Discussions about Silent Computing
https://silentpcreview.com/forums/
https://silentpcreview.com/forums/viewtopic.php?f=13&t=68099
Some more modding fun to consider. It's just a molex connector for the pump. Why not use a fanmate II or modding the pinout to see how low it can go/provide enough gpm for cooling?The pump is not controllable but even at full speed, it's one of the least offensive sounding models I've encountered, and the included fan has silky smooth acoustics.
On second thought - this may not be a big deal if the user is planning ahead rather than doing a retrofit. Buy the cheapest reference design board (with a warranty that covers modding) and have at it.CA_Steve wrote:Nice review. Too bad the cooler isn't more compatible with the variety of gfx cards.
I see. Wouldn't having a much larger area of contact on the chip's surface be better than getting it from the pins? ie. I've seen a few thermal images of GPUs and the whole chip is highlighted in red no just the pins.CA_Steve wrote:A significant portion of the heat in an IC package is conducted away from the IC through it's bond pads to the connected pins to the board/copper traces/power and ground planes. Increase the copper on the board (like Gigabyte did a while back) and lower the IC temps. Add thermal pads and a big honking heatsink on the backside and lower temps.
The cooler directs air downwards towards the chips though albeit a bit warm from the heatsink. With this setup there's no air being blown on them at all.nagi wrote:I have their air-only cooler on my R9 290X, and it also leaves the VRMs without any additional cooling. There is however a way to deal with this: take the original heatsink and a metal saw. That lowered my VRM temps to ~70 on load. I'm not sure about the "they can take a lot more" comment, do you have the specs for the ICs?
Would that one fan over the GPU direct any sort of air towards the VRMs? And wouldn't those benefit more if they actually had a proper heatsink on them? Even if there is no air blown directly onto them, attaching a large heatsink would improve their cooling at least a bit.fufufu wrote:The cooler directs air downwards towards the chips though albeit a bit warm from the heatsink. With this setup there's no air being blown on them at all.nagi wrote:I have their air-only cooler on my R9 290X, and it also leaves the VRMs without any additional cooling. There is however a way to deal with this: take the original heatsink and a metal saw. That lowered my VRM temps to ~70 on load. I'm not sure about the "they can take a lot more" comment, do you have the specs for the ICs?
But it's not the chip's surface...it's the plastic package that encapsulates the chip. Odds are the metal infrastructure of the bonding pad/bonding pins/package pins/motherboard pad is a better thermal conductor than the plastic of the package. Depends on many factors, like size of the IC, number of pins, type of package, etc. In the case of the VRMs, they might even have a metal slug underneath that connects to a mobo pad that goes to the ground plane to facilitate heat transfer. Think of the old style 3 pin voltage regulators. Bend that baby over and bolt it down to the ground plane.fufufu wrote:I see. Wouldn't having a much larger area of contact on the chip's surface be better than getting it from the pins? ie. I've seen a few thermal images of GPUs and the whole chip is highlighted in red no just the pins.
I guess that makes sense.CA_Steve wrote:But it's not the chip's surface...it's the plastic package that encapsulates the chip. Odds are the metal infrastructure of the bonding pad/bonding pins/package pins/motherboard pad is a better thermal conductor than the plastic of the package. Depends on many factors, like size of the IC, number of pins, type of package, etc. In the case of the VRMs, they might even have a metal slug underneath that connects to a mobo pad that goes to the ground plane to facilitate heat transfer. Think of the old style 3 pin voltage regulators. Bend that baby over and bolt it down to the ground plane.fufufu wrote:I see. Wouldn't having a much larger area of contact on the chip's surface be better than getting it from the pins? ie. I've seen a few thermal images of GPUs and the whole chip is highlighted in red no just the pins.
