmark 06 with your setup yet? :)

---

I just picked up dryice and I'm heading to office :)

subscribed.....nice X1900 cards you've got there Sampsa

Quote:

---

Originally Posted by dinos22

subscribed.....nice X1900 cards you've got there Sampsa

---

Where do you see X1900 cards?

he said he allready benched new gen so its not hard to figure i guess:D

Quote:

---

Originally Posted by Jort

he said he allready benched new gen so its not hard to figure i guess:D

---

Ahhhhh okay I see.:)

Quote:

---

Originally Posted by Sampsa

Kinc, FX-60 sure is tricky. 3,4 GHz and 3,45 GHz are pretty easy to bench but 3,5 GHz locks up too damn often.

I'm currently working with new setup. Here is a score with graphics cards stock cooled and FX-60 @ 3,45 GHz.

http://www.io-media.fi/temp/orb.png

:)

---

Looks like another awesome cherry! ATI style!

Quote:

---

Originally Posted by Sampsa

I just picked up dryice and I'm heading to office :)

---

Looking forward to the results :)

Quote:

---

Originally Posted by Kinc

Looking forward to the results :)

---

As am I...this is really starting to pick up. :banana:

That REALLY sux about the cards Kinc, but ah well I guess. :(

Can't wait to see what that "mystery" setup of Sampsa's is capable of if the cards are still on stock cooling. :cool:

Quote:

---

Originally Posted by Cornelious0_0

As am I...this is really starting to pick up. :banana:

That REALLY sux about the cards Kinc, but ah well I guess. :(

Can't wait to see what that "mystery" setup of Sampsa's is capable of if the cards are still on stock cooling. :cool:

---

Yeah it really does. But I had a good conversation with em the other day that made me very happy ;)

So.... you're getting the 7900 to play with then..... :D

NC™

Quote:

---

Originally Posted by Kinc

Looking forward to the results :)

---

17475 with Radeon X1800 CrossFire

And here is a teaser with new setup:

http://www.io-media.fi/sampsa/tests/orb2.png

It's all about system power now :)

Damn!!!! :slobber:

NCâ„¢

omg:slobber:
sampsa that is sick!
thats 15 fps over the 17.5k score! and if the others go that much higher ... :slobber: :woot:

so will we see this on monday:D
monday is release date for x1900 right? :fact:
so are you keeping it back till then?:toast:

Quote:

---

Originally Posted by phelonius

omg:slobber:

thats 15 fps over the 17.5k score! and if the others go that much higher ... :slobber: :woot:

---

GT1 and GT2 are completely CPU/system limited.

And unfortunately, in somewhat opposite directions (GT1 likes MHz while GT2 likes the second core)....anyway, there IS improvement in those two tests, right Sampsa? Otherwise you're not much over 19k, if at all.

EDIT: without improvement in GT1 and GT2, you're barely over 18.5k.

Quote:

---

Originally Posted by Sampsa

GT1 and GT2 are completely CPU/system limited.

---

Are you still running on a very low htt? Try to get the htt up since it will increse the performence in GT1 quite much. 14x250 = 3514MHz gave me almost the exact same performence as of my 3930MHz FX-57 in GT1, the test would end at about 7000 frames. It worked fine to use the 5X multipel for the bus also.

Quote:

---

Originally Posted by Kinc

Are you still running on a very low htt? Try to get the htt up since it will increse the performence in GT1 quite much. 14x250 = 3514MHz gave me almost the exact same performence as of my 3930MHz FX-57 in GT1, the test would end at about 7000 frames. It worked fine to use the 5X multipel for the bus also.

---

For some reason system becomes very unstable if I try to raise the HTT under Dry Ice. Even 216 MHz (5x HTT) is too much for 16x. Strange.

Using memory dividers is no problem at all and I'm running my mem @ +250 MHz, only HTT.

Quote:

---

Originally Posted by Sampsa

For some reason system becomes very unstable if I try to raise the HTT under Dry Ice. Even 216 MHz (5x HTT) is too much for 16x. Strange.

Using memory dividers is no problem at all and I'm running my mem @ +250 MHz, only HTT.

---

With LN2 I think it would be possible, since my FX-60 will not even boot in to windows with dry ice. Yours seems to like the cold better. When adjusting HTT I'm keaping the basetemp at -45C then when I start 3dmark I quickly lower the temp to about -55C. The hard part is to save the score :) The basetemp needs to start dropping as soon as the last test starts to fade. I never had the time to complete a run with max oc on the cards at that cpu speed :(

So far -35c np for this one:
http://pic8.picturetrail.com/VOL256/.../126595665.jpg
More voltage should allow better load temps. below -45c had issues with this vcore.