• Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!

Prototype Project Orthrus - Smallest SLI mATX case (V2)

Hazy

Efficiency Noob
Apr 15, 2018
7
10
@Hazy it appears the website he used to upload all the pictures was taken down. it looked like a crappy website too, it was always filled with ads and stuff so it isn't too surprising. I think I am like you too lol, where I have to look at this case and check the forum constantly because I love this case and can't wait for my future dream build. Lucky for me tho, I saved some pictures on my phone and computer long ago lol. My favorite image is this one that I combined to give one great all around view of the case.

You are a lifesaver @xSanchez78
I can finally get back to normal
 
  • Like
Reactions: Biowarejak

Ray2r

Caliper Novice
Apr 4, 2018
31
12
@Hazy it appears the website he used to upload all the pictures was taken down. it looked like a crappy website too, it was always filled with ads and stuff so it isn't too surprising. I think I am like you too lol, where I have to look at this case and check the forum constantly because I love this case and can't wait for my future dream build. Lucky for me tho, I saved some pictures on my phone and computer long ago lol. My favorite image is this one that I combined to give one great all around view of the case.

I have not seen this before. What is this? From the picture it is an ITX board with support for only 1 PCIe slot. But the case itself extends it into 3 slots or does the 1 slot is used to connect with the 2 slots on the case allowing for 2 PCIe slots total?
 

Ray2r

Caliper Novice
Apr 4, 2018
31
12
I'm using a PCIe bifurcation riser to split a single 16x slot into 2 x 8x slots.

I'll repost the pictures soon by the way :)
Is this going to be an optional accessory for ITX users or is it default? Like many other custom case sellers, they sell a base case then charge additional for accessories.

I want to make sure that this is still meant to be a mATX case as I want to fit the x399m TR4 motherboard.
 

xSanchez78

Average Stuffer
Jan 15, 2018
57
56
Yeah you can still pick between mATX or mITX, it's just that when using an ITX board, you get room for the reservoir/pump combo for more conventional water cooling setups. The only water cooling you can get with mATX boards is an AIO for the CPU, however with a CPU as hot as threadripper, I don't think you're going to want to water cool the GPU's too. And the only extra thing needed with the ITX setup is the bifurcation riser, it's not really a huge change other than that. However I'm curious if @LukeD will be able to sell the riser with the case or not, otherwise you have to obtain it on your own. But anyways yeah the mATX x399 board should definetly fit in this case.
 

ShieldSean

Cable Smoosher
Dec 23, 2017
9
14
Sorry I misunderstood I thought he said it was coming on April 1 St But he said that he was going to make an announcement on early April.

Hope my tempering of your expectations helped, here we are in late April without THE announcement. @LukeD keep us updated, there's a bunch of peeps here who want to give you our money! Hopefully you've kept count so you can have an idea of what to charge to recover development costs while not going too high to kill sales entirely. I realize this isn't small, but this is what I'm looking at in competition with the Orthrus: http://ttpremium.com/product/thermaltake-core-p1-tg-mini-itx-wall-mount-chassis/
 
Last edited:
  • Like
Reactions: Biowarejak

Ditlan

Chassis Packer
Feb 28, 2017
18
20
I know people are eager for this case to land, but there should be some serious consideration for making the creator Luke feel in any was pressured - there is a certain privilege to being able to read on a forum the development of a product, I have seen it a few times (many more if you count backed Kickstarter projects) and unless its an established company bringing out another iteration of their product then it always, almost without exception, takes much longer than anyone wants it to.

Suppliers promise parts it turns out they cant deliver, the Chinese go on holiday, when production kicks up there is a higher failure rate than anticipated or one part needs remodelling... the reasons are endless and there are so many moving parts to the process that delays are inevitable.

I think the right attitude to these kind of things is that they will be done when they are done, but you can keep your fingers crossed that it will be sooner rather than later.
 

BennyBurner

Efficiency Noob
New User
May 5, 2017
5
16
Some very true points Ditlan - and I agree that impatient forum posts don't help Luke at all. I see it as a number of potential buyers who are invested in seeing this project succeed, becoming vocal when announcements are delayed.

You said it yourself, part of the appeal of this forum is in following the development of this case. I think I'd rather hear about all of the development, setbacks and all, rather than just the highlight reel. That's the beauty of sharing a passion project with your audience - they become more invested in the product if they feel connected to the painstaking and often frustrating journey to market.

