Extreme cooling phase change CPU cooler – Joe
SUMMARY: First look at Prometeia’s phase change case looks promising.
The kind folks at Chip-con were nice enough to send a sample of the Prometeia
100202. This is a phase-change (i.e., refrigerator) cooling solution complete with case (black) – a good one – an Enlight 7230.
I am not going to cut-and-paste all the technical details – read
Prometeia’s Technical Details and Explanation at your leisure.
The claims made for this unit are impressive:
“The Prometeia is optimized for maintaining below -40C/-40F in the range up to 80W, but capable of maintaining an impressive -28C at a CPU load of 150W at 20C ambient temperature…”
The “Micro Freezer” (cooling head) is basically a copper square at the end of the wand coming from the compressor. All the cooling power is concentrated at this point. This super-cold copper square sits on top of the CPU, cooling it down to sub-zero temps.
The surrounding material is designed to isolate the cold to this point to avoid any condensation problems due to below ambient temps.
The refrigeration unit sits in the base and is composed of the compressor, radiator, fans and control unit. The compressor is a Danfoss NF9FX; click on the link for detail specs. Danfoss has been around since 1933 and makes good stuff, so the unit should be a good one.
The long black hose is the base of the cooling wand. The wires on the lower left hand side are the 120 volt inputs (NOTE: DO NOT take these panels off unless you disconnect power – there are exposed connectors here. Prometeia does not advise removing these panels – I have removed them only to show what’s inside). A view from the other side
clearly shows two fans – one directly on the radiator and one venting the compressor compartment. The small black box at the upper left is the LCD and control unit, clearly visible in this shot of the lower front:
The front panel comes off in two sections. Taking the lower panel off is not difficult, but it is firmly mounted, so strong fingers are required. The upper panel removal is much easier – there’s a lip at the bottom you squeeze and it pops right off; all this is explained in detail in the instruction manual.
The case itself is a good mid-tower (Enlight 7230) – it features four CD bays and three floppy/hard drive bays, the latter in a removable cage. The CDs are mounted with clips that allow you to easily remove them. However, you must remove the front panels to take the side panels off – an inconvenience.
There are two mounting kits for CPUs, one for AMD and one for P4s:
They are basically the same except that the P4 kits uses the motherboard’s mounting holes and the AMD kit uses the socket’s lugs for mounting the clip. The parts include:
- Rear cover for the motherboard’s back
- Clip to which the cooling head mounts
- Upper bracket to seal the CPU from ambient air
- Shims to protect the CPU from misalignment
- Spring loaded bolts to hold the head to the Clip
- One tube of Arctic Alumina
- One long hex key to tighten the bolts
- One roll of sticky sealer to mount components
The sticky stuff looks like licorice and has a consistency similar to clay, but stickier and more rubbery. It’s more like caulking than anything else, but does not harden as some caulks do.
All these parts are put together as follows (from Chip-con’s website):
{mospagebreak}
The manual is quite good and the instructions are easy to follow; I took my time doing it and finished the whole thing, including mounting the motherboard in the case, in about two hours. There is no rocket science in this at all.
In my install, I did not mount the insulation on the back of the motherboard because I am using my Shuttle AK31 with the diode reader hack – I instead covered it with plastic wrap to keep condensation off the back. This IS NOT the way to go long term, but for me a short term move so I could measure CPU diode temps.
When installing the upper bracket, first check to see that all the motherboard components clear it. The base of the bracket has small feet that allow it to sit a bit above them, but some of these feet may hit a component directly. In this case, use a knife or razor blade to cut the offending foot off.
In my case, I had to cut off two feet – the pic below shows how one foot sat on an IC near the socket:
I cut it off and, after making sure the bracket was level, I then mounted the clip to the socket:
Note also there is a thin plastic shim surrounding the CPU – it has an adhesive back and is used (I assume) to prevent any CPU damage during the mounting process.
Now for the bracket. Once you’re satisfied that it’s level, then apply two layers of the licorice stuff to the bottom of this bracket like so (this is the side with the feet that contacts the motherboard):
Then you mount the bracket over the socket – it’s like a dam around it to prevent air infiltration. Before mounting the cooling head, you put one layer of the licorice stuff on its top to seal the cooling head to the bracket; it looks like this:
The cooling head then mounts to the clip and bracket with two spring loaded bolts:
You have to line up the holes to the posts on the clip and firmly press the head, keeping it level, to the lower bracket, squishing the licorice stuff in the process. This stuff is pliable and as you tighten the cooling head to the bracket, it should seal the socket from any air infiltration.
