VF-900cu results on a x1800 or x1900 video card

[Eastcoasthandle]

Ok, it's time for you to let others know how well the VF-900cu performs.
All you need to do is:
-report which video card you have
-idle and load times with and without the VF-900cu

 

[PureBooYah]

what i posted in another thread with my x1900xtx....

This is what I got when running a little BF2....

Using ATItool...
710mhz core OC
Core voltage 1.45v

Stock cooler with AS5
idle: 46c (fan 100%)
load: 78c (fan 100%)

VF900
idle: 40c (fan 100%)
load: 84c+ (fan 100%)....my monitors would actually black out because of my OC becoming unstable, but i could still hear sounds.

 

[NoDamage]

I compiled a preliminary list of user reports over and AnandTech about a week ago:

http://forums.anandtech.com/messageview.aspx?catid=31&threadid=1839956&enterthread=y

The general result seems to be that most users aren't getting very good performance out of the VF900 on the X1900XTX. There have been a few cases of really good results (50-60C load temps) but the vast majority of them have been kind of discouraging. The discrepancy is enough to make me think that some of the VF900's may have manufacturing issues or some such.

Anyway there have been a dozen or so new reports since then. I will go ahead and update the list one of these days when I find some more free time.

 

[cyks]

The discrepancy is enough to make me think that some of the VF900's may have manufacturing issues or some such.

I hope you didn't pull that conclusion from those reports, because they don't mean anything. With the inherently high temps or OCs aside: from what I hear, the stock cooler is pretty loud and the 900 is an open-air HSF that needs special airflows to perform best. In addition, I didn't notice any ambient pc temps for further investigation, so yea...

 

[PureBooYah]

I hope you didn't pull that conclusion from those reports, because they don't mean anything. With the inherently high temps or OCs aside: from what I hear, the stock cooler is pretty loud and the 900 is an open-air HSF that needs special airflows to perform best. In addition, I didn't notice any ambient pc temps for further investigation, so yea...

but with that many people having bad results and especially after experiencing it myself, I think it's a bad cooler for the 1900 series (not saying there's some manufacturing issues though). My ambient temps are around 70f and the air flow in my case is pretty good. My VF900 performed the same even with the side of the case off.

 

[NoDamage]

I hope you didn't pull that conclusion from those reports, because they don't mean anything. With the inherently high temps or OCs aside: from what I hear, the stock cooler is pretty loud and the 900 is an open-air HSF that needs special airflows to perform best. In addition, I didn't notice any ambient pc temps for further investigation, so yea...

I acknowledge the difference due to various ambient temps/clock speeds/voltages, as most users didn't provide that much information. But like I said in that thread, there are some surprisingly consistent results that can still be drawn across all of the different reports.

The vast majority of users are reporting similar temps for the stock cooler on auto (50C idle 80-90C load), stock cooler at 100% (40C idle 60-70C load), Accelero X2 on auto (40C idle 70-80C load), and Accelero X2 at 100% (40C idle 60-70C load). This is generally consistent regardless of the clock speeds/ambient temps.

The only major discrepancy is with the VF900 results. Most of them are in the 40C idle 70-80C load range, but a couple of users are getting very good results of 30C idle 50-60C load. I find it hard to believe that case airflow and ambient temps could be the sole factor in explaining this huge difference. Especially since most of time, their stock cooler temps are the same, and it is only the VF900 results that differ drastically.

I mean, it is total speculation on my part that there is something wrong with some of the VF900's and I could be totally off. But these results are just completely confusing.

Let me put it this way. After looking at 20+ user reports across 5-6 different forums, most of the time when I see a new report with the stock cooler and Accelero, I think to myself "yeah that is pretty consistent with what I have seen so far". It is only when I come across the occasional VF900 report that I am surprised and wonder "wow, this guy is getting awesome temps, but why are all these other people getting such poor performance?"

 

[cyks]

Who said anythign about the accelero? Nevertheless, the results lacked a reliable constant. I do agree that claiming their HSFs malfunctioned is speculation. But, OCing with a poor understanding of convection is more likely suspect (although that too is nothing more than speculation, due to the nature of inconclusive reports). Agree to disagree-- for skience's sake.