I hope that makes sense. It may just be that there's no new information - good or bad - in which case all we can do is hang around and wait for more :)
 

HeroXLazer

King of Cable Management
Sep 11, 2016
707
476
Some very true points Ditlan - and I agree that impatient forum posts don't help Luke at all. I see it as a number of potential buyers who are invested in seeing this project succeed, becoming vocal when announcements are delayed.

You said it yourself, part of the appeal of this forum is in following the development of this case. I think I'd rather hear about all of the development, setbacks and all, rather than just the highlight reel. That's the beauty of sharing a passion project with your audience - they become more invested in the product if they feel connected to the painstaking and often frustrating journey to market.

I hope that makes sense. It may just be that there's no new information - good or bad - in which case all we can do is hang around and wait for more :)
Anyone who has never made a mistake has never tried anything new. - Albert Einstein

I think this fits perfectly right now.
 
Last edited:

Ellimist

Trash Compacter
Feb 11, 2018
37
39
Are the TR4 mATX boards going to be okay for cooling? I am seriously considering an AMD as my next processor but the heat is not exactly the highlight of threadripper. Will an AIO be okay in this case with everything else?
 

xSanchez78

Average Stuffer
Jan 15, 2018
57
56
@VegetableStu The enermax cooler probably won't fit due to width limitations, and custom loops obviously aren't possible with an mATX board in this case, because there is no room. However asetek based coolers will work with the little brackets that come with threadripper and it will probably cool just fine. After all, @LukeD proved that a single 240mm radiator can cool two 1080 Ti's and an 8700k, so a single 180 watt TDP processor would definetly cool fine.
 
  • Like
Reactions: VegetableStu

Hazy

Efficiency Noob
Apr 15, 2018
7
10
@xSanchez78 @VegetableStu I think there is a chance to do an mATX build with a custom loop, but it would require that you use one of the GPU slots to house the Phanteks Glacier 160 reservoir, since it seems to be the only reservoir that slim, or a flow rate indicator as a substitute. I think it would be a similar setup to @LukeD 's, just with the res taking the place of the second GPU.

However, this is all speculation because I am still not sure if that res would fit since it is still 1.5 inches deep and I am not sure if that is with or without the pump on. There is also the problem of mounting it, but I'm sure you modders out there will find a way. Also, I am not sure if there is even enough space to run a tube between the GPU and the fans/radiator. If you used one of the smaller cards, like one from the Zotac mini series, you could, but I am not sure how much the open air GPU would affect the temps or if there are water blocks for them. If not, then you might just want to cover up the sides of the Zotac cards around the heatsink and turn it into a mini-blower style card ;P

So a couple questions for @LukeD :
1. How deep of a video card could you fit in the case? I know it is the standard 2 slots, but is there any wiggle-room to maybe fit a 1.5 inch deep reservoir in one of the graphics cards slots?
2. With the average FE graphics card w/ the stock cooler on and the radiator and fans, how much space is between them? In other words, would you be able to fit any tubing between them?
3. Is it still a 53 mm limit on the RAM or could a 55 mm RAM stick fit?

One last thing, @LukeD , would you care to elaborate on that "one more change that [you] think people will treat as a very welcome feature in the final version of the case"? I was able to control my excitement and anticipation for this case until I reread that post and now I can't get it out of my head.
Anyways, good luck working with the suppliers and I can't wait to see your final design!
 
Last edited:
  • Like
Reactions: Biowarejak

xSanchez78

Average Stuffer
Jan 15, 2018
57
56
Hey @LukeD I have a question, and I've been thinking about this for quite awhile now. I was thinking about doing a build in this but with AMD GPU's with Crossfire, and here is why: there is no support for a HB SLI bridge. The custom length SLI cables are cool and all, but I don't believe they provide a "High-Bandwidth" connection like the newer solid bridges do ( Please correct me if I'm wrong, because if those SLI cables actually do provide an HB connection, I'd definetly go with Nvidia instead ). I did some research into what issues might be caused when using ribbon cables vs an HB bridge, and it appears you get lower frame rates, some games worse than others, but it is still lower than when using an HB bridge, which is still unappealing to me, no matter how drastic the drop in fps is. Secondly, I read somewhere that it can cause other issues than just lower fps, such as game stuttering and latency issues, especially in 4K ( And yes, I do plan to game at 4K ). So then I thought, what about Crossfire? I did some research into AMD's lineup of GPU's, and I found that the AMD Radeon RX Vega 56 is actually a pretty good option. It seems to perform as good as a GTX 1080 ( or atleast better than a GTX 1070 Ti ), and although it isn't as power efficient as a 1080, it's still not too hot, with a TDP of 210 watts, and definitely not as hot as a 1080 Ti, which has a TDP of 250 watts, so it should be able to be cooled with the 140 watt TDP CPU I plan to use. Also, looking at a recent video comparing SLI and Crossfire, it actually seems that currently, Crossfire seems to scale slightly better than SLI in most games, which is really surprising. But most importantly, it doesn't require any cables between the GPU's, meaning a "Low-Bandwidth" connection won't become an issue with this setup. So my question is, two things actually:

1. I'm assuming they're not but, just to be sure, are the custom SLI cables you have used for this case able to provide a High-Bandwidth connection?
2. If not ( meaning I'd rather go with AMD Crossfire ), is it possible to enable Crossfire on an Asrock ITX board like you have done with SLI? From what I read from you in the post about PCIe Bifurcation on this forum, enabling SLI only required the "SLI Certificate" to be in the BIOS of the motherboard. Does Crossfire require the same kind of thing or similar, and if so, can it be added to the BIOS in order to be enabled just like SLI?

AMD Radeon RX Vega 56 w/ EK Water Block
 

Ellimist

Trash Compacter
Feb 11, 2018
37
39
Hey @LukeD there is no support for a HB SLI bridge. The custom length SLI cables are cool and all, but I don't believe they provide a "High-Bandwidth" connection like the newer solid bridges do ( Please correct me if I'm wrong, because if those SLI cables actually do provide an HB connection, I'd definetly go with Nvidia instead ).

I was looking into this. It turns out theres a slight pinout change to make it read as a HB bridge to the graphics card. So a custom cable could be configured to be a HB bridge as long as its pinned correctly. See here: https://overclocking.guide/nvidia-hb-sli-bridge-technical-review/

"The LED / HB Bridge is slightly different. One pin is not routed to ground but no supplies 1.8 Volt to the SLI Bridge. This supply voltage is also used to feed the LED. In addition the driver can notice if the pin is bridge to ground or not. Bridged to ground, the driver detects a legacy bridge. Not bridged to ground it will detect a HB bridge."
 

xSanchez78

Average Stuffer
Jan 15, 2018
57
56
@Ellimist interesting article. So are you saying that it's possible to trick the drivers into thinking that it's an HB bridge by modifying the cable somehow? And if so, would the cards then just try pushing the full bandwidth across the cables, and would the full bandwidth be unstable or be limited by the cable? And if the full bandwidth is solid across the cable, does this mean we can have full HB SLI performance with these cables?
 

Ellimist

Trash Compacter
Feb 11, 2018
37
39
My guess and I really mean guess is that if you were to build a cable with the appropriate specifications for pushing signals at 650mhz and then have the pinout correct for a HB bridge then it would work as a HB bridge.

I'm no engineer though but I'd sure as hell test it out. hehe.

If it does work it would be a major boost to the case long term as the HB bridge becomes more utilized. The other thing I see happening is that the bridge goes away as we start to get dual x16 slots (Looking at you AMD) and PCIe 5.
 

ShieldSean

Cable Smoosher
Dec 23, 2017
9
14
Suppliers promise parts it turns out they cant deliver, the Chinese go on holiday, when production kicks up there is a higher failure rate than anticipated or one part needs remodelling... the reasons are endless and there are so many moving parts to the process that delays are inevitable.

One of my favorite stories is the part where you're casting something in plastic, and when you go to a production cast, the plastic for the end product may have enough variation in it's composition so as to change the amount of shrinkage you get when cooling the part. And then of course the part doesn't fit with other pieces and you have to remake a custom 1-off steel mold to compensate for the difference in plastic type, and that steel mold will take weeks to rebuild.

So yeah, invariably there are things you just can't account for, unless maybe you've been embedded in manufacturing for years and you have solid control of the supply line to make sure all the initial materials are up to grade. Basically production is crazy hard... Not unlike writing software, but with software you can iterate in minutes instead of weeks.
 

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
And even when you've been manufacturing for years supply chain issues can still catch you out: Remember when there were a spate of Macbook Pros saround a decade ago where the display would start showing corruption/garbage? That was traced back to the Nvidia GPU, and Nvidia then traced it back to the solder balls holding the GPU die itself the the package cracking due to thermal cycling. And that was due to a chemical supplier of the solder-ball manufacturer changing the formulation of the solder to be used in the solder balls, but not changing the designation of the solder (which meant the designation of the solder balls also remained unchanged), which ended up with solder balls being used with the wrong soldering process without anyone being informed up the chain.
 
  • Like
Reactions: VegetableStu