And you’re done!{mospagebreak}
To test the Prometeia, I used my Shuttle AK31, modified to read AMD’s CPU diode, as a test board, with an XP 1400 MHz as a test CPU. In addition, I tested it on the CPU Die Simulator. My test objectives were:
- See how far this CPU could be overclocked
- Determine the effectiveness of the moisture prevention system
- Determine the difference between Prometeia’s temp readings and CPU temps
Ambient conditions during these tests averaged about 23 C and 70% humidity.
When starting the Prometeia, it does take some time for the system to reach cold temps. The longer the system is off, the longer it will take before the PC boots up. There is a built-in delay by using the motherboard’s reset function. Until the cooling head’s temp reaches -33 C, the PC will not boot.
Once the cooling head hits -33 C, the reset function is released and the PC starts the boot process. Th following graph shows this process in action:
After sitting off for a day, the initial temp in the Prometeia display was about 22 C – ambient temp. It rose to 37.9 C and then started to decline. If temps go over 40 C, the Prometeia shuts down; you wait about 30 seconds and restart.
In this instance, temps peaked at 37.9 C and then started to decline. The software I have for reading diode temps does not go below 0 C, which is why you see the vertical red line at 0 C. From Power-On to 0 C took almost 4 minutes; to boot-up, about five minutes.
When you shut down, the cooling head will slowly return to ambient temps; if you boot up again within about two hours, the startup time will be substantially less – the cooling head retains its cool temps for some time.
It appears that in the power-off state, CPU and Prometeia temps are roughly equal. How this holds in the power-on state will be shown below.
I first booted up at spec settings to ensure that everything was OK. It was, and I then when into BIOS to see how far I could go. I was able to coax the CPU to 1740 MHz (12 x 145) at 1.85 volts. Increasing voltage beyond that had no discernible impact on better results.
The problem I had with this setting was that after a hard shutdown, the settings would not hold. I had to Clear CMOS and start over again – OK if you run 24/7, not good for daily use. I am not sure if this is due to holding the board in RESET until boot temps are reached or that the CPU is being pushed too far – I think the latter, as I was able to run at 1680 MHz (12 x 140), 1.70 volts, with hard shutdowns.
Note that I ran the CPU 0.05 volts under spec while overclocking 20%. There is no doubt in my mind that the Prometeia will allow very high overclocks with minimal voltage increases. For example, hitting 3 GHz with a P4 1.8A should be very achievable with this system.
With ambient conditions during these tests averaging about 23 C and 70% humidity, I ran Prime 95 at 1680 MHz, 1.70 volts, for 12 hours with the following temps recorded:
I DO NOT believe the MBM temps as this is from the in-socket thermistor. I could not get a diode reading, indicating that diode temps were below 0 C; more on this later.
The toughest condition to run a CPU cooled to well below ambient is at idle. With no stress on the CPU, it is radiating very little heat, cooling the CPU to well below ambient. I ran the system at idle for 24 hours with the following results:
I DO NOT believe the MBM temps.
The objective of this test was to determine if water vapor condensed inside the CPU chamber. During the test, I felt around the cooling head and motherboard to see if I could feel any moisture. I could feel, but not see, some slight moisture on the motherboard close to the socket base.{mospagebreak}
The cooling head itself has a heater strip which prevents moisture from collecting on it. However, the bracket that sits on the motherboard does NOT have one, and consequently I found this:
I circled in RED what I found – four or five drops of water had condensed after 24 hours. Under more humid conditions, there would probably be a bit more. However, I found no instance of water condensation such that it would drip on any peripherals or run down the motherboard.
As a reminder, this test was after running at idle for 24 hours, with the Prometeia Display reading -44 C and MBM CPU Temp reading -2 C. Running at idle at these temps is about the most extreme test possible.
Chip-con suggests the following remedies for this situation:
“Check that the legs of the Clips is not in direct contact with the bracket at either end, as that will generate a direct thermal path cooling down the bracket too much. This is the first and most obvious thing to check. If this is not the case at least one or more of the following measures should take care of the problem:
1: applying sealstring on the outside of the bracket and squeeze it flat to make contact with the sealstring above and below the bracket, on the area envolved.
