A little X1950XTX + P5B DLX benches

Printable View

10-06-2006, 11:01 AM
SIOUX

A little X1950XTX + P5B DLX benches

Played around with a new Sapphire X1950XTX the last couple of weeks. Now, i must say i am a little disapointed in the card. Its deffenetly slower than X1900XTX as long as mem speed on the X1950 is limited to around 1143:( i have gotten far better results with the X1900 so far.

But here is what I got. No windows tweaks yet:) will have a CF card in 1 week, and will test the P5B DLX and ABIT AW9D-MAX then...stay tuned:)

Test Setup:

P5B Deluxe C1
E6600ES on VAPO LS
Sapphire X1950XTX on chilled water with EK waterhead
G.skill 1gb HZ
OCZ GX 700w
WD Raptor 74gb 16mb

http://img434.imageshack.us/img434/4...0123d01sg1.jpg

http://img402.imageshack.us/img402/2...4773d05bx9.jpg

And some spi just for the fun of it:D realy suffering with only 1gb in 32m though.

http://img414.imageshack.us/img414/786/111561mii5.jpg

http://img47.imageshack.us/img47/463...437532mke5.jpg
10-06-2006, 03:50 PM
G H Z

What happens at the 1143 limit? Does it artifact or just do nothing at all? Thats still a very nice '01 :cool:
10-06-2006, 03:58 PM
SIOUX

Quote:

Originally Posted by G H Z

What happens at the 1143 limit? Does it artifact or just do nothing at all? Thats still a very nice '01 :cool:

It seems that after 1143-1150 (variates in my case) the screen just artifacts like crazy.Nomatter voltage. I havent seen any X1950 over 1150-1155 so i think it must be a bug of some sort:)
10-07-2006, 01:47 PM
SIOUX

And a 06 with humble cpu clocks:)

http://img166.imageshack.us/img166/3...1950xtxzu3.jpg
10-13-2006, 01:23 PM
G H Z

Quote:

Originally Posted by SIOUX

It seems that after 1143-1150 (variates in my case) the screen just artifacts like crazy.Nomatter voltage. I havent seen any X1950 over 1150-1155 so i think it must be a bug of some sort:)

Have you tried editing the ATI Tool profiles.ini with higher clocks and using it rather than the sliders? That worked for me in a similar situation on X1900 around 900mhz with lower voltages, maybe it might work here. Your ORB 1900 score is fast on system tests, in fact almost as fast as mine with 200mhz less on CPU :toast:
10-13-2006, 01:52 PM
Viss

Tried that already but stil same, artifacts all over.
10-13-2006, 02:10 PM
SIOUX

Quote:

Originally Posted by G H Z

Have you tried editing the ATI Tool profiles.ini with higher clocks and using it rather than the sliders? That worked for me in a similar situation on X1900 around 900mhz with lower voltages, maybe it might work here. Your ORB 1900 score is fast on system tests, in fact almost as fast as mine with 200mhz less on CPU :toast:

Thx :) but already tryed that, and no go:( My X1900 Score was also with a 100% tweaked OS :)
10-13-2006, 02:26 PM
J0d3r

Nice 6600 you got here, 4570 mhz 1m stable with stock vapo LS its a nice result.

What vcore are you using? temps?

Thanks
10-13-2006, 03:34 PM
SIOUX

Quote:

Originally Posted by J0d3r

Nice 6600 you got here, 4570 mhz 1m stable with stock vapo LS its a nice result.

What vcore are you using? temps?

Thanks

Using 1.7v from BIOS. 1.65v Actual vcore. Temps around -21 idle, and around -9 load:)
10-13-2006, 03:40 PM
J0d3r

amazing! with 1.9v like epox4life 6600's and cascade/ln2 that chip would fly

Thanks
10-13-2006, 04:00 PM
Pedro Rocha

Very, very good single card scores my friend, as usual :toast:

What voltages are using on the X1950 ?

Did you raise just the VCore or the VDCCI also ?

Seems beta16 ATITool set of VDCCI are not accurate ..
10-13-2006, 04:04 PM
IvanAndreevich

Wow.. can't wait to see the 1043 limit on memory gone.
10-13-2006, 04:35 PM
SIOUX

Quote:

Originally Posted by Pedro Rocha

Very, very good single card scores my friend, as usual :toast:

What voltages are using on the X1950 ?

