For x1800 xl users

Shinryu

Limp Gawd
Joined
Oct 23, 2005
Messages
297
Hey I want to know how much energy in watts does the x1800 xl consume and is there something wrong with the card (like voltage leaks and such)?

Also how much was household people able to overclock there card too?
 
The card hardly pulls more than a 7800GTX so power consumption should not be a concern. Remember, it only requries a 500watt power supply so it can't be so bad right?
 
R1ckCa1n said:
The card hardly pulls more than a 7800GTX so power consumption should not be a concern. Remember, it only requries a 500watt power supply so it can't be so bad right?

how is ur card? managed higher oc? my sapphire retail x1800 xl is coming, i couldnt stand IQ on 7800.
 
I'm ordering one of these cards today. Should I buy retail or oem ?
I want a good oc core of course ! Most of these are hitting xt speeds ? Would they be the same or will retail get slightly higher grade chips ? Anyone ? :)

FWIW I like the alien on the sappire retail better than the art on the oem. :LOL:
 
Stereophile said:
I'm ordering one of these cards today. Should I buy retail or oem ?
I want a good oc core of course ! Most of these are hitting xt speeds ? Would they be the same or will retail get slightly higher grade chips ? Anyone ? :)

FWIW I like the alien on the sappire retail better than the art on the oem. :LOL:
OEM / Retail, same card different box.
 
R1ckCa1n said:
The card hardly pulls more than a 7800GTX so power consumption should not be a concern. Remember, it only requries a 500watt power supply so it can't be so bad right?

the x800xl does not require a 500 watt powersupply
 
Ok thanks R1ck. Have an oem on the way.
Does aftermarket air cooling do much for oc these cards ?
 
Please. I'm running an X1800 XL OC'd and an OC'd X2 4400+ with 2GB of OC'd OCZ RAM on a 300W (fanless no less) PSU. You need quality not quantity.
 
Stereophile said:
Ok thanks R1ck. Have an oem on the way.
Does aftermarket air cooling do much for oc these cards ?

Yes, the Zalman VF700 Cu is noticeably better than the stock unit.
 
Bar81 said:
Please. I'm running an X1800 XL OC'd and an OC'd X2 4400+ with 2GB of OC'd OCZ RAM on a 300W (fanless no less) PSU. You need quality not quantity.
what clock speeds did you get on your x1800xl?
 
ClearM4 said:
what clock speeds did you get on your x1800xl?

The overclocking tool doesn't work properly (the scores in game don't change at all) so I'm just using CCC. The Zalman is set at 5V. At that I can get 550 core. I tried 575 but it locked up; haven't tried in between. Will play around more, but I'm running a quiet rig so max overclock isn't my biggest concern, quiet is. Still, runs under 60C on load in a very low airflow case. Memory is rated at 1400mhz so once an OCing tool is available it'll be OC'd that high, for right now it's at 1100 which is the max allowed in CCC.
 
Bona Fide said:
You're running all this off a 300w fanless PSU? I find that hard to believe.

It may be hard, but believe ;) Like I said, it's quality, not rating.
 
Bar81 said:
Please. I'm running an X1800 XL OC'd and an OC'd X2 4400+ with 2GB of OC'd OCZ RAM on a 300W (fanless no less) PSU. You need quality not quantity.
http://www.extreme.outervision.com/index.jsp

2 sticks DDR + X2 4400+ No OC and 7800GTX @ 490MHz (because there is no X1800XL) and a 7200RPM drive comes out to 260W on there. While you may have a quality PSU (IDK what PSU you have) according to your information and that website, you are probably within 300W anyway.
 
Shameless Liar said:
http://www.extreme.outervision.com/index.jsp

2 sticks DDR + X2 4400+ No OC and 7800GTX @ 490MHz (because there is no X1800XL) and a 7200RPM drive comes out to 260W on there. While you may have a quality PSU (IDK what PSU you have) according to your information and that website, you are probably within 300W anyway.

