NV30- the fan will last how long, we must dust it how often?

Status
Not open for further replies.
i just wonder how much higher a 9700Pro would clock with that type fan solution. Maybe i can pick up a fried NV30 on ebay after release and strip the fan off to see.... I actually imagine it might decently cool a P4.

The general overriding theme in consumer graphics is (neglecting marketing numbsculls) only use what you must to obtain your targeted performance. I really believe they're only using this fan because they have to.
 
Now for those who changed out air conditioning filters well realize the amount of dust that is built up even in the cleaness of the clean. At work our A/C unit filter got so clogged up (lazy maintenance) that you could barely feel the air come out of the vents in the building, once replaced, a full force of air could be felt. We have to replace these about every month, two months for the dramatic decrease in air flow. I clean my home A/C filters once a week during the heaty months and they are packed with dust in just one week.

Now what does this have to do with the NV30?

Well I am convinced now that it would be naive to think that the GF FX design as shown will not have a flow problem due to dust build up. There is nothing magically different then a typical A/C unit with fins. Without filtration just like an A/C unit the cooling mechanism will point blank get clogged.

Ichneumon pointed something that Tomshardware brought to light which I almost can't believe.
"The demo board, which NVIDIA demonstrated in an nForce2 system, produced a lot of heat. The air coming out of the fan grille is hot to the touch. While the system was quite loud overall, we could still make out the Flow FX fan - not a very positive trait. NVIDIA has promised to refine the design to make it quieter. "

Hot to the touch? What temperture is that? I would say at least 120F on the outlet (best guess). Now if the inlet air was 75F with a delta T (difference of temperture from outlet to inlet) of 45F at its normal air flow rate, now if a dust buildup caused a 50% reduction then this would cause the delta T (difference in temperture) to double from 45F to 90F!!! Meaning that the inlet temperture would obviously not change but the outlet would go up to 165F!!!

Maybe the reason why the NV30 sucks and blows out behind your case is to protect the rest of your computer from overheating.
 
Dio said:
To anyone thinking of doing this I must add 'Your mileage may vary'. ATI recommend a 300W power supply for R9700Pro.

Thanks for clarifying... Actually, seems like mileage also varies for those with (generic) 300W power supplies...

Me still drools at the thought! :LOL:
 
Well, if you think dust build up on the sides can cause a 50% reduction in flow, I can see why you think this will be significant. If you think dust build up somewhere else will result in this, perhaps this is the source of our disagreement. To further clarify my thoughts:

I still don't personally see how the effect of dust buildup on the sides of the air passage could affect the creation of a laminar such that the effect of insulation of the dust layer rendered that effect insignificant.

I also don't see how dust buildup would act to increase the layer of laminar flow.

I do understand how a reduction in average flowrate can promote a laminar flow layer, but I don't see in the first place why laminar flow is significant concern for heat transfer in this situation, and in the second place how dust can achieve this effect from buildup on the sides of the air passage.

In fact, I'd wonder if the viscosity change due to temperature increase during travel along the assembly might be a more significant impact on laminar flow and further decreased heat transfer on the exhaust path (I suppose I should go plug some values into some equations, but I'm not sure where to get an idea for the impact of dust buildup on the sides for comparison).

I can see how the fan can be clogged to impact flow rate, and how filtering problems can be a concern, but I can also conceive that a fan design could be achieved to overcome that and that the propported similarity to laptop cooling designs seems to support that this is readily achievable.

Noko, or perhaps someone with more direct experience in air systems of this nature, perhaps my delineating these items more clearly provide an opportunity to clarify this.
 
Hello, first post so please go easy.


Some of the comments that I have seen on websites about the 2 different versions of the nv30 (geforce fx and geforce fx ultra) claim that the core speed will be scalable to 500mhz+. In my mind, that might mean that the standard version will be clocked at 500mhz and the ultra version slightly faster. Would that make sense?, 500mhz seems very high for a graphics card but 500mhz+ seems a little too far out.

