Gigabyte 6600GT Silent Pipe II

[Phlebas]

Haven't noticed any reviews of this passively cooled card yet (I may just be blind, of course ) so thought I would give my experiences.

I've just put together the following system:
Antec SLK3000B
Seasonic S12 430W PSU
Asus A8N-SLi Premium
Athlon 64 X2 3800+
Thermalright SI-120 with Nexus 120mm fan
G.Skill 2GB DDR ZX PC3200 (2x1GB) CAS2 Dual Channel Kit
Gigabyte GeForce 6600GT 256MB Silent Pipe II
200GB Samsung Spinpoint P120 SP2004C

I've replaced the stock rear exhaust TriCool fan with another Nexus (running at 7v, as is the heatsink one), suspended the hard drive and taped over the CPU intake holes on the case but other than that no alterations have been made.

The 6600GT doesn't come with the temperature sensor enabled so I edited the latest BIOS revision using NiBiTor to enable the sensor and reflashed it (exceedingly painless using the supplied Windows flash utility).

At idle, the 6600GT is running at 34C which is basically the case temperature (other idle readings I've got are CPU 35C, motherboard 38C - room temperature is ~22C). With the games I've tried I've not seen it go above 56C.

All in all, it looks like the Silent Pipe II works really well at cooling the card.

I haven't tried overclocking but I don't see why that shouldn't be possible in my system. With the addition of an intake fan (or some other slow-spinning fan pushing air at the card) I expect it could be overlcocked rather well.

Price wise, here in the UK it is the cheaper than the other 256MB 6600GT cards I've seen and even retails for less than the original 128MB Silent Pipe card! Don't ask me why, just take advantage of it

[JVM]

Your temps for that card are quite amazing in light of this review showing temperature range to 89C:

http://www.tomshardware.com/2005/12/09/ ... age42.html

[diver]

On the silent pipe II temperatures one should consider the possibility that Ton's Hardware tested the card on an open platform. The design of the silent pipe II is such that it depends on air flowing into (or out of) the case through vents in the rear of the case through a duct on the card. Several other sites reported this card had acceptable cooling.

Remember passive cooling is not usually completely passive, it just catches a ride on a fan that is cooling something else.

[JVM]

On the silent pipe II temperatures one should consider the possibility that Ton's Hardware tested the card on an open platform. The design of the silent pipe II is such that it depends on air flowing into (or out of) the case through vents in the rear of the case through a duct on the card. Several other sites reported this card had acceptable cooling.

Remember passive cooling is not usually completely passive, it just catches a ride on a fan that is cooling something else.

I've seen other sites say that, but they don't give any specifics like what the temps were. Tom's review does say, aside from giving the temperature, the card gets very hot.

I don't know what you mean by an open platform. There were quite a number of cards compared in the test, and the Gigabyte was hottest of all.

[frostedflakes]

What he's saying is that case airflow (or lack there-of) can have a huge impact on passive GPU temperature. The 6600GT SilentPipe II looks like it relies heavily on intake air from the back of the card for cooling. This assumes that your case has negative airflow (more air moving out than in -- creating negative pressure and pulling air into the case through any available holes, in this case the 6600GT's open grill). In a case with neutral airflow (about as much forced intake as forced exhaust), obviously the heatsink will receive little or no airflow. Same goes for an open-bench setup, where the components are not installed in a case. Essentially all you have is convection, which creates very, very little airflow. The reason the others would be cooler with an open setup is because they have forced airflow from a fan installed on the heatsink.

For example, the X800GTO in my sig is cooled by a passive Zalman heatsink. With the PCI slot cover under the card installed, my exhaust fans pull air in from the front of the case and right past the video card heatsink, and the GPU cooks. With the PCI slot cover removed, fans can now pull room-temp air into the case and right over the Zalman heatsink. Makes a huge difference in temps, like 20*C+ load.

[JVM]

Not quite clear to me. With a setup of rear exhaust and a front intake fan, are you saying that's bad or good for cooling the passive card? In this case, the Gigabyte card has an open area in the PCI slot. So, is the front intake fan drawing air in conflicting with the silent pipe trying to draw air from the rear?