Actually, that calculator WAY overshoots the power draw. According to it when I input my entire system specs I need 370W power draw. I know that's false because my APC unit tells me exactly the power draw. At idle it's 186W without CnQ and high 160s with CnQ. Based upon numbers I've seen there's simply no way that my system as a whole exceeds 300W under load (it gets quite close but not over 300W - besides if it exceeded 300W the APC unit would be beeping and freaking out; I've done it before :D). Just take a look at these numbers that include an X2 4800+ and 7800GTX:
http://www.gamepc.com/labs/view_content.asp?id=pm780&page=3

If the power draw was more than 300W I'd be in trouble no matter the quality of the PSU (unless of course it exceeded spec). What I was saying was that a monster PSU isn't needed if it's quality. If you get the crap that seems to be popular amongst gamers then you need to overshoot on the rating.

PSU is this:

http://www.silverstonetek.com/products-st30nf.htm

Oh, just for the record complete specs include Raptor 74GB, 2x Maxtor MaXline III 300GB in RAID 1 and BenQ 1640B, along with 3 120mm Nexus case fans running at 5v fanmated along with Audigy 4 and Vortex 2 with Roland MIDI daughtercard.
 
I've noticed that lots of people assume you need high capacity power supplies on this forum (and others)
 
Bar81 said:
The overclocking tool doesn't work properly (the scores in game don't change at all) so I'm just using CCC. The Zalman is set at 5V. At that I can get 550 core. I tried 575 but it locked up; haven't tried in between. Will play around more, but I'm running a quiet rig so max overclock isn't my biggest concern, quiet is. Still, runs under 60C on load in a very low airflow case. Memory is rated at 1400mhz so once an OCing tool is available it'll be OC'd that high, for right now it's at 1100 which is the max allowed in CCC.

Ouch... I saw a lot of these on rage3d supposedly doing 600-650 on stock cooling.
 
Stereophile said:
Ouch... I saw a lot of these on rage3d supposedly doing 600-650 on stock cooling.

I really think the utility just isn't working properly. In my tests in F.E.A.R., which is heavily GPU limited, there was no difference in framerate between 500/500 and 550/675 at the settings I'm running 1280x768 High everything except no soft shadows 2x AAA and 8x HQ AF. That strikes me as very strange.

Besides, if all those boards were doing that speed then why did ATI have to issue another revision core, the 15 (all the XLs are 14), and slap on a dual slot cooler to be able to reach 625mhz and delay the launch to build up quantity.
 
Bar81 said:
I really think the utility just isn't working properly. In my tests in F.E.A.R., which is heavily GPU limited, there was no difference in framerate between 500/500 and 550/675 at the settings I'm running 1280x768 High everything except no soft shadows 2x AAA and 8x HQ AF. That strikes me as very strange.

Besides, if all those boards were doing that speed then why did ATI have to issue another revision core, the 15 (all the XLs are 14), and slap on a dual slot cooler to be able to reach 625mhz and delay the launch to build up quantity.

Did you try the overclocking tool posted on the forums? Both nunyabiz and rickcain hit overclocks above 650mhz with voltage raised to xt level in ccc.
http://www.hardforum.com/showthread.php?t=969139

I think the problem for ati is getting 1500mhz ram. plus 512 mb of it for one card.
 
ClearM4 said:
Did you try the overclocking tool posted on the forums? Both nunyabiz and rickcain hit overclocks above 650mhz with voltage raised to xt level in ccc.
http://www.hardforum.com/showthread.php?t=969139

Yup, used the revised one from that link (2nd release) I'll play around with it a little more and see if I was doing something wrong. Nope setting both 2d and 3d yields the same numbers.
 
Bar81 said:
Yup, used the revised one from that link (2nd release) I'll play around with it a little more and see if I was doing something wrong.
You should just test overclocks with 3dmark05 to see if score increases. If score increases then overclocks are working.
 
ClearM4 said:
You should just test overclocks with 3dmark05 to see if score increases. If score increases then overclocks are working.

I'd prefer to test with a real application. I don't care about synthetic numbers.
 
