Nagao Industries has created an innovative new cooling system for motherboard VRMs, that works in almost every single chassis. Dubbed the N-VRM-FSTY60, the cooling system relies on twin 60mm fans mounted to a 120mm/140mm chassis fan bracket, which in turn gets installed onto the rear 120mm or 140mm chassis fan mount in your case next to the rear I/O panel.
These two fans blow at the rear portion of the VRMs. Thankfully, the mount works with the rear fan installed, so there's no need to remove the chassis fan. The idea is to keep the VRMs cool, even if you're using an AIO or tower cooler that otherwise wouldn't move much air across the VRM heatsinks.
A device like this might seem like overkill, as most power delivery systems on mid-range to high-end boards are overbuilt on purpose. They need to provide adequate power to Intel or AMD's most power hungry CPUs without causing the VRMs to overheat. But power requirements are going up for CPUs — Intel was already there with Alder Lake and Raptor Lake maintains the status quo, but AMD's Ryzen 7000 will be goosing power delivery to more or less equal Intel's socket power.
(Image credit: Nagao Industry Inc.)
With power creep becoming a legitimate concern on both Intel and AMD CPUs, a VRM cooling solution like this could become very useful. Alder Lake chips are already capable of running upwards of 240W or higher under specific workloads, which is extremely high for a central processing unit. Just a few years ago, you were lucky to hit 200W on a mainstream Intel chip (not HEDT) even with overclocking.
Thankfully, modern day motherboards won't explode or crash when the power delivery systems get hot, but once the system hits a certain temperature, the motherboard will throttle down the CPU to keep temperatures in check. This can be especially problematic on entry-level and some mid-range motherboards, where VRM power and cooling capabilities are sacrificed to keep manufacturing costs down.
Adding this funky fan setup to such a PC could seriously improve VRM performance, unlocking more performance at the same time. Since VRMs are traditionally passively cooled, direct VRM cooling can make a night and day difference to motherboard component temperatures.
Each fan measures 60 x 60 x 20mm, featuring hollow bearings (an improved variant of fluid dynamic bearings) and a maximum fan speed of 3000RPM. Despite that very high RPM speed, the fans are rated for just over 21 dBA, which seems impossible to believe. We don't know what RPM level the noise level specification was taken from, so take that 21 DBA with a grain of salt. Each fan runs on a 3-pin connector so they will work with almost every motherboard in existence.
Unfortunately pricing and availability are unknown at this time, and of course this would add the to cost of budget and mid-range boards. Still, $20 for higher performance and better cooling (if it's that cheap) might be a good idea.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Unfortunately I see this probably costing far more than anyone would be willing to pay. I expect at least a 40-50 dollar price tag just for the "novel idea"
I also see this causing issues with people thinking they can get away with cheap, poorly managed motherboards just because they added a fan.
I made something similar from an old I/O slot blanking plate. If you're handy with a drill it's not hard at all and looks just as "professional" as this.
Oddly, it's not necessary, and might actually work against the heatsink shown in the article picture. AMD's stock heatsinks for Ryzen are all down blowing, directing exhaust air to the rear to cool VRM...as well as to the front to cool warm memory.
Power delivery on high end boards is already overkill so I'm not sure how much of a benefit this really is. On a lower end board there's a chance that the cost setting up something like this you could just buy a better board with better power delivery. It's possible in an ITX situation that this could come in handy due to limited airflow, but that's about the only situation I could see it.
Lets go back to the Tom's article to see how much better it keeps the VRM's cool... oh wait there are no metrics. What is the point of this article since there is no proof it works and there is no link so it is not to get referral links.
Well, it doesn't claim to be a review, so more like just a "in point of interest" tidbit, which almost all tech sites do these days. I have no problem with it other than the fact that for boards it might actually help, you probably shouldn't be using a CPU at that end of the performance scale anyhow, and for boards intended for use with that high end of a CPU or meant for somewhat serious overclocking, they likely already have an enormous heatsink that doubles as part of the integrated I/O shield and basically won't see much benefit from additional airflow.
In most those cases they will also already be using a tower cooler that passes air over that heatsink but maybe some small benefit to those using an AIO.
I made something similar from an old I/O slot blanking plate. If you're handy with a drill it's not hard at all and looks just as "professional" as this.
Oddly, it's not necessary, and might actually work against the heatsink shown in the article picture. AMD's stock heatsinks for Ryzen are all down blowing, directing exhaust air to the rear to cool VRM...as well as to the front to cool warm memory.
Oddly, it's not necessary, and might actually work against the heatsink shown in the article picture.
Exactly what I'm saying. For older boards, for example, this MIGHT have had significantly more "usefulness" on one of the old AM3+ boards for overclocking or for FX-9370/9590 at stock settings, since there REALLY were no motherboards that could adequately support the 9590. Even the highest end boards had VRM throttling issues. But even for other older boards that either lacked heatsinks or simply had fairly small ones that only covered the VRMs, might have had a lot more usefulness.
For these modern ones, the heatsinks are often very large. Maybe for some boards that don't have the VRM heatsink integrated into the I/O shield's internal particulars it could still be useful IF you are having VRM issues. I suspect upcoming boards with very high TDP expectations will ALL have enormous heatsink integrated into the I/O though.