Doesn't it depend on whether the heatsink fan is blowing on the h/s or blowing up?

[diver]

open platform=not in a case, out in the open

[JVM]

And here is a picture of the test setup at that review:

http://www.tomshardware.com/2005/12/09/ ... age36.html

[JVM]

open platform=not in a case, out in the open

If it was done out in the open, I would think the temps would be better than in the case

[MonsterMac]

so are the Gigabyte Silentpipe II cards OK now? I remember reading some threads a few months back about how they were unstable and people were getting freezes with them, etc. but it looks like they've changed the cooler since those posts?

[len509]

open platform=not in a case, out in the open

If it was done out in the open, I would think the temps would be better than in the case

Read frostedflakes reply again. Temps would not be better in the open. This card needs directed airflow to help cool the core. If the case is out in the open, airflow is all over the place and no longer directed.

Silent-Pipe II Technology

[JVM]

open platform=not in a case, out in the open

If it was done out in the open, I would think the temps would be better than in the case

Read frostedflakes reply again. Temps would not be better in the open. This card needs directed airflow to help cool the core. If the case is out in the open, airflow is all over the place and no longer directed.

Silent-Pipe II Technology

Sorry, but I still have to believe being outside the case would result in lower temperatures regardless of the technology.

[GHz]

I think the key words here are "standing air". You can't expect a passive heatsink to perform well in non-moving air. When you test a passive heatsink outside of a case, you most likely lose very important airflow that would otherwise be plentiful inside a case. Now if they had a box fan blowing onto the test rig, then it might be a more reasonable test. The point I'm trying to make is that it's not difficult for a passive heatsink to perform poorly on a test bench, but perform great inside a case... it all has to do with the heatsink taking advantage of available airflow... directed or not.

[NeilBlanchard]

Greetings,

Sorry, but I still have to believe being outside the case would result in lower temperatures regardless of the technology.

Why did the poster get so much lower temps than Tom's? Why do F1 cars overheat sitting on the start line?

If you read the Ninja or the HR-01 reveiws here at SPCR -- look for references to how mounting on a system sans case requires the use of a fan.

[Phlebas]

I decided to run a little test to see just what effect the card's air intake has.

Case setup: SLK3000B with 120mm Nexus exhaust fan @ 7V, Seasonic S12-430 PSU, no intake fan. I've now taped over all of the holes on the left side panel.

The GPU was idling at 33 C. I ran rthdribl for 10 minutes after which the GPU temperature had risen to 51. I removed the case's side panel and the GPU temperature immediately rose, topping out at 63 after 10 minutes. Replaced the side panel and the temperature dropped to 53 after a further 10 minutes. (NB: room temperature rose by about 0.7 C during the test)

CPU temperatures were 6 degrees C lower without the side panel but the GPU was 10 degrees hotter. Clearly the air being sucked through the intake is essential for the Silent Pipe II cooling to function well.

I'll try with an intake fan at some point in the future. It will be interesting to see what effect this has - there should be more overall airflow within the case but with much reduced negative pressure only a little air will come in through the Silent Pipe II cooler.

NB: One final point - since the air intake clearly does take in air it's going to be necessary to clean dust out of the system...

[JVM]

Greetings,

Sorry, but I still have to believe being outside the case would result in lower temperatures regardless of the technology.

Why did the poster get so much lower temps than Tom's? Why do F1 cars overheat sitting on the start line?

If you read the Ninja or the HR-01 reveiws here at SPCR -- look for references to how mounting on a system sans case requires the use of a fan.

You should re-read his post. He never gave any temps of the card, just temps for his CPU and MB, which was higher than the CPU. Tom's gave temps of the CARD.

[StarfishChris]

You should re-read his post. He never gave any temps of the card, just temps for his CPU and MB, which was higher than the CPU. Tom's gave temps of the CARD.

But you acknowledged it!

At idle, the 6600GT is running at 34C which is basically the case temperature (other idle readings I've got are CPU 35C, motherboard 38C - room temperature is ~22C). With the games I've tried I've not seen it go above 56C.

