Kickstarter For Open Source GPU

I don't understand why their first goal is so pricey for what it delivers, but then it seems to go the other way and twice the cost rewards with a lot more than twice the work (just guessing though :-( Maybe they have some agreements in place with some of the developers to pay for releasing some of the code as open source?

That's a very good question.  Since they already have the 2D functionality working and they claim that there is no risk for the first target, what's the $200K for?

I believe it would be for Manufacturing the Chip.

What they've done up to this point is mainly Software.

R_Phoenix wrote:

 

I believe it would be for Manufacturing the Chip.

 

What they've done up to this point is mainly Software.

I see your wink, of course.

Just in case others miss it, $200k wouldn't come even close for a project involving delivering a physical device.  They explain clearly that the project entails delivering only Verilog code and related development materials:

Silicon Spectrum writes:

 

Our deliverables are Verilog source code and testbenches.

So it's an interesting question where the money will be spent for the (already nearly finished?) initial goal.  The other interesting question is about their expectations regarding the number of backers, since the number of people for whom Verilog is a worthwhile personal acquisition is not really very large, at least not by the standards of most Kickstarters.

I hate to say it but this looks like an attempt to get $200k for some Verilog they can't sell any other way.

They are incredibly vague (unless I missed something) about the size of FPGA and the performance that will be achieved.

It's also not clear if there is any comitment to open sourcing the drivers for Linux or Windows.

Any serious 3D graphics accelerator needs GB of RAM which will be off chip - it takes a lot more than the basic IP to turn that kind of technology into  a working board so I'm not at all sure who the later stages of the project are aimed at.

MK

Unfortunately I have to agree with you.  Although I would like to advocate strongly on behalf of OSHW enthusiasts and encourage them to do a lot more to bring hardware under their control instead of just taking what they're given and having to like it, I don't see how this project achieves that.

Equally importantly, on the Kickstarter front it doesn't deliver value for money, in that the number of backers for a Verilog deliverable is likely to be extremely low, and therefore the amount pledged per person would have to be extremely high to meet the target, which mades no sense on value for money grounds for an open HDL design.

The days of paying a lot of money for code are long gone, and that applies to HDL just as much as to code for CPUs.  Some people just haven't noticed the sea change yet, and this is a particularly bad case of it.

Michael Kellett wrote:

 

I hate to say it but this looks like an attempt to get $200k for some Verilog they can't sell any other way.

Seems that way. In addition to the general vagueness, two things that caught my eye were the PCI interface, not PCIe or the type of interfaces common for SoC's, and Direct3D 7/8.  So the design seems rather out of date already.

Then stuff like bump mapping being in the $800,000 goal and the early 90's style screenshot just adds to the amazement.

Performance wise, did you watch the video ?  If it's representative of expected results, well...

Even the premise of designing an OSHW GPU targeting an FPGA is rather questionable, given how no FPGA in existence is programmable with an open source toolchain.  Although they're also proposing to license it for ASIC production, it's debatable whether a primitive design which only committed OSHW fans could love would ever be chosen by such a manufacturer given it's likely meagre sales.  The business plan doesn't seem very convincing, no matter how much I'd love to see an OSHW GPU.

Althought their bailliwick is Verilog, I wonder if an OSHW GPU wouldn't be better implemented in a completely different way, a way that open source software people would find more suited to their skills.  It seems a bit silly to implement the shaders of the programmable pipeline in an FPGA when the Epiphany device is already available and provides 16 cores of ASIC efficiency ready and waiting to run shader processing in a novel way.  Nobody said that GPUs can only have the design preferred by nVidia and AMD.

selsinork:  Our replies overlapped, but I'll just say that an OSHW GPU doesn't have to look anything like the commercial ones.  3D graphics continues to evolve, and raytracing scenes or voxel processing isn't necessarily handled optimally by the current generation of graphics cards.

Even the premise of designing an OSHW GPU targeting an FPGA is rather questionable, given how no FPGA in existence is programmable with an open source toolchain.  Although they're also proposing to license it for ASIC production, it's debatable whether a primitive design which only committed OSHW fans could love would ever be chosen by such a manufacturer given it's likely meagre sales.  The business plan doesn't seem very convincing, no matter how much I'd love to see an OSHW GPU.