Bar81 said:
I'd prefer to test with a real application. I don't care about synthetic numbers.
I only suggested using it cuz you said something about the overclock untility was working and with 3dmarks you could check if there is an increase in performance. synthetic numbers are useless for gaming, but not for testing the same card at different settings.
 
ClearM4 said:
I only suggested using it cuz you said something about the overclock untility was working and with 3dmarks you could check if there is an increase in performance. synthetic numbers are useless for gaming, but not for testing the same card at different settings.

Yeah, but that application in particular is known for shady reporting. I just don't trust it.
 
Bar81 said:
Yup, used the revised one from that link (2nd release) I'll play around with it a little more and see if I was doing something wrong. Nope setting both 2d and 3d yields the same numbers.

that oc utility is pretty much idiot proof so I doubt you are using it wrong. and it DOES work, my fillrate in 3dmark01 is near 10,000 MP/sec with my core at 628. With my core at 500 fill rate is around 7700, as one would expect it to be. your prolly just got a crappy core.
 
LawGiver said:
that oc utility is pretty much idiot proof so I doubt you are using it wrong. and it DOES work, my fillrate in 3dmark01 is near 10,000 MP/sec with my core at 628. With my core at 500 fill rate is around 7700, as one would expect it to be. your prolly just got a crappy core.

Like I said, it's not the OC I'm talking about but the fact that in game scores don't change with the OC. Try some games and see if your scores changed.
 
Bar81 said:
Yeah, but that application in particular is known for shady reporting. I just don't trust it.

Nothing shady about it, as mentioned before ok tool for checking out different settings and clocks. You have been reading too much Hard forums if you dont "trust" the application. Its not going to invade and violate your computer or anything :D
 
benchmark programs are good for a quick view to see if your overclock did anything

now if you wanted to compare the actual speed of the competition, its a bad indicator actual performance
 
Alarmer said:
Nothing shady about it, as mentioned before ok tool for checking out different settings and clocks. You have been reading too much Hard forums if you dont "trust" the application. Its not going to invade and violate your computer or anything :D

Nah, I don't trust the results. I know it's not spyware or anything, just the results aren't reliable. Real apps are what I use and what I rely on in performance evaluations.
 
Bar81 said:
Nah, I don't trust the results. I know it's not spyware or anything, just the results aren't reliable. Real apps are what I use and what I rely on in performance evaluations.

Results are reliable, Specially when one overclock tells you the fillrate is 7000, and the next overclock tells you its 10k. Now that REAL application is telling you the overclock worked and its stable. Now it wont tell you a damn thing about perforamce on a game. But will tell you your overclock worked (and if its stable or not)

Could always run time demos..........
 
DASHlT said:
Results are reliable, Specially when one overclock tells you the fillrate is 7000, and the next overclock tells you its 10k. Now that REAL application is telling you the overclock worked and its stable. Now it wont tell you a damn thing about perforamce on a game. But will tell you your overclock worked (and if its stable or not)

Could always run time demos..........

That's what I'm saying. I'm running the timedemo and the numbers are the same. Something is wrong with the app.
 
I inititally had similar problems with the OC tool and thought that it didn't work.

I uninstalled the driver + CCC and installed the 5.10a based Omega drivers. Now it works perfectly, you can just pop in clock values, hit the button, the screen blanks for a few seconds and ..... your FPS go up. :)

I suspect Overdrive interferes with the settings on the fly (or when 3D apps are opened).

Try the above and do let us know how you get on ..... I've left mine burning in at 625/700 (8500 in 3DM05 with an A64 3000+ @ 2.7GHz) with stock voltage, water cooled with a TT rocket (only cooling the X1800XL).
 
nemesis.ie said:
I inititally had similar problems with the OC tool and thought that it didn't work.

I uninstalled the driver + CCC and installed the 5.10a based Omega drivers. Now it works perfectly, you can just pop in clock values, hit the button, the screen blanks for a few seconds and ..... your FPS go up. :)

I suspect Overdrive interferes with the settings on the fly (or when 3D apps are opened).

Try the above and do let us know how you get on ..... I've left mine burning in at 625/700 (8500 in 3DM05 with an A64 3000+ @ 2.7GHz) with stock voltage, water cooled with a TT rocket (only cooling the X1800XL).