Also, the competition to win the xpc with the nv30 could be using the standard nv30 without the large fx flow heatsink i.e. smaller in size. The small case might also of been specifically modified to contain the large heatsink. Also, the xpc version of the nv30 might be a lower clocked version of the nv30 i.e. an geforce fx ti4200 to lower the power requirements and cooling requirements even further.

A lot of people seem to be making a lot of assumptions i.e.

1. the fxflow will be on the standard version of the geforcefx, no one but nvidia are 100% sure either way.

2. the heatsink and fan will not be refined for the final release version, for all anyone here knows, that heatsink and fan could be a prototype and the actually real version will have some sort of external filter that can be removed and cleaned easily.

3. that the noise and dust will cause a serious problem to the final release version of the fxflow fan and that it will need to be cleaned every other week, as far as I can see, there is absolutely no proof either way to deny or confirm this argument.



As for the statement about a big heatsink making the card look more impressive to an uneducated buyer, I agree. If you handed the 9700 and the geforce fx with the fxflow to an uneducated buyer and said "which of these 2 graphics cards is faster?", you don't need to be a rocket scientest to know what they would tell you.

One thing that does trouble me, if all geforcefx's do have the fxflow fan and that they have somehow managed to squeeze it into a small form factor case, wouldn't that cause problems when you are cleaning the filters. In a standard case, removing the side of the case and removing the graphics card is pretty easy. Will it be that easy with the tiny case?



It is dinnertime and I am without food, so I will appologise if these arguments have been covered before or whether the arguments are infact utter nonsense :oops:
 
sas_simon said:
One thing that does trouble me, if all geforcefx's do have the fxflow fan and that they have somehow managed to squeeze it into a small form factor case, wouldn't that cause problems when you are cleaning the filters. In a standard case, removing the side of the case and removing the graphics card is pretty easy. Will it be that easy with the tiny case?

Hopefully the filter could be removed/cleaned without removing the video card or opening the case. In fact, why would you need to remove the card? The entire heat sink system is self-contained...you could never get behind the filter anyway. At least, not without taking the whole thing apart.
 
Howdy Simon!

1. the fxflow will be on the standard version of the geforcefx, no one but nvidia are 100% sure either way.

All people have in the line of *facts* are the photos of the board unveiled, from which all shown so far have the fxflow(tm) fan/cooling system shown.

"Assumptions" are generally decisions made devoid of actual facts. To "assume" a 9700 Pro WOULD have a gigantic, plastic cooling system bolted on the side that takes up two slots given images of it having no such device would be the same as deciding the GeForceFX does not have such a unit when all imagery clearly shows that it does.

For at least the specs, performance and release information, it's not what one would consider an "assumption" that the fxflow is involved as this is how these were obtained- with the equipment as shown.

2. the heatsink and fan will not be refined for the final release version, for all anyone here knows, that heatsink and fan could be a prototype and the actually real version will have some sort of external filter that can be removed and cleaned easily.

Many images of the GeForceFX already show a small, grey sponge-style filter on the intake vent on the card. It's hard to distinguish the actual material, but there IS some sort of filter shown if you scour photos and look closely at the intake port.

3. that the noise and dust will cause a serious problem to the final release version of the fxflow fan and that it will need to be cleaned every other week, as far as I can see, there is absolutely no proof either way to deny or confirm this argument.

Noise factor has already been noted by several sites, along with enough circumstantial evidence it IS loud in it's current form (i.e. comments NVIDIA is working to reduce the noise factor, as well as the throttling technology and the need therein.).

It is true though that nobody knows how frequently the unit and/or filter will need to be cleaned, nor could such a figure be arrived at earnestly. The same goes for any piece of consumer hardware- where homes with pets, smokers or near factories or other areas would likely need a higher frequency of maintenance versus owners with hardwood floors, filtered AC systems, no pets, and vacuum the area twice a day.

Chalnoth-
Hopefully the filter could be removed/cleaned without removing the video card or opening the case.

I think this is a safe "given" when looking at photos of the current unveiled product. Sponge filters on the back of the card can be fished out, and any computer enthusiast nerd worth the pocket protector they bear is going to have either a can of compressed air or a 2-direction mini-vac for cleaning. As the intake filter is shown to be lodged into the intake port on the BACK of the card, this would simply mean grabbing with a pair of needle-nose pliers to remove, and simply snapping back in once cleaned... no case removal or card dissection necessary.
 