Althought their bailliwick is Verilog, I wonder if an OSHW GPU wouldn't be better implemented in a completely different way, a way that open source software people would find more suited to their skills.  It seems a bit silly to implement the shaders of the programmable pipeline in an FPGA when the Epiphany device is already available and provides 16 cores of ASIC efficiency ready and waiting to run shader processing in a novel way.  Nobody said that GPUs can only have the design preferred by nVidia and AMD.

selsinork:  Our replies overlapped, but I'll just say that an OSHW GPU doesn't have to look anything like the commercial ones.  3D graphics continues to evolve, and raytracing scenes or voxel processing isn't necessarily handled optimally by the current generation of graphics cards.

Morgaine Dinova wrote:

 

Nobody said that GPUs can only have the design preferred by nVidia and AMD.

 

selsinork:  Our replies overlapped, but I'll just say that an OSHW GPU doesn't have to look anything like the commercial ones.  3D graphics continues to evolve, and raytracing scenes or voxel processing isn't necessarily handled optimally by the current generation of graphics cards.

Agreed.  Ultimately though there's a couple of things a GPU needs to be able to do. For the sake of argument lets say that boils down to H.264 and OpenGL. Actual implementation details aside, you're going to have to deal with accelerating those API's in hardware.  Do some random thing of your own that no software supports and you have a product with no market and are likely doomed to failure.

So whether you implement OpenGL & h264 in the same way as everyone else, or do it with an array of Ephiphany cores and some novel approach to the problem is almost irrelevant.  It all comes down to whether the competition can do the same task faster and/or cheaper through the same API. If they can do either then your chance of success will be reduced, if they can do both...

I'm all for new and interesting ways to do things, especially if it's done in an open way that benefits everyone by giving us access to the source.  Still, if it's not competitive you become a follower not a leader and nobody will care other than a few die hard OS evangelists.

selsinork wrote:

 

I'm all for new and interesting ways to do things, especially if it's done in an open way that benefits everyone by giving us access to the source.  Still, if it's not competitive you become a follower not a leader and nobody will care other than a few die hard OS evangelists.

Ah, but my view of OSHW and SW is fundamentally different to that market-centric one.  I believe strongly that market penetration and success is completely irrelevant and even a hindrance to the cause of openness, as it narrows the scope of what is possible to only those open projects that make sense commercially.  That's a terrible straightjacket!

OSHW fans conceived, designed, implemented and promoted 3D printing of the RepRap kind entirely without reference to commercial viability, and business only latched on to the idea later.  That's the open model that I value most, and it could happen in the field of computers just as much as in 3D printing.

Although enthusiasts don't (yet) have the ability to fabricate their own integrated circuits and therefore there will be a lot of "vitamins" (necessarily commercial components) in an OSHW computer, if it's designed for easy parts replacement then the individual vitamins become less fundamental because they can easily be swapped out for others that perform the same function.

The important point in such a design is that only the vitamins are subject to commercial pressure for viability, but the design of the computer as a whole does not need to suffer from it.  As a result there is no need to convince the vitamin producers of the merits of openness (a difficult task at best), nor do they even need to know that their component is being used as a vitamin.

Having no interest at all in commercial success nor market penetration frees OSHW designs from market-oriented considerations, and in so doing it frees the minds and the capabilities of its advocates.  It's a very powerful approach to take in an otherwise commercially restricted world.

Morgaine Dinova wrote:

 

Ah, but my view of OSHW and SW is fundamentally different to that market-centric one.  I believe strongly that market penetration and success is completely irrelevant and even a hindrance to the cause of openness, as it narrows the scope of what is possible to only those open projects that make sense commercially.  That's a terrible straightjacket!

Agreed, and philosophically you won't get much of an argument from me. Practically though, OSHW is chained to the commercial side by the need to build something physical. Whether that's a GPU ASIC or a 3D printer or whatever, you need to fund the physical aspect somehow. That's why the likes of kickstarter exists after all.

Your idealistic position of not caring about the commercial world is fine up to a point, it lets new ideas grow.  The only real problem is that for HW, it'll either stay an idea or you'll need help from the commercial world to turn it into reality, and that commercial world won't help unless you can supply the funding up front or you can convince them of commercial mass-market success.