Your temps for that card are quite amazing in light of this review showing temperature range to 89C:

Why do heatsinks have fans? Because forced air cooling is far, far better than convection cooling, fact. A passive heatsink on an open test bench has only convection cooling. A passive heatsink in the path of an intake vent is subject to forced air cooling - the air is sucked in by negative pressure in the case (due to exhaust fans), if there's something in the way it'll have to go through it.

[JVM]

You are comparing to the tests done at Tom's and don't know under what conditions the tests were done. The other two reviews that say the card runs cool do not give any temperature reading, kind of a meaningless conclusion to me. The tests done at Tom's are shown here, and you can also go to his test setup page:

http://www.tomshardware.com/2005/12/09/ ... age42.html

[JVM]

And if you look at the test setup, the cards are completely outside the case. I will once again post the link to his setup:

http://www.tomshardware.com/2005/12/09/ ... age36.html

[StarfishChris]

You are comparing to the tests done at Tom's and don't know under what conditions the tests were done.

If you can't see that it's taking place on an open platform then you are not reading or you're a troll. EDIT: I assume you're talking about the other review sites, but since you haven't named them or linked they aren't particularly relevant. The OP has his computer in a case.

My final words on the matter is that you can only do apples-for-apples thermal tests if you use a open bench, where air moves in the same fashion. The THG review is meaningless in a real-world situation as airflow is completely different inside a case.

[Phlebas]

And if you look at the test setup, the cards are completely outside the case. I will once again post the link to his setup:

http://www.tomshardware.com/2005/12/09/ ... age36.html

But that's the problem and why he got such high temperatures. His setup will produce low temperatures for actively cooled cards but will not work well with passively cooled ones.

I guess you'll only accept this if you are ever in the position to try it out for yourself. I would have hoped that my test results would have sufficed but clearly not. Hopefully they have at least been of some assistance to others who were wondering about this card after seeing the reports of the original Silent Pipe cards running hot. Bottom line is that this card runs cool when mounted inside my case and should perform well in any case that has a negative pressure fan arrangement.

[JVM]

You are comparing to the tests done at Tom's and don't know under what conditions the tests were done.

If you can't see that it's taking place on an open platform then you are not reading or you're a troll. EDIT: I assume you're talking about the other review sites, but since you haven't named them or linked they aren't particularly relevant. The OP has his computer in a case.

My final words on the matter is that you can only do apples-for-apples thermal tests if you use a open bench, where air moves in the same fashion. The THG review is meaningless in a real-world situation as airflow is completely different inside a case.

I named the other review sites in another thread.

A troll? Me? I was just expressing my opinion based on what I believed to be true. If I'm wrong, so be it. Life goes on ...

[JVM]

And if you look at the test setup, the cards are completely outside the case. I will once again post the link to his setup:

http://www.tomshardware.com/2005/12/09/ ... age36.html

But that's the problem and why he got such high temperatures. His setup will produce low temperatures for actively cooled cards but will not work well with passively cooled ones.

I guess you'll only accept this if you are ever in the position to try it out for yourself. I would have hoped that my test results would have sufficed but clearly not. Hopefully they have at least been of some assistance to others who were wondering about this card after seeing the reports of the original Silent Pipe cards running hot. Bottom line is that this card runs cool when mounted inside my case and should perform well in any case that has a negative pressure fan arrangement.

I will probably give that card a shot because I don't have a better solution at this time

Are you saying that having a rear case fan and intake fan would be detrimental to the temperature of that card as opposed to just a rear case fan, besides the heatsink/fan and PSU fan?

[Phlebas]

Are you saying that having a rear case fan and intake fan would be detrimental to the temperature of that card as opposed to just a rear case fan, besides the heatsink/fan and PSU fan?

To be honest I don't know without trying. Assuming intake & exhaust fans are equal then there will still be some negative pressure in the case due to the PSU (maybe also because intakes aren't as open as exhausts). So you'll still get some cool air flowing directly into the card's heatsink. You'll also have more overall airflow inside the case. This may result in better, equal or worse cooling of the graphics card. Too many variables at play for me to be able to make a judgement.