2: applying selfadheesive closed Cellfoam sponges along the side where it feels too cold.
3: using MBM 4 or higher to perform a CPU warm-up when temperatures get too low in idle operation.
4: applying a considerable amount of heatsink compound on the rear cover to obtain thermal transfer of callories from the metal mounting plate onto the Rear cover to raise the temperature a few degrees.
5: Turn on the airconditioning in the room where you sit to reduce humidity. 😉”
Removing the cooling head does require some effort – the black sealing compound is very sticky and separating the cooling head from the base took some effort, although not unduly so. Once removed, I was pleased to find that there was NO moisture inside the CPU chamber.
This is quite a testament to the efficacy of the sealing system devised by Prometeia. I believe that if the directions are followed exactly and carefully, the probability of moisture condensation within the CPU chamber is very low. My hats off to them for a simple but elegant solution.
I found some moisture in a place I did not expect – the cooling stalk. The cooling tubes leading to the cooling head are encased in a foam sheath covered with what looks like a woven nylon cover. I squeezed the stalk and felt moisture on my fingertips – not saturated, but enough to indicate that some moisture had condensed on the foam covering.
In all instances where I found some moisture, I did not find enough to short circuit any components in the system. However, purchasers of this system should consider additional insulation around the socket, particularly when used in very high humidity environments. It might also be prudent to add some additional insulation to the back of the motherboard as well.
The temp probe is located at the top of the cooling head, where the temps are the lowest. By the time temps work back to the CPU, temps at the core are higher. I could not get a diode reading for reasons I indicated above, so I used the CPU Die Simulator to give some indication of the relationship between what’s displayed and CPU temps.
To do this, I jumped the power supply plug to power up the system with the cooling head attached to the CPU Die Simulator. I cranked it up to 80 watts and ran it until temps stabilized, finding:
This is a pretty good indication of how much gets to the CPU core – roughly a fall-off of a about 24 C. Note that this is a “true” 80 watts – a CPU typically radiates less than its full power potential under use. I think it’s safe to say that with a display temp of 40 C, CPU core temps are somewhere between -10 C to -20 C, depending on usage.
The compressor is virtually silent – noise from the system is due to the 120mm fans in the base. Add to that any fans used to cool the case. I did find it interesting that the base did not radiate much heat into the bottom of the case. The compressor does get warm – I measured some spots as high as 40 C – so venting of the bottom is quite good.
Overall, I would say the Prometeia is not silent but nowhere near as noisy as a case with a Delta 38 and 2 or 3 80mm case fans. I would place it between a good watercooled case, such as the Koolance or Swiftech Q-Power, and an aggressive air cooled case. The noise levels are not high pitched.
Overall, very tolerable considering performance.
{mospagebreak}
Buying something this expensive is not a trivial decision. In addition to raw performance results, buyers should consider the organization behind the product and its warranty. I bring this up because the consumer’s experience with the Maxxxpert product has been mixed at best.
As far as I can ascertain, Maxxxpert is not producing their phase change system at the present and distributors are not carrying it. Further, warranty claims are in limbo. Not a pretty picture.
I asked Chip-con about warranty, and they are offering a one year warranty on the system – a MAJOR step above the 30 day Maxxxpert warranty. Naturally this presumes Chip-con will be around to honor it.
Finally, I want to relate what happened as I tested the Prometeia.
The first system I received was defective – it shipped from Europe in a cardboard box with what I thought was limited protective packaging. I noticed an oily stain on the cardboard and thought this was just an overly liberal dose of protective oil. Turns out that there was a crack in one tube (happened during shipping) and the system was DOA (Chip-con has beefed up packaging as a result).
(An aside here: You can test the system before doing any motherboard installation by hooking up a 12 volt hard drive plug to allow the system to power up and jumping the ATX power supply to a power on state. Chip-con will include instructions in future manuals for this.)
Chip-con quickly shipped another system to test and I returned the first one. The second system tested OK, but the cooling head does not rotate at all, which means that it can only be used with a socket which is oriented in one direction only.
Emails to Chip-con on this resulted in a check of systems to be shipped, which resulted in a finding that a number of systems were, in fact, defective on this point. QC has been strengthened to eliminate this possibility.
I bring these points up not to discourage buyers, but to indicate that Chip-con is early on in the marketing cycle, and buyers should ensure that RMA and warranty policies are to their liking.