That's the big difference between OSHW & SW. SW will flourish as long as it targets an easily available commercial platform, all you need is volunteer time to write code. Those same volunteers can design OSHW, but you still need to build it, pass EMC regulations etc.

BTW, finished building your 3D printer yet ?

selsinork wrote:

 

Agreed, and philosophically you won't get much of an argument from me. Practically though, OSHW is chained to the commercial side by the need to build something physical. Whether that's a GPU ASIC or a 3D printer or whatever, you need to fund the physical aspect somehow. That's why the likes of kickstarter exists after all.

 

[Rest of your comment continued in the commercial vein. -- Morg]

I think there's something wrong with the premise there.  Isn't what you write presupposing that OSHW only gets bought as a fully functional commercial product rather than producing non-commercial derivatives?  Although the purely commercial case certainly exists (eg. the BBB), it's not what OSHW is primarily about --- that initial release (whether commercial or not) is merely the starting point.  If a product were sold as OSHW and that were it, the process ended there, OSHW would be merely informational  rather than recursively priming the pump of modification and re-release, both commercial and non-commercial.  It would be largely pointless except for nosy busy-bodies like us who like to peer under the hood.

But that's not the sum-total of OSHW at all.  Instead, it's a community thing, evolving the initial OSHW design in new directions and spreading the love around mostly outside the commercial arena, and only occasionally giving rise to new commercial versions made by other manufacturers.  We haven't seen this happen with complex products like the BBB (yet), but it happens continually in the Arduino community and in the Seeed/Sparkfun area.

So no, OSHW is not limited by the need to make commercial products at all, nor require Kickstarter-type funding.  Those products and projects are typically the best known junctions on the OSHW highway, but along the road between them are the bespoke hacked versions and the small runs of bare PCBs and the ad hoc kits and so on.  This happens a huge amount in the 3D printer community especially, as the barriers are even lower when you can manufacture parts on your printer and sell them on eBay or dish them out to your friends in the community or hackerspace.  That's the real OSHW in action.

The kind of OSHW where a commercial product meets the definition but doesn't give rise to any community derivatives at all is really quite an aberration of the OSHW concept.  There are good reasons for it of course (complex boards are hard to work with and costly to recreate), but recognizing the reason doesn't mean that it's satisfactory.  It's OSHW in name only, not working as it should.

I think we need to make our computer boards more modular.  If they were made out of several smaller parts then these would be more easily hacked or replaced by other better versions, and small-scale manufacturers of derivatives would then be much more likely to pop out of the woodwork.  When OSHW products are complex and monolithic it creates a huge barrier to this happening.

PS. Guzunty Pi is a wonderful example of OSHW modularity encouraging easy add-on projects without major investment, notwithstanding that Pi itself is not OSHW.  Just imagine if the entire computer were made of such small replaceable pieces --- there would be continuous evolution in the community.

Morgaine Dinova wrote:

 

We haven't seen this happen with complex products like the BBB (yet), but it happens continually in the Arduino community and in the Seeed/Sparkfun area.

Ok, but Arduino and Seeed/Sparkfun all sell their stuff commercially for the sort of people who don't want to get into the messy chemical processes of etching their own PCB's or soldering stuff up themselves. Get into BBB like stuff that has high density BGA parts and the number of people capable of building one outside a commercial environment becomes vanishingly small.

So no, OSHW is not limited by the need to make commercial products at all, nor require Kickstarter-type funding.  Those products and projects are typically the best known junctions on the OSHW highway, but along the road between them are the bespoke hacked versions and the small runs of bare PCBs and the ad hoc kits and so on.  This happens a huge amount in the 3D printer community especially, as the barriers are even lower when you can manufacture parts on your printer and sell them on eBay or dish them out to your friends in the community or hackerspace.  That's the real OSHW in action.

lol.. I see where you want to go..  So small run of PCBs, home lab, or commercial PCB fab ?  Selling on ebay..  again that's commercial, albeit smaller scale.

I'm not suggesting at all that you need to make a commercial product out of OSHW, just that you can't avoid the intersection with the commercial world. Funding that interaction out of your personal pocket change will necessarily keep things very small scale. 