The other thing to consider is the effect on the other components. An intake fan should help with general case temperature and the hard drives specifically if they are positioned by the fan. Also the northbridge if you've got a passive cooler on it. Keeping those cool may be more important than a few degrees on the graphics card.

If/when I decide to fit an intake fan I'll report on what difference it makes.

[NeilBlanchard]

Greetings:

Please re-read the original post yourself:

The 6600GT doesn't come with the temperature sensor enabled so I edited the latest BIOS revision using NiBiTor to enable the sensor and reflashed it (exceedingly painless using the supplied Windows flash utility).

At idle, the 6600GT is running at 34C which is basically the case temperature (other idle readings I've got are CPU 35C, motherboard 38C - room temperature is ~22C). With the games I've tried I've not seen it go above 56C.

Also:

I decided to run a little test to see just what effect the card's air intake has.

Case setup: SLK3000B with 120mm Nexus exhaust fan @ 7V, Seasonic S12-430 PSU, no intake fan. I've now taped over all of the holes on the left side panel.

The GPU was idling at 33 C. I ran rthdribl for 10 minutes after which the GPU temperature had risen to 51. I removed the case's side panel and the GPU temperature immediately rose, topping out at 63 after 10 minutes. Replaced the side panel and the temperature dropped to 53 after a further 10 minutes. (NB: room temperature rose by about 0.7 C during the test)

CPU temperatures were 6 degrees C lower without the side panel but the GPU was 10 degrees hotter. Clearly the air being sucked through the intake is essential for the Silent Pipe II cooling to function well.

I'll try with an intake fan at some point in the future. It will be interesting to see what effect this has - there should be more overall airflow within the case but with much reduced negative pressure only a little air will come in through the Silent Pipe II cooler.

I think that you stand corrected, JVM.

Let's keep it civil, or this thread will be locked; thank you.

[StarfishChris]

A troll? Me? I was just expressing my opinion based on what I believed to be true. If I'm wrong, so be it. Life goes on ...

I misunderstood you, but I thought the other posts were clear enough

Are you saying that having a rear case fan and intake fan would be detrimental to the temperature of that card as opposed to just a rear case fan, besides the heatsink/fan and PSU fan?

To simplify things, intake fans usually force X amount of air (depending on fan speed, size, P/Q curve etc. - calculating would be difficult) from outside and directs it at anything in front i.e. hard drives.
The exhaust fans blow Y amount of air out of the case. If Y > X then the negative pressure that results sucks air into the case wherever possible, taking the least restrictive routes where possible.
So I would expect that if your intake fan doesn't run too fast, air will come through the card's vent whatever.

In short: an intake fan per se is not detrimental, though you may find it beneficial to adjust fan speeds (intake or exhaust).

[JVM]

Greetings:

Please re-read the original post yourself:

The 6600GT doesn't come with the temperature sensor enabled so I edited the latest BIOS revision using NiBiTor to enable the sensor and reflashed it (exceedingly painless using the supplied Windows flash utility).

At idle, the 6600GT is running at 34C which is basically the case temperature (other idle readings I've got are CPU 35C, motherboard 38C - room temperature is ~22C). With the games I've tried I've not seen it go above 56C.

Also:

I decided to run a little test to see just what effect the card's air intake has.

Case setup: SLK3000B with 120mm Nexus exhaust fan @ 7V, Seasonic S12-430 PSU, no intake fan. I've now taped over all of the holes on the left side panel.

The GPU was idling at 33 C. I ran rthdribl for 10 minutes after which the GPU temperature had risen to 51. I removed the case's side panel and the GPU temperature immediately rose, topping out at 63 after 10 minutes. Replaced the side panel and the temperature dropped to 53 after a further 10 minutes. (NB: room temperature rose by about 0.7 C during the test)

CPU temperatures were 6 degrees C lower without the side panel but the GPU was 10 degrees hotter. Clearly the air being sucked through the intake is essential for the Silent Pipe II cooling to function well.

I'll try with an intake fan at some point in the future. It will be interesting to see what effect this has - there should be more overall airflow within the case but with much reduced negative pressure only a little air will come in through the Silent Pipe II cooler.

I think that you stand corrected, JVM.

Let's keep it civil, or this thread will be locked; thank you.

