1C over ambient?

[winewood]

Link from overclockers.com
"The Nebulous Cube"
Dino Velez ("Thundr7") and Kevin Heckeler ("Pinky") - 2/22/03

Dino has since upgraded his CPU, but at the time of the build and these photos he was running an AROIA Y 0213 XP1600+ (Palomino) at 1.9v and 1.92 GHz stable. Room temps were 21 C at the time of testing, CPU core temp (internal diode) was 22 C.

Talk about efficient. I searched our forums to see if anyone addressed this claim. I know what I think, but will defer to you guys.. any comments?

[Crosstrack16]

Read the review... sounds like an awesome concept. And for the price, i dont think there's a con, except maybe the size.

[hara]

1C above ambient can't be...with watercooling.

[bigben2k]

With 5 degC air temp, it might be possible, but it still sounds like a stretch. He reportsw CPU temps at 32 degC with that.

Winter cooling, you just can't beat the price!

[winewood]

The point I was looking out was a 1 degree difference from ambient with the current setup. The only difference is that the system has a bigger fan, bigger rad than may be possible in an "in case" setup. With these numbers, it beats any setup out there. I frankly think these numbers are beyond believable. I don't accept these numbers, not without a TEC or active cooling involved. Sorry.. doesn't happen. I'm suprised Joe published that on the site.
Nice box though! Sweet idea. Good construction. Nice resourcefullness on part scrapping. If you don't have case room, or want bigger accessories, this is the best idea for external mounting I have seen thats affordable. Those guys get a nod of approval overall.. if not for the temperature reports though.

[bigben2k]

Well, he's measuring a deltaT with his case's ambient temp, where his heatercore air temp is near or sub freezing.

The 1deg deltaT isn't measured correctly.

A water temp would have been nice.

[wymjym]

I run chilled water through a homemade water block. My ambient room temps vary from 70 (winter) to 83 (summer). My water temps vary from 18.5~21, depending on ambient air....cpu load.....and when during the measuring cycle I grab the data.
My cpu to water temps generally show the cpu operating at about 10 degrees above water temp. If I really pull the water down to 12 degrees, start the computer (so that everything else is nice a cool) and measure the temps I can see the cpu at only 5 degrees above water temp.
This is a long winded way of saying that 1 degree above ambient isn't very realistic.
My techcool unit is cooling off my pelts. Right at this moment I have 29.5 cpu, water 18.5 (chilled), water 34.7 (heated from pelts) going to techcool unit, water 32.3 (cooled) coming from techcool unit, ambient room temp 29.
wj
my chiller
my chiller's cooler

[jaydee]

Dino has since upgraded his CPU, but at the time of the build and these photos he was running an AROIA Y 0213 XP1600+ (Palomino) at 1.9v and 1.92 GHz stable. Room temps were 21 C at the time of testing, CPU core temp (internal diode) was 22 C.

Yes, you read right.

We verified this was NOT a fluke, bios issue, or diode problem (especially since we have tested many other CPUs in the past few months and have seen insanely cool temps with those as well).

:shrug: So it wasn't a concern that all the CPU's tested in the past few moths read insanely cool and some how adds to the argument those temps are right? And how was this "non" fluke varified as not a fluke and why wasn't that equipment used for the original testing being it must be better in order to test the mobo sensor accuratly?

[pHaestus]

Oh come on guys you don't still take issue with the temperatures you read on the web do you? I just shrug them all off and repeat after Bill:

ooohhhhhhhhhmmm

The exception is when the dubious claims are from those selling goods. That still gets me hot under the collar like picante from NYC

Otherwise tho just grin, bear it, and relax.

[myv65]

What's so hard to believe? I've got a '96 Explorer that'll do 0-60 in less than one second. Hey, I ought to know 'cause I've tested lots of other cars and was able to get all of them going 0-60 in one second.

Did I mention that I measured speed in inches per second?

As Bill would say, and borrowing a GIF from www.carbuyingtips.com

[pHaestus]

Well is it even so unreasonable that you can get your CPU to 23C using ~0-5C water? That seems about right to me (still a 15-20C delta T from CPU to water). I was making a more general observation about the reliability of temperatures that are reported

[myv65]

Could you get a CPU to ~23° with 0-5°C water? Sure. But why would these guys state their CPU was at 22°C with a room ambient of 21°C and then go on to say, "Yes, you read right."? They are clearly implying that they magically got a CPU temperature 1°C over ambient without a chiller. Oh, but what's this? The very next paragraph goes on to say they are now measuring 32°C with outside air ~5°C.

I believe the point everyone would make is two-fold:

1) Don't make unsubstantiated performance claims like: "To our mutual surprise, this outperformed anything we could have bought for nearly 50% more cost of construction." I say, "Show the proof, and I'll decide if it's really that good."

2) If you *do* claim outrageous performance, at least back it up with some measurements that have a known accuracy range.

[murray13]

C'mon guys do you really think that you can get a Tbred 1800+ OC'd to within 1 deg C of the water flowing in your system???

NOT!

What's the wattage of that proc'? Well over 50W? The best TIM available, .02 degC/W? Temp rise of just the TIM joint = 1 deg C! And thats ONLY the TIM joint and that proc' puts out more than 50W.

Think people.

1 deg C, must be a typo, a misread reading, something like that.

11 deg C, I could believe, and would say it's a nice system.

[pHaestus]

No read the article Murray. It is somewhat misleading in that the radiator is outside so the water is actually close to 0C. This means the CPU is only 1-2C over the room temperature air, but still 20C above the water temp (well probably less depending upon insulation quality).

[Alchemy]

It's a good idea for use of an outside-mounted cooling box - that is, the particular choice of plastic container.

There's nothing impressive at all about the system, which will be even clearer when summer comes around and the CPU is nearer to 50-60C. I can beat that with my uber-quiet HSF.

I like the "Several things at play that made this work," part, where the first and third reasons are given for this amazing performance. The "Yes, you read right" line is pretty classic as well.

It's all well and good to build these cooling boxes while high, but damn, come off it before you post the article about it.

Alchemy

[murray13]

pHaestus: True, my bad. It is very misleading, having a pic of the completed project sitting in the corner but using numbers with it outside.

Their line

The results are impressive and far exceeded our expectations.

says to me they really don't know what their talking about. A delta T of over 25C isn't really that impressive at all.