The Prometeia performs pretty much as advertised – it delivers sub-zero CPU cooling in a manner which should ensure trouble free operation, with the caveats indicated above. I have used a Kryotech for almost three years without a hitch, so I have no hesitation in asserting that well built compressor based systems should last a long time.
Prometeia is a young product, and early buyers take some “pioneer” risk. To their credit, Chip-con seems very responsive to changes to enhance the product.
I asked for comments from people who bought a Prometeia and received the following emails (note: I have edited some emails for English usage but not for content):
Mark Faulkner
Just read your review on the Prometeia and was impressed.
I have had one for about 2 weeks now, and it is currently running a
2.26b @ 2.822 Ghz. It runs 24/7 and I have been very please with the efficiency of the system.
I read your comments about moisture on the evaporator mountings and will
be checking mine quite thoroughly tonight.
I havent been able to achieve a serious overclock yet – best I have
managed is 3 Ghz and stability was dubious, but I don’t believe this is due to the cooling, as my evap never goes above -38 and CPU never above -10 (more often it is at -20).
Hedzer AKA Jirnsum, The Netherlands
I have been using the Prometeia system for about two months now, and I must say: I am impressed. I am using it on a TH7II RAID, with a P4 2.0A. Currently my memory is holding me back on overclocks (currently at 2.7 @ 1.65V), but hey, first gotta’ get some money again 😉
I ordered mine from www.oc-shop.com who sent a guy over to personally deliver it. Having read some other reviews and hearing the advice from the oc-shop guy, I took my time carefully reading the manual. This is VERY important. If you have done that, installing it is very easy.
One comment about the fact that the microfreezer is hard to remove: I found that you really have to let the ruberry stuff “thaw” out. Let the unit rest for about 2-2.5 hours, and removing the freezer gets much much easier.
One last comment on customer support: I found that the chip-con people reply within the hour during (Danish) business hours. Very, very good.
Here is a link [see next email] of a guy who did a make-over of his Prometeia case. Note the extra insulation on the tube, just like you said. I myself have once found a real small patch of ice on the tube, but not so much I worry about it.
Tom Versteeg
Hello, I`m using it now for a few weeks and I like it very much.
In about 2 or 3 weeks I`m going to order another one and build a Prometeia-secret-secret.
Matt (UK)
Having read your Prometia article, I thought I’d respond with some feedback on my unit. I’ve had it for a month now and have to say I’m very impressed. As you indicated, noise to performance ratio is off the scale. I did however remove the somewhat crappy (and sharp edged) case and stick a Coolermaster atcs710 on top. Aside from the case, not a single worry. It’s been 100% reliable from day one, and is run for about 8 hours a day.
I’m running a P4 2.26b, so some P4 feedback for you.
On air I could run 2.9 Ghz @ 1.72v with an MCX4000. As you can tell, a very poor chip as 2.26b P4s go. With Prometia I can manage 3.154 Ghz @ 1.74v (both actual readings, not Bios), so a substantial gain. It seems the P4 achieves lower temps than the XP at any given wattage, an example being my idle temps at the above speed and voltage being -17 C CPU (-44 C evap) and -9 C load (-31 C evap).
A point to mention would be I’ve replaced the standard 120mm fans with some Papst 120mm low noise fans. CFM is about 50 rated I beleive compared to 70ish for the Sunons, but they’re almost silent. It lost me around 1 C on evap temp and made no noticeable difference to CPU temps – a very worthwhile mod.
My overall conclusion, a very well made system with superb performance and a company with excellent customer service.
This is Eric Kronies, aka OPPAINTER from California
I’ve been running this unit for 2 weeks now. A very sweet system, I
use a Swiftech Quiet Power case for the top. I’ve run the system 48 hrs
straight under load and it seemed to have no effect on the compressor,
meaning after a rest it booted up and ran my idle and load temps on the evap
the same as before the small stress test. And two weeks later, it’s still
holding the same.
Condensation:
Now this could be a problem: My mobo was already conformal coated from
running peltiers and chilled water setups from before. But recently I’ve
been changeing a few CPUs and I’ve noticed a tad of condensation in the
socket area. I also have one of my Spring Screws for the Evap. starting to
show signs of rest.
I don’t know how Kryotech works for condensation proofing, but I can see
users without condensation proofing experience getting into problems down
the road with this unit.