I think demalion brought a good point in regards to Laptop cooling. I am not sure the flow rate of air is even remotely as close to what the GF FX card will do but still laptops are maintenance free for at least 6-12 months before a good cleaning is usually required. My laptop lasts about a year before I have temperture problems due to dust. Then again I don't keep my laptop on for long periods of time unlike my computer which sometimes runs 24/7.

Now if the GF FX does have a filter then the only problem would come from not cleaning it, the rest of the cooling surfaces should remain relatively clean and not really effect the overall heat transfer coefficient. The design displayed does allow anyone to easiliy install anykind of custom filter if needed (hmmmm, a crome car air filter might work :)).

As for the laminar flow consideratons, yes any decrease in air flow rate the more laminar flow you have. Only significance is that laminar flow increases the temperture the heat sink is exposed meaning the heat sink temperture will have to rise in order to transfer the same amount of heat. In addition the decrease air flow rate causes the average air through the cooling system to go up meaning the heat sink surface will be at a higher temperture in order to transfer again the same amount of heat.

I can put these all into equations if anyone is interested. A water block system would be the easiest to explain, test and get accurate heat outputs from any chip but still air cool systems for the most part with few exceptions behave the same transferring heat from a hotter source to a cooler heat sink.
 
How many hours per day and days per week on average does your laptop run?

How does that compare to someone running their computer 24/7 (which many do)?

What kind of surface is your laptop typically sitting on, and how close to the floor is it usually when running?

And as pointed out, how similar are the fluid velocities between your laptop and the GFFX?

Considering all of these questions, I think it is highly likely that the GFFX will need regular cleaning much more frequent than your laptop. ;)
 
Yea ... so is this thread really going on still ? Anyway if the chip is hot enough to cook an egg on it i wll never have to leave my room again :)
 
Yep, still going on.

... and with the phase change cooling solution needed to cool that egg frying chip, you could probably keep your beer cool as well.

Definitely no reason to leave the room!
 
Reading about the cooling solution they were using, did give me some considerations...things to look out for. Basically:

- I do run my computer 24/7. The internal temp is pretty good (42-44 on the CPU when under load if room temp is about 70 F due to the OCZ Gladiator and Arctic Silver II) and 28 C case temp. But that's due to good ventalation, not because I don't have hardware that puts out a bit of heat.

- To date, the heater has not needed to turn on, the computer puts out enough heat that my appartment stays relatively warm despite the outside temps.

During the night it isn't bad, but during the day it was been uncomfortable and I do have to open windows sometimes. My apartment hit 80 F when it was 63 F outside last Sunday. On such days, I need to open the window for awhile before going to bed so it isn't too warm for me to be able to sleep. However just leaving it open isn't an option either, as it gets cold enough at nights, I'd just wake up freezing. (Temperatures can very a lot from day to night down here in New Mexico...)

- All air flow is front to back, and the server ATX case is sitting on a computer desk, the very top (above the fans) sitting smack up against the hutch in the corner, and blowing all this air that heats up my apartment as such into that space.

- I do blow out my computer, but in the 2 years I have lived here, have never vacuumed behind my computer as well, I don't pull any air in from back there. Unless there is a reason to pick up the case and move it...this server size computer feels heavier to pick up and move then many pieces of furniture.
 
Sharkfood said:
I wonder what ATI could do by adding a dustbuster to their current chip...

How about creating the need for a 11 page thread at a given internet site........... :rolleyes: :eek: :LOL:
 
Fuz said:
martrox said:
How about creating the need for a 11 page thread at a given internet site........... :rolleyes: :eek: :LOL:

11 pages? Give it some time, I am sure it will reach 12. :D
Never in a million years.

Edit: Drat. I was hoping my post would roll over to the other page. :/
 
That's nearly as sad as my increased activity in posting over the last week being because I'm heading for 200 posts... :)

one more down :)
 
Status
Not open for further replies.
Back
Top