Concept Flex-ITX motherboard form factor concept

[Necere]

Both solutions would command a premium on the pricing, so I don't think that "you can use it in a standard mini-ITX config" argument has much weight. The only people willing to pay for the premium will be those who actually want the feature, so then why not make it ideal for them increasing your chances that people find it interesting instead of making an intermediate no one else will use and won't drive sales numbers up.

What makes you think it would be appreciably more expensive? I don't think a removable plastic PCIe slot is that much different cost-wise from a fixed one, when produced in volume.

You seem to be missing the simple fact that the bigger the market, the more you can sell, and the cheaper it can get. A board that can be used for all existing mini-ITX cases AND niche sandwich/console cases has a tremendous advantage over a board that can only be used in niche cases. That alone would allow the flexible board to be priced lower, even if the development cost is roughly the same.

 

[2bitmarksman]

I'm digging the concept, it makes a lot of sense for most mini ITX boards to be in a Flex ITX form factor, assuming the cost difference at scale is like $5-8 at max. It also opens more interesting options for the x16 slot if it doesn't end up being used for a GPU. I can see a right angle/bifurcation setup for dual NVMe M.2s or something without going above the height of standard IO, for example

 

[Kilrah]

You seem to be missing the simple fact that the bigger the market, the more you can sell, and the cheaper it can get

Not at all, but I just do not believe that someone who wants a board for a conventional build will choose that board among so many other offerings, so while technically you could sell as many as a conventional board since it's compatible it's not what would happen in practice.

 

[Necere]

Not at all, but I just do not believe that someone who wants a board for a conventional build will choose that board among so many other offerings

Why not?

 

[paulesko]

If I understand correctly it doesn´t have any shortcomings if wantesd to use in a traditional way does it? And it does give you more flexibility to use in diferent form factor cases without the use of a riser. I find this to be a very good idea if the costs don´t ramp up too much. It-x motherboards are very expensive as they are right now because they have a ver good quality pcb, adding more cost to that, could be too much for the final price, but if that is not a problem, then I´d like to see it.

 

[Thehack]

The thing is that this implementation doesn't mean much in terms of cost, since mITX motherboards are already very expensive to begin with: in a $50 motherboard this would be problematic... but have you checked mITX prices latetly for high-end skus?

BUT, at the same time, it offers a huge functionality bonus since it streamlines the usage of different layouts that change the traditional orientation of gpu in relation to the motherboard without requiring shielded risers that are delicate, expensive, and a general pita in many regards (let alone space-consuming).

All in all this is a very very good idea, that I hope gets some traction.

This implementation would mean a brand new spec. New designs need to be... Designed and validated.

If high end boards are already expensive due to being mini itx, imagine a different sku that has a very specific use case (mostly only via sandwich designs).

Only two layouts benefit, sandwich and opposite side horizontal. Only sandwich case has any designs out in numbers.

The issue is that this spec won't replace standard mini itx. This means this will always be secondary to the standard mini itx and likely not get made in big enough numbers to offset the costs.

Standard and regular horizontal don't need a specific boards, node 202 and rvz01/02 series already employ a cheap pcb riser.

Mobo manufacturers are still gonna make mini itx boards, but adding a sku they can't copy and paste means it has to be built from the grounds up anyways. At some point, the end result is we should get at least a product in our hand?

Why spend $$$$ on an unproven design whose only features is for sandwich case essentially?

It is better to approach it holistically and provide something that benefits more use cases.

For small makers, in small quantities I feel also doesn't help them. If you want to use a slightly different design, mounting points that differs slightly, you'd have to design a pcb specific to your case to take advantage of this.

If the whole idea is to move to pcb risers, then it'd be tougher to spec a pcb for your case than it is just to select the riser from LiHeat.

And then now your customer are segmented between those who can buy an edge mini itx and those who can't or don't want to, and PCB or flex riser designs.

 

[troybot]

The bifurcation idea is really unique. But is a hard riser really a better idea? At least a soft one can bend around an m2 ssd on the rear side for example.
But personally I would rather wait for the new intel power spec to be mainstream. Look at m2 ssds, less cables have made such a big difference to layouts (in my opinion).