I haven’t seen anyone mention dielectric grease for the socket pins. I make
sure I have the pins coated – there is a big time problem if the user doesn’t coat the CPU pins with this stuff. Eventually your CPU pins will rust out and weaken to the point they will be easily broken when handling the processor. I have seen this in my experience of sub-0 cooling.
That’s about it. I personally can handle the condensation myself. I love
this Unit and runs rock solid and smooth over the Maxxxpert. I’ll recommend
it anytime.
Pedro Amorim
I’m a Prometeia user from Portugal. You can check some benches and picture of my system on these links:
As you can see, I adapted the cooling unit to a Lian Li PC70. This thread also reveals the very good performance of the Prometeia on Intel P4 systems.
I think the CPU Kit for the Intel is a better design. I don’t have any condensation problem but I use a lot of seal string.
I have a friend that has a Prometeia with one AMD CPU – he had some condensation problems initially that he solved by applying more force on the CPU Microfreezer using some extra spacer rings (a dangerous practice with the fragile AMD XP core, no doubt!).
But, in my opinion, one of the major problems is the lack of flexibility of the tube that connects the cooling unit to the CPU. When I tried reassembling the CPU Kit, I found one stiff screw and a very serious problem – some liquid was flowing out the tube. When I examined it carefully, I saw that the insulation material of the tube was slashed!!
So I returned the unit back to Chip-con – it arrived there last Thursday (August 29); the support team said that they will ship me a replacement unit tomorrow or next Friday. I hope they honour they compromises!
My name is Tobias Johansson and I live in the little country of Sweden.
I have had my Prometeia system for a little more than a month now and
I’m quite happy with it, but I need to get me a new CPU to really test
the system because my current CPU didn’t give me any boost in speed with
the Prometeia system. I could post at higher speed but not run stably
but I’m quite sure I have hit the limits of the CPU (anyway I hope so
or my system must be defective, but I have no other indication of this
other than not being able to clock any higher).
I have also seen a bit
of condensation in the bend of the cooling pipe to the CPU, but I
talked to Chip-con about this and they promised to send me some
adhesive foam tape to help insulate that area a bit better. I can’t but
say good things about their support that I have talked to a couple of
times – always friendly and helpful.
I’m currently running a T-Bird 1000 @ 1428, 24*7, on an Epox 8K3A+ with a
256MB stick of Corsair XMS 3200 cs 2.5. I’m hoping to get me a new T-Bred 2400+ when they come out and hoping for a real good overclock with that 😉
Tom Holck – Denmark
I have had a Prometeia for some weeks now. Before I had 2 Vapochills, a good experience with OC. You can see my name on many OC top 10 /top 20 lists on both AMD and Intel systems. The Prometeia are better than the Vapochill – there are a few articles on that – and the Prometeia cost less. I have had no problems, only I have to set the reset temp from +40 to +50. Then the Prometeia will boot on first try.
I think your problems are from the CPU or the Board, not the Prometeia. I have tried different Abit boards and a few Asus. For serious OCing, I like Abit AT7-max or IT7-max.
Compared to Vapochill with the same system (P4 2.533, IT7, cosair400 c2) the Vapochill was able to do: 2533 @ 3201; the Prometeia can do 2533 @ 3448, 100% stable.
The Vapochill can not hold the temp down with full load when the voltage on AMD goes over 2.05V or the P4 goes over 1.92V. I found that the best results with Prometeia/P4.2533 was at 1.85V or less. The Vapochill was maxing out over 1.96V.
Right now my limiting factor to get the Prometeia-system faster is the DDR-ram. On FSB set at 175-185, I have to use the 1:1 and the overall performance drops compared to ram setting at 3/4. When I use 3/4 (lov) the FSB are at 171-173 and the DDR ram goes 228-234; I cannot go higher.
I meed some better Ram…. with a cooling device…
I’m the member “TASOS” from the Overclockers forums.
I’m an early user of Prometeia cooling system (bought in July).
I live in Athens Greece and I have “tested” the Prometeia in very tough weather conditions. Ambient temps during summertime 34C – 40C and high humidity.
I’m no expert in hardware issues and OCing – just an average+ I believe. I have OC 2 Intel P4 CPU so far with success. My post are 1000% stable and NOT simply the highest posts possible. I used Asus P4B533-E in both overclocks.