Did you raise just the VCore or the VDCCI also ?

Seems beta16 ATITool set of VDCCI are not accurate ..

Thx alot my friend:toast: I´m using 1.575 for core, stock volt on MEM, 2.150v on MVDDQ and didnt touch VDDCI since its incorrect as you say:)

I realy hope the mem limit is broken somehow soon:)
10-14-2006, 02:47 PM
M.Beier

Impressive as usual Kenneth :)

I'd sure like to have a word with ya one of the days in the holiday.. But tomorrow morning Im off @ vacation, is there by any chance I can catch ya saturday next week?? I'd like to hear som' about your mobo experiences as I consider changing mine but wanna make the right pick first time, spend 37K in the last two months due movin' and startin' university :eek: (In other words, Im very short on cash, so I dont feal like ordering 3 mobo's :p: )
10-16-2006, 03:01 PM
vanovich

nice scores there ,nice 6600 youve got there . keep up the good work
10-19-2006, 05:38 AM
perry_78

And now to fire it up under the pot eh? :D
10-19-2006, 05:51 AM
SIOUX

Quote:

Originally Posted by perry_78

And now to fire it up under the pot eh? :D

:D me and brian just need to find a weekend now, and we will go for it...thx for letting us borrow the mousepot perry :toast:
10-19-2006, 06:36 AM
perry_78

No worries, looking forward to some nice benchies coming out of that ;)
10-21-2006, 01:05 PM
Revv23

hmm i was thinking about trading up but i guess its not ready yet...

i need to learn how to tweak 05, you 3d benchers outscore me by miles with lower clocks.. (not you soiux, i cant touch your clocks lol)
10-21-2006, 09:27 PM
creidiki

The problem is ATI is using the very cheapest (and slowest) GDDR4 chips for the current X1950XTX runs - probably because of yelds... you're hitting a hardware wall here, even in ideal conditions they just wont go much faster.
10-22-2006, 03:28 AM
SIOUX

Quote:

Originally Posted by creidiki

The problem is ATI is using the very cheapest (and slowest) GDDR4 chips for the current X1950XTX runs - probably because of yelds... you're hitting a hardware wall here, even in ideal conditions they just wont go much faster.

I must say i do not agree with this statement:) i think its a bug, course in my case 1142 is working 100% arties free... also in games nomatter if voltage is 2.0v or 2.4v 1142 is rockstable. but if i turn to 1143 the screen just get full of artifacts, and lock up in the end. but lets see what happens.....i´m not sure about anything with these cards:D
10-22-2006, 08:29 AM
Viss

Exactly the same here, 1143 is max. If it was chip you would think the wall would be somewhere around 1143, a little less for some and more for better cards. This isnt the case though, everybody seems to be stuck at 1143.
10-22-2006, 09:02 AM
SIOUX

Quote:

Originally Posted by loc.o

Exactly the same here, 1143 is max. If it was chip you would think the wall would be somewhere around 1143, a little less for some and more for better cards. This isnt the case though, everybody seems to be stuck at 1143.

yes..nomatter voltage...so strange:( :)
10-22-2006, 09:26 AM
Dynasty

nice single card sioux :toast:

the ram can handle way past 1143 on X1950's
...BC 09 very fast stuff....:)

With ati tool beta 15,
2.2 mvddc/mvddq in ati tool = 2.1 measured with DMM.
With active cooling on ram and real 2.25 mvddc/mvddq im still maxed @
1143...
10-22-2006, 09:43 AM
SIOUX

Quote:

Originally Posted by Dynasty

nice single card sioux :toast:

the ram can handle way past 1143 on X1950's
...BC 09 very fast stuff....:)

With ati tool beta 15,
2.2 mvddc/mvddq in ati tool = 2.1 measured with DMM.
With active cooling on ram and real 2.25 mvddc/mvddq im still maxed @
1143...

Thx Dynasty:toast:

Yes..i notived the voltage is very strange....only the core voltage seems to be right. Its the same with all the oc tools, and i cant seem to find any reason for these mem bugs:( :)
10-22-2006, 12:42 PM
Patriote

Quote:

Originally Posted by SIOUX

And a 06 with humble cpu clocks:)

http://img166.imageshack.us/img166/3...1950xtxzu3.jpg

Wow! Nice Score! Imagine what you'll get whit you Crossfire Edition card! :D Im also getting one this week!