 

[pwnedundies]

This implementation would mean a brand new spec. New designs need to be... Designed and validated.

If high end boards are already expensive due to being mini itx, imagine a different sku that has a very specific use case (mostly only via sandwich designs).

Only two layouts benefit, sandwich and opposite side horizontal. Only sandwich case has any designs out in numbers.

The issue is that this spec won't replace standard mini itx. This means this will always be secondary to the standard mini itx and likely not get made in big enough numbers to offset the costs.

Standard and regular horizontal don't need a specific boards, node 202 and rvz01/02 series already employ a cheap pcb riser.

Mobo manufacturers are still gonna make mini itx boards, but adding a sku they can't copy and paste means it has to be built from the grounds up anyways. At some point, the end result is we should get at least a product in our hand?

Why spend $$$$ on an unproven design whose only features is for sandwich case essentially?

It is better to approach it holistically and provide something that benefits more use cases.

For small makers, in small quantities I feel also doesn't help them. If you want to use a slightly different design, mounting points that differs slightly, you'd have to design a pcb specific to your case to take advantage of this.

If the whole idea is to move to pcb risers, then it'd be tougher to spec a pcb for your case than it is just to select the riser from LiHeat.

And then now your customer are segmented between those who can buy an edge mini itx and those who can't or don't want to, and PCB or flex riser designs.

It wouldn't be a specific use case. While it may do something better than the traditional layout it would also do exactly what the traditional layout does. So while the market would have options if the industry said this was better and this is how we are going to do things in the future then there wouldn't be any segmentation in the new market.

Why not spend $$$$ on an unproven design? It's how we innovate every day. People push the boundaries and this is their current idea of how to do that.

I get that it is a mobo and only so much is dependent on but if R&D money is the only issue then what does it really matter?

 

[gwertheim]

Here's how I see it, unless you have an MBA and work in the industry, you or I have no clue what the demand is. Same with modifications to the board, unless you are an electronic engineer who develops circuit boards, you or I have no clue how extensive the modifications have to be. I was an apprentice pipe fitter so I know about that and can speak with some concrete knowledge on the subject.

Instead of poking holes in the idea, find a way to make it work and leave the technical details to the experts.

Sorry if this comes off as hostile but I believe this project will make our builds better if @Necere and @Wahaha360 can get it going

 

[KSliger]

Dunno about this.

You'd be competing against half size PICMG 1.3 SHBs and backplanes on one end, and then standard ITX + riser on the other end, not really doing anything better than either.

PICMG you'll possibly be cheaper, maybe, but larger in case volume and lacking the ability to have multiple PCIe slots.

https://portwell.com/products/detail.php?CUSTCHAR1=ROBO-6912VG2AR

ITX + riser you'll maybe be a wash on price, but that higher price will limit your standard layout sales.

There's also companies that make side port PCIe slots, just nothing 16x.

NF797-Q370 :: JNF797-Q370 :: Intel Q370 Coffee Lake 8th / 9th Gen Mini ITX :: JETWAY COMPUTER CORP.

Jetway Corporation is the worldwide leader in high end technologies for mini-ITX, IPC, POS, Industrial Motherboard, Embedded Computer, Medical Computer, Industrial Computer, consumer electronics, and motherboards

www.jetwaycomputer.com

I have some experience with motherboard development from when we used to be part of Quanta/Kontron/Quanmax, and typically development was $300k to $500k USD. Easy enough to recoup if you can get the 15k~20k+ unit sales to do it, but I've seen a good number of motherboards have their components go EOL (mainly Intel parts) while they were still tens or hundreds of thousands of dollars in the red.

Also while searching around for possible comparisons I found this interesting iBase motherboard:

https://www.ibase.com.tw/english/ProductDetail/EmbeddedComputing/MT800M-P

 

[Windfall]

This reminds me of PIO boards, but they're hard to get,

 

[scatterforce]

Gamers Nexus is asking for manufacturing questions as they go to Tawain. I posted a question referring to this thread/idea. Go and upvote, lets shine a light on this idea and see where it can go.

eidt: the video is their current QnA. I don't know how to link a specific comment...