You're right, CCC overrides the tool. I unistalled it and the overclocker works fine. At 1.2v I was running at 600 core (didn't try any higher). Memory must be severely undervolted as it wouldn't even do 650 (which is well below it's rated speed) as the artifacts hit.

Problems:
Omega driver makes my desktop blurry (not a problem with official CATs)
Omega drivers based upon 5.10a and not the better 5.10b
Memory voltage control does not work

The biggest problem is ATI Tray Tools still doesn't recognize my card for whatever reason with the beta drivers so I can't enable AAA or HQ AF which is a dealbreaker. I'm going to see if installing CCC and not unlocking the card via CCC allows me to use the overclocker.
 
Bar81 said:
Problems:


Memory voltage control does not work

.
How do you know its not working? You put a meter on the memory? I noticed that raising the mem voltage did not let me clock higher, is that what you mean?
 
Well, good news. As long as you don't unlock the CCC Overclock utility the overclocker works fine. Now, if only we could get mem memory adjust we would be set. My mem doesn't go higher than 600 even though it's rated for 700. The core will do 600+ with a slight voltage bump to 1.2v (from 1.1v)
 
Alright, alright, I stand FULLY corrected. The overclocker works completely. I was able to OC to 600/700 and the core could probably go higher but I didn't test past 600 as that's enough for me.

What I did:

Install:
(1) ATI Catalyst 5.10b with CCC
(2) Installed Catalyst 5.10 WHQL WDM drivers ONLY (not the Catalysts or CCC)
(3) Went into CCC and set a bunch of things the pertinent items being:
(a) Went into AA and deselect app control and then select Adaptive AA and then reselect app control (which keeps the AAA selected)
(b) Went into AF and selected HQ AF
(c) Confirmed that the overclocking utility is LOCKED

(4) Ran overclocker. Important notes:
(a) Any core clock over 550 was done by upping core voltage to 1.2v
(b) Any memory clock over 600 was done by upping memory voltage to 2.0v

Scores tested in F.E.A.R. 1.01 Patch
Resolution: 1280x768 [16:9]
Graphics: Everything max except no soft shadows. 2x AAA, 16x HQ AF
Sound: Everything including EAX HD selected

System:
Athlon 64 X2 4400+ at 2.4Ghz cooled by Scythe Ninja
2x 1024GB OCZ EL PC3200 DDR Running at 220 2.5-3-3-8-1T
ATI X1800 XL with Zalman VF700 Cu running at 5v setting with Zalman ramsinks
Audigy 4
ASRock 939Dual-SATA2 BIOS 1.30
Raptor 74GB
Silverstone 300W Fanless PSU

First clock is core/second is memory. The first line is min/avg/max and the second line is percent spent uder 25fps/between 25 and 40 fps/ over 40fps.

500/500 (stock)
31/49/96
0/36/64

550/600
33/52/102
0/28/72

600/625
32/57/113
0/13/87

600/650
35/57/112
0/13/87

600/675
35/58/112
0/12/88

600/700
37/59/119
0/12/88
 
Ya, ati really news to release x1800 offical drivers with overclocking tools (5.11 drivers?). I am waiting for this to happen. In the mean time I'll decide if the x1800xl or x1800xt is best for me. Considering the x1800xt reviewers were testing ran at 1.3v I am leaning towards the x1800xl with it under $400 and everyone hitting over 600mhz on the cores with 1.2v. My watercooling setup will keep it extra quiet too.

PS. If your spending over $300 on a video card you really should have a water cooling setup. :)
 
ClearM4 said:
PS. If your spending over $300 on a video card you really should have a water cooling setup. :)

No need. All that noise from the radiators, maintenance, and extra expense; not to mention risk however slight of a leak. I'm running a quiet aircooled system that's a monster. With the new aircoolers for CPUs and GPUs watercooling is completely unnecessary. It's nice if you want it but I for one don't. Planning a system intelligently gives you 95% of what WC would give you for a lot less money and headache and maybe even a quieter system in the end.
 
Back
Top