The second quote was after the reply in question, and is not the original post.

Ah, but the FIRST quote is from the FIRST post -- and he very clearly DOES mention the video temp. The second quote very clearly and directly refutes your oft repeated argument -- and yet again you ignore the main points, and blithely make your erroneous statements AGAIN.
NB

Civil? I am not aware of being non-civil. Can you point me to my post showing non-civility?

It not just you, but to repeat my above point -- your continued refusal to listen/acknowledge/learn from everyone's very patient explanations might just get up someone's nose. Frustration has already been made equally clear, and I hope you take the point!

You need to become more aware.

NB

[JVM]

A troll? Me? I was just expressing my opinion based on what I believed to be true. If I'm wrong, so be it. Life goes on ...

I misunderstood you, but I thought the other posts were clear enough

Are you saying that having a rear case fan and intake fan would be detrimental to the temperature of that card as opposed to just a rear case fan, besides the heatsink/fan and PSU fan?

To simplify things, intake fans usually force X amount of air (depending on fan speed, size, P/Q curve etc. - calculating would be difficult) from outside and directs it at anything in front i.e. hard drives.
The exhaust fans blow Y amount of air out of the case. If Y > X then the negative pressure that results sucks air into the case wherever possible, taking the least restrictive routes where possible.
So I would expect that if your intake fan doesn't run too fast, air will come through the card's vent whatever.

In short: an intake fan per se is not detrimental, though you may find it beneficial to adjust fan speeds (intake or exhaust).

Thanks for the explanation, but I'm going to solve this dilemma by going fanless without a video card

[mattthemuppet]

I can hear alot of heads hitting desks here

I'd give up (I've tried explaining this to JVM in another thread) but I'll try one last time, even if I'm parroting what's already been said..

JVM - think very carefully and hard about this.

Scenario 1 - If you have fans (exhaust/ PCU) sucking out more air from the case than is being blown in by any intake fans, you will have a negative pressure inside the case. To equalise the pressure inside the case with that outside the case, air will try and enter the case by any route possible (side vents, gaps in the case, through the optical drives). Think about what happens when you open a vacuum packed container of coffee - air rushes in.

Now, the Silentpipe2 cards have a vent at the back which takes up a PCI slot. This vent leads to a heatsink. When there is -ve pressure in the case (remember, opening a packet of coffee), air will be sucked in through this vent and over the heatsink. Therefore the heatsink will be cooled down and the GPU temperature will drop (or be low). All clear?

Scenario 2 - If there is a intake fan in the case that is pushing more air into the case than is being sucked out by the exhaust/ PSU fan, you will have the positive case pressure, as more air is being forced in than is being taken out. Therefore, air will try to escape through whatever holes it can (think of a balloon - high pressure inside will escape if you open the blowhole).

In the case of the Silentpipe2 cards (remember, that have a vent leading to the outside, with a heatsink inside the vent), air will be forced out of this vent and over the heatsink, therefore cooling the card. However, the air being forced out will have already been heated to a degree by the components inside the case, therefore the degree of cooling will not be as great as if there is air being sucked in (due to -ve pressure). Does that make sense.

Scenario 3 - air coming into the case from intake fans = air being taken out of the case by exhaust/PSU fans. Here, the case pressure will be the same as the outside air pressure, so no air will be sucked into or blown out of the case other than that by the fans. In this situation, no air will pass over the Silentpipe2 vent and heatsink (as none is being sucked into or out of the case, other than that by the fans), so the heatsink will not be cooled and the GPU temperature will rise.

The test setup in the Toms Hardware review is essentially the same as Scenario 3 - there is NO air being sucked over the heatsink of the Silentpipe2 card therefore the GPU temps will be higher than they would be if the card was installed inside a case with either scenario 1 (-ve case pressure) or scenario 2 (+ve case pressure). Essentially, the reviewers have not paid any attention to how the card is supposed to be used (closed case with -ve or +ve pressure) and have then complained about the temps being too high.

In short - the Silentpipe2 design is a very intelligent one which needs to be used in the way intended and not by stupid reviewers.

Does that make sense?

[diver]

I have a bump on my head and a dent in my desk.