 

[VegetableStu]

eidt: the video is their current QnA. I don't know how to link a specific comment...

the date of your post should be a direct link if i remember right. post that link here and the "new tab" link should bring people to the youtube video with the highlighted comment brought to the top

EDIT: also damn GN gave up on testing SFF cases ,_,

 

[ermac318]

the date of your post should be a direct link if i remember right. post that link here and the "new tab" link should bring people to the youtube video with the highlighted comment brought to the top

EDIT: also damn GN gave up on testing SFF cases ,_,

I think they gave up on testing because they value consistent, comparable testing more than anything (see their recent videos on Air Cooler testing) and doing that in SFF cases is very difficult. Is it a worthwhile comparison to compare a Ghost S1 with one cooler vs. an NCase M1 with a different cooler? What does that actually tell you? Their standardized testing harness for ATX cases uses the exact same fans and the exact same coolers with the exact same GPU, CPU and Motherboard. That kind of consistency is impossible in SFF cases.

A comparison of which case gets the best thermals under their unique ideal conditions might be redundant - most of us can guess (outside of weird design issues) that certain cases are better than others based on the coolers supported.

 

[riba2233]

I think they gave up on testing because they value consistent, comparable testing more than anything (see their recent videos on Air Cooler testing) and doing that in SFF cases is very difficult. Is it a worthwhile comparison to compare a Ghost S1 with one cooler vs. an NCase M1 with a different cooler? What does that actually tell you? Their standardized testing harness for ATX cases uses the exact same fans and the exact same coolers with the exact same GPU, CPU and Motherboard. That kind of consistency is impossible in SFF cases.

A comparison of which case gets the best thermals under their unique ideal conditions might be redundant - most of us can guess (outside of weird design issues) that certain cases are better than others based on the coolers supported.

There has to be a way.

I would do it like this:

for GPU - use reference 2080ti, it should fit most cases and produces a lot of heat

for CPU - use 3700x (enough for gaming and most workstation tasks, but still low enough TDP) with noctua nh-l9i as a standard option, since it should fit most cases. But, if case allows for bigger cooler, also make a test with the biggest cooler it will fit, to show it's advantage.

I don't know, I'm just writing some ideas, but I think that there should be a way for sure.

 

[ermac318]

There has to be a way.

I would do it like this:

for GPU - use reference 2080ti, it should fit most cases and produces a lot of heat

for CPU - use 3700x (enough for gaming and most workstation tasks, but still low enough TDP) with noctua nh-l9i as a standard option, since it should fit most cases. But, if case allows for bigger cooler, also make a test with the biggest cooler it will fit, to show it's advantage.

I don't know, I'm just writing some ideas, but I think that there should be a way for sure.

I would say get on their Patreon and discuss it on their discord. They are usually receptive to any subject matter experts, and I'm sure if they felt they could get a good idea of what something realistic is, they would be interested. But it's also an issue of time and bandwidth, and they may already have such a backlog of projects that it's not a high enough priority.

 

[Steelmint]

The only issue I can see coming from an incredibly simple PCB design background is the missing real estate I now have on my board. Most boards also have some M.2 Slots, connectors or the chipset in the area that is now being reduced in size.

 

[Steelmint]

Sorry for the 2nd comment but I did also realize another thing, the PCIE slot on motherboards tends to be very close to the RAM slots, how would this increased cutout work together with RAM tracing and positioning?

 

[arthasdklol]

First motherboard with ATX12VO : https://www.asrock.com/mb/Intel/Z490 Phantom Gaming 4SR/
Direct link to the board: https://www.asrock.com/mb/photo/Z490 Phantom Gaming 4SR(L1).png
They replaced 24 pin by 10 pin (+6+4+4 ?), I wonder how big the PSU will be...

 

[riba2233]

First motherboard with ATX12VO : https://www.asrock.com/mb/Intel/Z490 Phantom Gaming 4SR/
Direct link to the board: https://www.asrock.com/mb/photo/Z490 Phantom Gaming 4SR(L1).png
They replaced 24 pin by 10 pin (+6+4+4 ?), I wonder how big the PSU will be...

Those two 4 pins are probably power outputs for the sata drives. Now, 6 pin is the unknown for me, it's either PCIE 6 pin or something special for other rails like 5V and 3.3V.