NV 5 Silencer + 6800 GT : Detailed Numbers

T

tamislan

Weaksauce
Joined
Aug 10, 2004
Messages
81
Another Silencer thread, but this time a little more scientific.

BFG 6800 GT OC (the dual / "custom fan" version) and NV Silencer 5 from Arctic Cooling.

System Specs: See Signature plus:-
- Silverstone SST-B032FW Case
- 4x 80 mm Vantec Stealth Fans (2 rear ex., 1 inp. over cards, 1 inp. over hard drives)
- 1x 120 mm Enermax Adjustable Fan (UC-12FAB-B) set at about 80%.
- 430 W Antec True Power PSU

I installed a Rev 3 Silencer on my ATI 9800 Pro 128 MB last year and I was quite happy
with it. I very much liked the external exhaust so when the hot, hot, hot 6800 GT came
along I wanted an external exhaust for it too. So, after 3 weeks with my new card I
received my Arctic Cooling NV 5 Silencer this morning and installed it this afternoon.

I thought people may be interested in seeing results that included ambient internal case
temperatures in order to give some reference - if your case is running at 30 degrees C, you'll
get different results to me, but at least you have a frame of references since you know my
case temp, so you can repeat my conditions to some degree.

All temperatures are in Celsius/Centigrade). A thin film of Arctic Silver 5
was used on the GPU and a thicker, slightly goopy layer, was used on each of the RAM chips.
Each set compromises of numbers below consists of:

- Case Temperature (the internal case temperature reported by the motherboard using Abit EQ)
- Card Ambient Temperature as reported in the GeForce 6800 GT panel
- GPU Temperature reported in the GeForce 6800 GT panel

Stock cooler measurements were carried out about 3 hours prior to the Silencer 5 results.
Unfortunately, the room temperature raised by 2-3 degrees Celsius in the mean time due to
afternoon sun hitting the room. I would therefore take at least 2 degrees from the
post-installation Case temperature numbers in order to bring it into line with the Stock
Cooler. I may test the card again at the same time tomorrow if the room temperature is about
the same.

Load measurements were carried out after 15 mins running HDR demo in as large a window as
possible at 1280x1024 but still allowing me to barely see the temps. Please note that I had
already overclocked the chip in 3D to 400 MHz and I hadn't realised this until after I had
installed the Silencer.

BFG Custom Fan Stock Cooler
==================================

Idle - Card clocked at 370:1000 (2D setting)
--------------------------------------------
Case Temperature: 26
Card Ambient: 36
GPU: 57

Load - Card clocked at 400:1000 (3D setting)
--------------------------------------------
Case Temperature: 29
Card Ambient: 55
GPU: 84


NV Silencer 5
=====================================

Idle - Card clocked at 370:1000 (2D setting)
--------------------------------------------
Case Temperature: 27
Card Ambient: 32
GPU: 51

Load - Card clocked at 400:1000 (3D setting)
--------------------------------------------
Case Temperature: 28
Card Ambient: 45
GPU: 69


BFG cooler versus Silencer differences
======================================

Idle - 370:1000 (2D setting)
--------------------------------------------
Case Temperature: +1 (not counting
room temp change)
Card Ambient: -4
GPU: -6

Load - 400:1000 (3D setting)
--------------------------------------------
Case Temperature: -1 (not counting
room temp change)
Card Ambient: -10
GPU: -15

My own comments: Remember that I slipped up on the clock settings (3D was 400 MHz versus
2D in 370 MHz) so the load temperatures are certainly higher than they would be at 370 MHz.
I haven't investigated overclocking with the NV 5 yet. I'm saving that for later today but
I'll let you know. I am very happy with the temperature results though. I am very much in
favour of the DHES to prevent unnecessary heat increase inside the case.
 
Great review. It looks like the NV5 really does the trick.

As somebody with exactly the same card, waiting on the NV5 in the mail, I have a few questions:

-How hard/easy was it to do?
-Are the instructions that come with the NV5 any good?
-Are there are any quirks/issues with the BFG 2-fan HSF to watch out for?
-Do you have any tips/suggestions for doing the installation?

Thanks. :)
 
CastleBravo:

1) The only pain was removing the white thermalgunk from the RAM chips because it seems it had solidified - the stuff on the GPU wasn't solid. So the GPU wasn't so bad but I did find that there was too much thermal goop on the GPU and some of it had gotten onto the diodes (?) on the green PCB within the rectangular metal square around the GPU (I'm sure that there are photos of this on the net or this site even). I couldn't get every trace of it off the PCB :(. Overall, it was quite easy, it just took a long time to take off all the gunk and apply the AS 5 correctly.

2) The instructions are, quite frankly, a bad translation of the original German instructions. They'll get you there though - they aren't flawed when it matters. The way in which they say the HSF is attached to the card does not match with the BFG custom cooler. Just ignore that step on the instructions and unscrew the four screws on the back of the card to free the HSF from its bracket.

3) Not really apart from what I just said in number 2.

4) Take your time. I spent about 90 minutes doing it, mainly trying to
remove as much of the old white goop as possible from the RAM and applying AS 5 as well as I could (I've got it down to an art now).
Good luck with yours. Make sure you have some good thermal goop cleaner handy (I recommend pure isopropanol aka isopropyl alcohol).
 
Very warm day today. Ambient temp in the case after playing Halo for 2 hours is 31 degrees (stopped playing about 20 mins ago). So I decided to test my max Overclock. Coolbits won't let me go above 422/1100. I ran the HDR demo for about 10 mins at 420/1100 and the temps topped out at about 70 GPU / 47 Ambient, no artefacts. Not too different to my 400/1000 temps posted above. I think in order to go higher than 420/1100 I'll have to up the voltage in the card's bios and I'm not prepared to do that yet. My max no-artefact overclock before the Silencer was put on was about 415:1080. So not much of an improvement. However, the card is running a lot cooler. This makes me wonder about upping the voltage.

I installed ATItool yesterday to check for artefacts. Within 15 seconds of running the application the computer warm reboots. This is before I actually tell ATItool to do anything. I tried versions 0.20, 0.21 and 0.22. It will then usually crash once after getting into windows, even if I don't load ATItool, though the time it takes to crash varies. Then it won't crash again unless I run ATItool again. Obviously, ATItool seems to intiate something strange on my system. so for each run of ATItool I get 2 crashes.

I hadn't ATItool on the machine before I installed the Silencer so I'm wondering if it's just some weird incompability. The PC runs perfectly stable - I've had it on now for over 8 hours and I've restarted once or twice. Played Halo and Doom 3 several times for long periods now, no issues. It's a pity I can't use ATItool though - it's the best way to find artefacts and test your overclock.

Any comments?
 
Well, it's supposed to work fine on nVidia cards for checking artefacts. It simply won't change the clock frequencies of nVidia cards. I've used it on the 6800 GT in another PC without issue.
 
tamislan said:
Well, it's supposed to work fine on nVidia cards for checking artefacts. It simply won't change the clock frequencies of nVidia cards. I've used it on the 6800 GT in another PC without issue.
Click to expand...

Works fine with any card to check for Artifacts.
 
You must log in or register to reply here.
Back
Top