Small Form Factor Pico-ITX x86 board on Kickstarter? (VOTE and get PRIZE, closing TONIGHT 1/Feb)

3
C
A
B

 

3 C B A


Sent on my CM10 JB powered i9100 by TapaTalk 2

 

1. I see no need in DIY x86 board on Kickstarter, because DIY ARM boards (Raspberry Pi, PandaBoard, etc.) cover all the needs.

However, for the record ABC.

 

3

ACB

 

3 - A,C,B

 

Mini update:
1) So far, 24 people have voted - thank you so much for giving this valuable feedback!
2) I am considering to decrease the required number of votes (and still keep the promised 5 prizes) - therefore making the vote more attractive with more chance to win the prize. I will push an update as soon as I come with some fair way to reduce the numbers.

Current vote count graph (updated 13-Jan-2012 11:36GMT)
​

 

x86 won't cost down enough I expect. If you want a contact at VIA for their pico hardware I can send you a contact.

 

Bindi you boss.

 

The downside of VIA is their software generally, but I wonder if they would let another company tweak it.. might be worth a try. Depends if you're software driven or hardware.

 

I would go for
3.
A.
C.
B.

 

Bindibadgi: I would appreciate a PM with a contact, thanks! But I doubt I would do VIA because then I would compete with their own Pico-ITX boards, and I don't see a reason for them selling me their components cheaper than they sell their boards AMD sounds a much better option here since AMD don't do boards

I am currently trying to narrow down good options, and I've effectively come up with 4:

1) Best-performance (expensive) - AMD A8-4555M looks like it could fit on Pico-ITX, and still provide 4 Cores
2) High-performance (somewhat less expensive) - Intel Atom S1260 (dual-core server-grade with ECC support and 8x PCIe 2.0 lanes) paired with NVidia Geforce GT625M
3) Moderate price/performance - Intel Atom dual-core with integrated graphics (i.e., D2560 / N2850 - both seem to be yet unreleased but will probably be soon)
4) Cheapest / lowest performance - DMP Vortex86MX/MX+ or RDC IAD100PE . I know those processors are very slow compared to Intel/AMD/VIA, and they cannot run Windows7 or any i686 or 64bit code, but they are dirt-cheap for what they offer, and can still run WindowsXP and other x86 32bit alternatives.

My choice from the 4 options above will be mainly based on what the public values most (size, price, performance) - hence my request to value them in your preferred order, in addition to answering the question Of course this is only relevant if the majority answers "3" (YES to Kickstarter campaign).

 

1 and 2 will cost $300-500 a board once you've done development and size-down.
AMD Brazos is still a 3-chip solution, and you need SoC or 2-chip.
Atom is probably easier as 2-chip and in future SoC, although I'm not sure what the minimum sales size for AMD and Intel is.

 

Bindibadgi: AMD solution I chose is 2 chip (in fact, all AMD Brazos are 2 chip, unless I am terribly wrong on what counts for "chip"). Minimum order is 1,000 CPU if I understood the documentation correctly. You are right about the prices for the top two options, but at large scale (1,000+ boards) and doing it at cost (not-for-profit), I think I might be able to offer one of them for $250 or less.

 

I'll be as honest as possible: selling Brazos for $250 is a dead market when mini-ITX boards have been around 3 years for well under $100.

People don't care that much about pico ITX, only when it's ridiculously cheap.

You're best off buying something A15 ARM quad core and packaging that. Chip cost is perhaps $40 most, but then you need less components unlike x86.

 

Anything with less power then AMD 350 is wasted efforts imho. And the finished product would need to stay below $300 to get people interested.

The intel NUC (4"x4") is selling for $300 and boasts a mobile i3, complete with an enclosure.

 

That's exactly the purpose of this thread - to see whether there is interest in x86 DIY board, since currently there are plenty of DIY ARM boards, but nobody has offered an x86 one.

That's for dual-core and larger size, so obviously smaller size and twice cores is more expensive.

To some up, I am trying to gather feedback whether there is interest. Putting up Kickstarter project costs nothing if it fails, and nobody will lose their money unless there is enough collected and project goal is passed, which would mean there is interest.

BTW, Bindi, you didn't formally vote, but reading your comments I would count your vote as (1) if you don't mind

This is exactly what I am trying to achieve

 

An update on the minimum number of votes:

Currently we have 26 votes. If I don't get the 200 votes originally requested (which is what likely to happen), the required number of votes will decrease by 50 each week; up to a minimum of 50 votes which will not be lowered further. Therefore, one week from the topic start, on January 18th, the required number of votes will become 150, and one week later (if 150 votes are not reached), the number will decrease to 100 on January 25th, and if one week from then, 100 votes are not reached, the number will decrease last time, on February 1st, to 50 votes.

The promised number of prizes (5) will stay the same, regardless of the required number of votes, to reward the people who have already voted.

 

3.) BCA

 

3
C,A,B.

 

3, CAB.

Windows on a tiny device wold be cool.