Concept Arctic: 5L RGB Console-Sized Liquid-Cooled Case

[Bryce]

Introducing the B-Tech Arctic computer case! At only 5.3 liters of internal volume, it's an utterly tiny ITX case with full-size GPU support! The Arctic case is designed for an accurate "console aesthetic" look with it's glossy plastic exterior construction, ample vent holes, and clean, screw-free exterior, including on the back and bottom!

But just because your computer looks like a console doesn't mean it should perform like one! That's why Arctic supports full-size GPUs and comes standard with built-in liquid cooling! It even includes a built-in 1200W power supply!

But of course, the Artic isn't called "Arctic" for nothing. It's also got built-in sub-zero cooling capable of handling over 500W of total system heat dissipation, meaning your system could be cooled to 0° C FREEZING even at LOAD!


For progress updates and to be notified when our campaign goes live!

Anyway, let's get into it! Here's a 3D Render of Arctic Case in Glacier White.
The top intake is a custom designed ~100mm RGB fan. The final design will also have a backlit logo on top.

The case can be opened quickly and easily thanks to the innovative magnetic-latch system, which sends power and data for the LED illumination through magnetized contacts, making the case side-panels totally wire-free.

Arctic in Cobalt Black:

From the side, you can see the 360° addressable RGB LED ring.
A non-slip rubber pad with an intricate pattern lines the full bottom of the case to give it a two-tone look.

From the front you can see 3 of the 8 individually controllable 4-pin exhaust fans.

A/C units connected to the watercooling loop with a custom copper heatsink block are positioned behind these to achieve CPU/GPU temps BELOW ambient!

And did I mention there's a flipping TURBO BUTTON!?

You'll also notice there's no power button. That's because it's capacitive!
To turn the system on, simply tap the left-front area of the case.

Of course, you're probably also wondering how on earth we're controlling 10 independent 4-pin fans (including the pump) and over 70 zones of RGB lighting, especially in such a tight space...

Well, existing 4-pin fan splitters are not able to control fan speed independently or even report individual fan speed, leaving you with no way to ensure your fans/pumps are still working properly, not to mention they're way to big to fit in a case this small, so we decided to come up with our own solution. That's why we created the Arctic Control Board! Able to control 12 independent 4, or even 3-pin fans with per-fan auto-config and per-fan RPM control. Plus, it has RGB lighting control, capacitive touch circuitry, A/C cooling control, case-open detect, and even front-panel IO!

And as for that Front Panel IO:

• 1x USB 2.0, 2x USB 3.0
• 2x Reassignable HD Audio Jacks
• RGB Illuminated Capacitive Power Button w/ Adjustable Sensitivity
• User-Configurable Turbo Button (Mechanical MX Blue-like Switch, because why not)

What's Included with the Case:

• The case. Obviously.
• Custom 1200W PSU w/ Dual (also custom) 600W AC Power Bricks.
• Pre-assembled copper CPU & GPU water loop.
• Universal CPU waterblock.
• Dual thermoelectric A/C radiator units.
• 8x 60mm side fans w/ magnetic dust filters, 1x slim ~100mm RGB intake fan.
• 64 pixel 360° addressable LED ring.
• Dual-channel smart RGB controller & 12-channel 3/4-pin PWM driver with per-fan RPM feedback.
• (W.I.P.) Optional 2-way SLI adapter.

The case will be available stand-alone (case, PSU, CPU+GPU loop), or in VR-optimized pre-built configurations featuring overclocked, pre-delided 7th gen Intel CPUs and waterblock-equipped NVIDIA 10-series GPUs.

Currently, this project is somewhere between the "concept" and "prototype" phases. Design concepts have been tested, circuits have been prototyped, but a full case prototype has not yet been completed. Final kickstarter prototypes should be ready sometime in December, but until then...

For progress updates and to be notified when our campaign goes live!
​

And of course... If you have any questions, feel free to ask them here!

See Also:

 

[ondert]

wow.. things getting speed these days here.

 

[Jonny727272]

This seems like a big April's fools joke, or maybe just total BS being spewed by a corporate business. First of all, custom mod makes a smaller case that can fit a full size gpu, so you're wrong there. Also, it's hard to still call your case 5L when you have two power bricks and the whole radiator system outside of it. I'll ignore all the technical difficulties about the RGB lights, multiple fans, universal waterblocks, and no screw system for now and instead ask how much thing thing will cost. I know we are a niche group and willing to pay a little extra, but this seems crazy. I mean, a turbo button with a mechanical switch? That's just excessive. I will wish you good luck in your process and trying to make a prototype.

 

[Bryce]

I was actually waiting to see how long it would take for someone to find a smaller case (that's why I put 5L in the title for this forum specifically). I couldn't find any, but I knew one surely must be out there. It is a good thing though, because that shows that people here know their SFF cases!