The 1st one a 1.8A SL63X that could do 2.52 Ghz air-cooled…went @ 2.84 Vc=1.65. The 2nd one(current) 2.26 SL683 could do 2.96 Ghz air-cooled…went @3.18 Vc=1.70.
**My rule is never to exceed 1.70Vc with these P4**
(My highest post with 2.26 was @ 3.33 Ghz)
The only problem I came up with was at mid August.
In one of the hottest days of summer, I noticed 3-4 small drops that came out of that black foam of the tube (at the point where the tube makes that curve to lead the cooling head at the socket). I solved that by covering the whole length of the tube with some paper tape (the one we use when we paint the house).
I had no problems till now…the system is working fine. No leaks at any point. If you ask me for a rating…I believe a 95/100 is fair enough.
Mike
Nice review and very close to my experiences with mine. Their packaging is very poor, mine was damaged and was only shipped to the UK.
James
Just a quick note on the Prometeia. I originally heard about Prometeia from
someone’s posting on the Overclockers.com forum. I did a little
research and ordered mine on the 30th of May. Chip-con has been extremely
helpful and I would have to say their service is excellent.
I use my PC at home mostly for gaming and occasionally for stock trading
when I happen to have a day off. So, while I OC my PC, I really can’t
afford crashes. I had my P4 2,4 and was using a Swiftech SuperDooper
cooler – that one with copper base and the aluminum poles sticking up. I
couldn’t stand the vacuum cleaner sounding Delta 60 fan, so I put on a 80 mm
Fan.
That was a little better, but I still had high CPU and system temps.
Since I’ve installed the Prometeia (very easy if you follow the
instructions exactly), I have only had to open my case once to change the
video card. I’m running it just at 3 Ghz and 1.65v, which is not the max,
but totally stable and the temps are great.
My CPU temp from MBM at idle is
-15° and moves to -5° under stress, but in anycase never gets to 0°
unless there are close to 40° ambient. I’ve never had a CPU temp over +3°
so far. My system temp moves from 10-15° (all celsius) which is quite good,
I think.
I’m using the Abit TH7-II. Oh – I’ve never encountered any
moisture, but I don’t open the case often enough to notice it if there was any. I can only tell you I’ve had no problems due to moisture (because I’ve had no problems at all). 😀
What a relief to be able to OC without crashes and whining fans. Before the
Prometeia, I was used to crashes; now I haven’t crashed once unless I was
trying to push the OC to the limit. I’ll admit it’s expensive, but I’d
definitely reccomend it to someone that can afford it.
I’m a completely satisfied customer.
I’m Hugo Sousa from Portugal
I bought a Prometeia System recently. I’m still testing it on my XP 1800+ and so far I hit 1850 @ 1.75V. Very nice. I’m sending you this mail not because of that, but, to tell you my story about buying this system.
I ordered it in the begining of August and it arrived at Portugal on the 20th day. I can tell you this because I used the tracking number to see where the package was all across Europe. Then, the wait began. On the 23th, I called General Parcel (the shipping conpany) here in Portugal and to my surprise on the other side, someone told me that they have lost my package.
I didn’t knew what to say. Anyway, next day I sent a mail to Chip-con and they said that they didn’t know what was happening but they will check it out. On the 25th August I received a call from Chip-con and they told me that the shipping company was checking all their warehouses to see if they could find my order.
On the 26th, 9.00 am, my order arrived – ufff, what a relief! At 2 pm Chip-con called me saying that they will send another Prometeia System ’cause the shipping company couldn’t find my package. I told them my order already arrived and that everything was OK.
All I have to say about Chip-con is good, because they even bothered to call to my mobile phone to try to resolve the problem. They were very nice. And it’s nice to see internet sellers worrying about the customers in this way.
Bruno Antunes
There’s not much to tell, since the case is simply great, although the case could be a little more beautiful.
Regarding to the temps, I have a P4, 2.26 Ghz, OC to 3240, and the temperatures are always below 0ºC, even when running prime for about 4 hours.
The worst thing about the cooling system is that is very tricky to get a good hermetic environment in the CPU. After having the case running about 3 weeks, when I diassembled the CPU KIT, there was an huge amount of water in the board. Another problem is that is a little tricky to have the freezer in a good contact with the CPU.
Overall so far, good reviews from those who have bought a Prometeia, with cautions on use that anyone using a sub-zero system should heed.
Be the first to comment