 

[Croak]

(cut to the chase for those that don't like reading long posts...VF700cu (NOT a VF900cu) on my 675/750 stock voltage 1900xt keeps me at 40-43 idle at 2200rpm/8v, 78c max at load so far, 8v-12v/2200-2800 RPM variable on the fan. Stock was 55-60c at idle, and 85c-90c+ at load. Winner: VF700cu)

Just as an oddball data point for you Zalman skeptics (or believers), I threw a VF700cu on my 1900xt today, because I just couldn't stand the freakin' noise of the stock cooler after a week of using it. I'm VERY happy with the performance, even though I've been an Nvidiot since the RIVA 128, and I got off $100-$200 lighter than what I was going to spend on a 7900 solution.

But the noise was so bad I would have taken it back if they had any other high end GPU's for sale in my area...the best I can find locally is a 7600gt or X1600's. The 1900xt was an open box unit that was stashed in the back of my local Fry's.

Since I didn't want the downtime waiting for a 7900GTX (or even a pair of 7900 GT's) to ship, I bought the X1900 (and I got a pretty good deal on it considering it's a retail store, $479). It's an older unit (I assume) with the BJ11 RAM.

I'm a "touch-feel gotta have it now" kinda guy, so I'm not real fond of internet ordering and the waiting involved, even with better supply and lower prices.

But getting back to the VF700cu...once again, it was the best cooler available locally, no VF900's to be found, no V2's, just the VF700cu and cu/al units.

Before I did the swap, I was seeing 55c-60c at idle, and 85-90+ under load (my "load testing" being BF2 for a few hours at 1650x1080 and everything cranked up, and some alt-tabbing out to CCC).

With the VF700cu, I see 40-43 idle, and 78c is my highest load. Nothing to complain about there, and the noise is GONE. My 120mm Antec case fans set to low/1200RPM are louder unless I crank the fan voltage to 12v, at which point it's barely louder. No other prepping besides using Arctic Silver-5 instead of generic grease...no lapping, no polishing, just some toilet paper to clean the factory grease off the R580, and a wipedown of the ram before sticking the Zalman sinks on them (and a AS-5 apply/wipe off of the VF700 base).

I should also explain that I'm running the VF700 off a motherboard header on my AN8-32X (the NB fan header, to be exact) and I'm using uGuru to regulate speed..since the lowest voltage I can send through the MB headers is 8v, that's my low setting, 2200 RPM. I've tied the voltage ramp to SYS temp, it kicks up to 12v/2800 RPM at 45c in the case, which happens about half an hour into BF2.

Using the MB header, my idle fan speed is 1000RPM higher than what the VF700 would spin using the Zalman supplied low speed connector, which might well account for my decent idle temps, and having it kick up as the case temp climbs helps too, something you don't get using the Zalman connectors. No doubt the AS 5 is helping too, and might just get better in a few days.

But the really important thing for me is that it's SOOOOOO much less annoying than the stock fan, even though it's running at 8-12v.

My current OC on the X1900 (675/750) is solid and stable, and that was set using CCC's auto-detect with the stock fan. Since I'm not planning on doing mega-overclocking with the X1900, the VF700 does the job for me.

Those of you trying to push closer to 800/900 probably won't care for the VF700, nor would those of you with cramped cases. I should also note that I'm using a P-180, and I think the VGA cooling duct is actually helping, and my case/CPU temps are nearly identical to what they were with the stock ATI fan.

I can only speculate, but I think on my system, the VF900 would be even more effective, and quieter too, but as long as the VF700 is working for me, I don't see the need to shell out another $50 (provided somebody around here ever gets one in stock).

Stuff I'm using:

AN8-32X
X2 4400+ @ 2706 (246x11), 1.475v
ATI X1900XT @ 675/750 (set using CCC autoclocking before the new fan)
Antec P-180 with factory fans set to low
Enermax FMA 535w PSU
2x Seagate 7200.7 80gb in RAID 0
Zalman CNPS9500
Zalman VF700cu

EDIT: Room temp/ambient is 20-22c, CPU idles at ~32c, 55c at load, SYS temp is 32c, ~45c at load (after quite some time), and PWM temp is 40c, 48ish under load.

 

[xX_Jack_Carver_Xx]

I just installed and tested the Accelero2 on my X1900XT.

701/802 1.52/2.151/2.151/1.50V

Stock 42C IDLE, 78C LOAD

Accelero2 40C IDLE, 69C LOAD

Doing exact same tests, in same system.

My core was slightly higher than the SHIM, so no issue with shim height, Accelero2 had perfect contact with core, using translucent coating of AS5 between them.

Im sold.