Anyway, I've seen plenty of cases and pre-builts (especially in the sub-8L range) with external power bricks, so I hardly think that counts. And you certainly won't be doing dual SLI watercooling, or even single-GPU watercooling, for that matter, with an off-the-shelf Pico-PSU.

And as for the cooling, that's actually all inside the case. Only the power bricks, which aren't even that large or heavy, are external. We've already done extensive part-fitting using laser-cut cardboard, formboard, etc, and the parts are physically fitting inside, so it's defiantly not BS. Like I said, the project is sort of between the "concept" and "prototype" phases.

As for cost, it's obviously not going to be the cheapest thing out there, but I've seen plenty of >$200 cases out there, especially croudfunded ones, and considering the fact that your CPU and GPU cooling as well as PSU is already covered, I'd say our $320 target price is pretty reasonable. The pre-built configurations are also going to be somewhat competitively priced, especially considering the free CPU deliding and watercooled GPU retrofit we're providing.

 

[Necere]

Sorry, but this is pure fantasy land. You simply don't have the volume to cool hundreds of watts of hardware and TECs, period. Even if it somehow worked, it would sound like a very loud and angry beehive with all of those 60mm fans running fast enough to push the air it would need.

 

[Bryce]

Actually, yes, it does get very loud. That's why the per-fan PWM control is essential. The system backs off once you hit your target temperature. And I doubt it's possible to fit 800W (BTW that figure is not 100% honest. That's how much power the cooling can draw, how much it cools is slightly less) worth of computer hardware into an ITX build anyway. The point of the cooling capacity is that because of the overhead, the components can be kept below 0 degrees despite their continuous heat output. (Yes we are taking measures to prevent condensation.) The loop is also surrounded with thermal insulation, which will help trap the cold in (well, the heat out, technically).

Sure, I'll admit, I'm not an expert on computer parts, but what I AM is an engineer, I know other engineers, and we know the laws of thermodynamics, and sure, I may not own a GTX 1080 Ti, or even an LGA1151 CPU, but I don't see how the laws of physics could apply to it any differently than the sad low-profile Radeon HD cards we've tested so far.

And yes, that much thermoelectric cooling CAN fit into a 5L case, along with an ITX mobo, a GPU, and our prototype 1200W power board, and if an ITX system with 1080 Ti and a 7700K in it only draws ~400W from the wall, it can't be dissipating more than 400W of heat.

Because of the nature of peltier-effect refrigeration, large copper blocks with heat pipes work far better than a typical radiator, and we can make these FAR smaller than any air-cooled radiator would have to be. We've also already built the PWM control board for driving the thermoelectric devices, which by the way requires some MASSIVE transistors.

The thing about cooling with refrigeration is, you're trading power-efficiency for maximum heat-dissipation. You're taking something that the surrounding air is happy to do for free at it's own pace, and forcing it to happen faster because that's not good enough. That means you can dissipate orders of magnitude more heat in a smaller space, provided you put enough energy into the equation.

 

[Necere]

The point of the cooling capacity is that because of the overhead, the components can be kept below 0 degrees despite their continuous heat output.

Under load? <0 degrees at idle only would be pointless.

Because of the nature of peltier-effect refrigeration, large copper blocks with heat pipes work far better than a typical radiator, and we can make these FAR smaller than any air-cooled radiator would have to be.

Ultimately you still need to move that heat to the air, which means surface area. I get that you can increase the rate of thermal transfer substantially by increasing the hot side temp relative to ambient, but you're also adding a whole lot of extra heat with the TECs that gets added to the total heat that needs to be moved to the air. It's hard to come out ahead (read: lower component temps) in this scenario without a lot of surface area (and/or airflow) on the hot side. With such a small volume, the needed airflow will be very loud, and the exhaust stream is going to be unpleasantly hot. Unless the heatsink/exhaust is very well insulated, some of that heat is going to transfer to the chassis, potentially making it scorchingly hot in places.

You might be able to pull it off - I'm happy to be proven wrong - but with how hot and loud it's likely to be, it won't be pleasant to actually use.

 

[ignsvn]

Let's just wait for prototypes with temperature, noise & power usage test data. Also perhaps price projection.

 

[jØrd]

Given the potentially dubious nature of this post I went ahead and stripped out the external links.
However, on the off chance this actually goes anywhere & isnt just an email harvesting scam ive gone ahead and left the thread open.

 

[Bryce]

"Email harvesting scam"? Ouch, that's a new one. (Are those even a thing in 2017?)
Is there any reason to believe that anyone else's project on this forum right now isn't a scam of some kind?
After all, the vast majority of kickstarters are! That's the risk of croudfunding! (Exciting, isn't it?)

