I'll second this, as they did it for me for my system a few years back.
I spoke to Noctua after the review, who provided this information.
If that's the expectation, then open bench testing for at least one of the CPUs would've been nice to include. I'd have rather seen that, than a second Intel Raptor Lake.
I'm well aware of the differences, believe me!
I believe stock is referring to the 9950X3D power consumption in Cinebench MT without PBO which is something basically every tower cooler with decent mounting should be capable of.
Vertical without fan is missing from that graph.
I'm not really clear on your point, here.
I once built a machine with all Noctua fans - the best available at the time. The motherboard didn't have a PWM header for the front intake, so that 140 mm fan ran at full speed. The result, at idle, was a lot of whooshing noise, with a dull hum underneath. Don't estimate how much noise air can make, especially when it hits obstructions.
Thanks for that very informative and thoughtful post. I'm not sure if it answered the premise, or maybe I just got lost along the way.
I aim for "unnoticeable" rather than silent, because with birds singing outside, and airplanes, trains and a motorway not too far off, a completely silent PC is of little benefit. A tinnitus doesn't help either...
I certainly agree, but NAND is very easy to cool. The SSD's controller is what produces the vast majority of the heat on a SSD.
Yeah, but you're not arguing that the NH-P1 isn't massive overkill for a SSD, right? That was certainly my point.
To nit pick, that's 39% of testing configs (12 out of 31) above the "PCIe 5 Throttling" line, by my count.As an example, here's the results of my NVMe heatsink test featuring a drive powered by Phison's notoriously hot E26 controller. About half of the heatsinks will encounter throttling if running PCI-e 5 and an actually stressful load.
Yeah that already exists, tablets, phones and SBCs are a good example of passively cooled systems and although they are low power due to the small size it could be scaled up to normal PC component size.
Indeed. While I understand that some users will be interested in upgrading their performance by using 3rd party pads (SSDs) or pastes (CPU coolers), I always test the product with the thermal interface provided by the manufacturer.
Even though I have plenty of different TIM available to me and can use whatever I want I think the way you do testing is a lot more useful to the average user even if it ends up having more variables: in a case with everything the cooler comes with.
To be clear, I was just contrasting CPU coolers with SSD coolers. In many cases, it wouldn't be practical to use CPU TIM with a SSD cooler, due to the need to ensure electrical isolation and potentially accommodate surface-mount components poking above some of the chips.
I've dabbled a bit with passive cooling and all I can say is that, even with a great heatsink, a little air movement can make a very big difference. If you wanted to completely get rid of case fans, then I think you'd at least need a case with wide open vents on the bottom and top, so a good convection current could get going.
As I think others have said, there are fanless PSUs at least up to about 450 W (and I think I've even seen higher). That's more than enough for a passively-cooled system.
Not a chance! It's rated at 130 W!
The quietest fans are inaudible at about 30% of their rated RPMs. Some fan models will even stop completely, if the PWM duty cycle drops low enough. The safest thing to do is run it in such an inaudible configuration, but with fan curves that will ramp up. If you really don't want to hear the fans, then set power limits low enough that the CPU won't heat up enough for them ever to ramp to an audible speed.
Okay, I found this:
Money, you just need lots of money to throw at it then you don't need fans!
Wow, thanks for that! It does seem rather impressive, for 2006.
Noctua makes some 200 mm fans, which is pretty close to what you described. I have no idea where they're used, but there must be a market for them.
Yeah, it's very hard to find a windowless case, these days, in the mid-range or high-end market. You have to get into some specialty territory, before you'll encounter them.
Noctua has the NH-L12x77, which I'm using in one chassis, and the NH-C14S, which I'm using in another. On the cheaper end of the spectrum, Thermalright has the SI-100, which I've yet to try, but I snapped one up for the low price of ~$25. In one of the few tests I saw of it, it seemed to outperform the NH-L12x77, but they used the low-profile fan that came with the Noctua cooler, whereas I always replace the fan with a regular-thickness. I think you could even use an oversized model, like the NF-A14x25r G2 PWM, which claims to have 120 mm-compatible mounting holes.
Oh, don't go passive, in such a box. A lot of that stuff needs some airflow.
The winning formula for dust is a combination of positive-pressure and dust filters on all your intakes. This keeps my case interiors noticeably cleaner, even though the dust filters I use are just the nylon variety. If it can be hidden away neatly, I might experiment with HVAC-grade filter material, for a machine I leave on 24/7.
A trend in cases seems to be having open tops, to facilitate 360mm radiators. Many such cases also have places for bottom-mounted fans. I wonder how you'd make out with a case enabling you to use 2-3 bottom-mounted 140 mm intake fans. If you can put something under the case feet, maybe placing a bit of "egg crate" foam underneath could help control the fan noise.
On the kW machines open top makes a lot of sense with heat rising and the fact that most of the backside surface area is already spoken for: unless you have a really tall tower, there is no other place for exhaust fans. Since my machines are under a table, there isn't a lot of dust settling from the top while they are turned off while the fat desk woodtop deflects any direct noise emissions.
My bottom-mounted cases are older Lian Li models, where the gap between the PSU and the bottom is large enough that the PSU basically doesn't pull any air through that vent. I've never seen those dust filters dirty, but then I've also never maxed the PSUs' output and so their fans wouldn't be running at max RPMs.
Yeah, I definitely like that only one of my machines has rotating media.
Facebook
Twitter
Instagram