EDIT: So since you people are so abnormally serious about the most minute details around here, I've rounded up to 5.3L instead of down to 5.2, adjusted the 800W to ~500W to more accurately reflect the real-world heat dissipation you'll see from the 800W cooling system, and changed will see to could see below 0°, because you can't really be certain about anything despite what the math says. Happy?

 

[Necere]

For what it's worth I don't think this is a scam. I'm just extremely skeptical that you'll be able to deliver anything close to the claims you're making. Extraordinary claims require extraordinary evidence, and so far you've not shown any - just some CAD screenshots and dubious claims about its capabilities. It's one thing to drop in with this as your first post, but to tell people you're collecting emails for the Kickstarter you're planning to launch next month - next month - is what puts it over the line for me. This is a guaranteed path to a failed Kickstarter. All the successful crowdfunded projects - ours, dondan's, the Dr. Zaber guys - have taken months or even years to refine and test the designs to make sure we had a manufacturable product before ever asking anyone for money, and yours is much, much more ambitious.

 

[jØrd]

Is there any reason to believe that anyone else's project on this forum right now isn't a scam of some kind?

Most of them aren't making borderline physics defying claims about their projects, the ones who are running a mailing list tend to have long threads detailing their design process and prototypes to show for it and generally speaking, a reputation to back them. Now to be clear, I wish you the best of luck w/ your project but I feel justified in the position I took. From here you essentially showed up out of nowhere going full dr. who in this shit and promising anyone who handed over their email address that you were going to take whats basically a tweaked xbox one s model, slam it full of top tier hardware and take that bitch sub zero, its one pay pal button short of being too good to be true. Now, with that said, i am a sceptic at the best of times, its entirely possible that you have some deep level of engineering skill (i sure as hell dont) and an actual plan to make this happen, or its possible that after consulting w/ the community and logging your build that your expectations of its performance might change or that as you develop your idea out your design will change to accommodate any changes that become apparent to reach your performance goal or whatever else. For better or worse your OP garnered some negative attention & scepticism , as a precaution the mailing list links were removed.

 

[ignsvn]

Guys, let's be civil here, and as I said before, let's just wait for prototypes with temperature, noise & power usage test data.

 

[AleksandarK]

Hmm, i am very skeptical about this.

When are you going to make a prototype?

 

[Bryce]

Well, Necere made a good point in that December is probably too soon. That date is somewhat naïve as it assumes that every aspect of the project goes just perfectly, which naturally has never happened ever. Another potential problem is finding manufactures for production, as obviously our CNCs and Laser Cutter are impractical for mass-manufacturing, but at the very least, a glossy plastic prototype (likely from ProtoLabs) should be ready for photos by then.

A proof-of-concept loop with CNC'd aluminum blocks is working now, though it doesn't quite fit inside the case (almost though).
We're also waiting for more thermoelectrics to ship (along with a CPU block and a better pump), 'cause it needs more power!

 

[K888D]

I think if you posted some pictures of the internal layout, or maybe even some photos of the cardboard cutouts you've mocked up it would give people a better understanding of what your trying to achieve and make it more believable.

If your using a custom cooling solution for both the CPU and GPU that is built into the case then I see no reason why you couldn't quite easily cool that level of hardware in a 5 litre case. Whether you can achieve that level of precision engineering that is compatible with a wide variety of components for the price point you have mentioned is the question.

 

[CC Ricers]

For me, I am guessing it's something like AMD's Project Quantum, in terms of precision and degree of customization. In Quantum, most of the watercooling loop (aside from the radiator and a couple of tubes) was completely designed and produced in-house, but all the electronic hardware was picked from off-the-shelf products. The case was built around that hardware. For more flexibility you probably need to create something like AIO closed loop systems to allow compatibility with many parts.

 

[Bryce]

I don’t think it will have the best motherboard compatibility. The spacing is pretty close, and any weird ITX boards, like those that have the RAM in a different place, probably wouldn’t fit. Other than that I’d expect most components to fit.

Anyway, I’m pretty busy on Monday but I will post some images and an airflow diagram tonight.

 

[Necere]

If your using a custom cooling solution for both the CPU and GPU that is built into the case then I see no reason why you couldn't quite easily cool that level of hardware in a 5 litre case.

It's the TECs that are the problem, since they're going to about double the total heat generated by the system. It's hard to see how 700W+ in 5L will result in anything other than hot, loud, and potentially unstable (even if the core components - CPU/GPU - are kept cool, secondary components like drives and VRMs may be subjected to proximity heating).

 

[K888D]

It's the TECs that are the problem, since they're going to about double the total heat generated by the system. It's hard to see how 700W+ in 5L will result in anything other than hot, loud, and potentially unstable (even if the core components - CPU/GPU - are kept cool, secondary components like drives and VRMs may be subjected to proximity heating).

Fair point.

But what single GPU and CPU combo consumes 700w? Max this will need to realistically cool is 350w I would have thought?