Since we have little choice other than to use commercially available parts like microcontrollers there has to be some acknowledgment that, like it or not, a lot of the boundaries are set by those commercial manufacturers we become at least partly reliant on when we use their parts. If you acknowledge that, you should also be aware that generally they really don't care about us unless we're buying stuff in volume. 

There are good reasons for it of course (complex boards are hard to work with and costly to recreate), but recognizing the reason doesn't mean that it's satisfactory.  It's OSHW in name only, not working as it should.

That's the crux of it, the commercial world has no obligation to make it easy for us. No, it's not satisfactory, but not a lot we can do about it.

John frequently comments on FPGA vendors not releasing bitstream formats and therefore limiting the OS communities ability to do interesting new things, and I think that's a really good example of the conundrum we find ourselves in as OS advocates.

On the software side, we see things like linux where today it's largely paid developers working for some company who makes money from selling services/support based on that work.  It's increasingly difficult to separate the two sides.

So really, all I'm trying to say is that we need to recognise that there's an interdependency, we don't need to like it..

Morgaine Dinova wrote:

 

PS. Guzunty Pi is a wonderful example of OSHW modularity encouraging easy add-on projects without major investment, notwithstanding that Pi itself is not OSHW.  Just imagine if the entire computer were made of such small replaceable pieces --- there would be continuous evolution in the community.

The inevitable conclusion to that becomes something like today's x86 PC where the smaller replaceable pieces are built to some set of agreed common interfaces, the hope being that everyone has the same opportunity and that there's some upgrade path that doesn't leave you with a 'scrap the lot and start again' problem.

Not sure it's worked out well for the PC though.

selsinork wrote:

 

Get into BBB like stuff that has high density BGA parts and the number of people capable of building one outside a commercial environment becomes vanishingly small.

Well that is what I wrote.  And to avoid that bad situation we should not place ourselves in such a position where we have no options.  We place ourselves in that position by feeding exclusively at the trough of companies that make complex boards.  It's very unsatisfactory unless we want to be nothing more than consumers.

Since we have little choice other than to use commercially available parts like microcontrollers there has to be some acknowledgment that, like it or not, a lot of the boundaries are set by those commercial manufacturers we become at least partly reliant on when we use their parts. If you acknowledge that, you should also be aware that generally they really don't care about us unless we're buying stuff in volume. 

Don't confuse components with entire products.  It's not a problem using documented components that we can't manufacture ourselves, as long as we can swap one for another.  They don't have to be identical --- eg. we can cope with different microcontroller instruction sets by recompiling, and different peripherals by switching BSPs.  When our  systems are modular it gives us options, and we're no longer subservient to the commercial decisions that companies make.  It's just like in the PC world when we assemble our own computers from sub-assemblies, it's very empowering not being at the mercy of limited options, expensive support, and the high prices that stem from being tied to a manufacturer.

OSHW always allows this in principle, but to turn it into practice requires us to seek modularity, because without it the openness provides far less value, and we turn into nothing more than consumers (ick) and totally impotent to affect the direction of our technology.  It would be depressing in the extreme, and it's not necessary.

That's the crux of it, the commercial world has no obligation to make it easy for us. No, it's not satisfactory, but not a lot we can do about it.

There is not a lot we can do about the semiconductors (in fact, nothing at all), but there is a huge amount we can do about computer boards.  We don't have to just wait with anxious eyes and impotent expression until companies release products that they've designed behind closed doors and which only by coincidence are what we actually want.

Designing a modular concept, laying out boards, and getting a small run of bare PCBs manufactured is well within the capability and reach of small groups of individuals without a large bank balance nor company behind them.  It requires a similar amount of time investment as writing a complex software system, and individuals or groups do that in open source software all the time.  It requires a little more money to have boards made than just typing 'make', but for small modules it's really peanuts.  PCB manufacture has really been democratized in recent times, and the only thorn in our side is BGA mounting.  Solve that problem and OSHW becomes nearly as easy as software.

So really, all I'm trying to say is that we need to recognise that there's an interdependency, we don't need to like it..

Sure there's a dependency, but it need only be a dependency on components or modules, not for whole products, and we should never have a dependency on a single company.  And yes, we do need to like the situation we create for ourselves, that's the whole point of being techies in this space.  It's the non-techies who are wholly at the mercy of companies and can do nothing other than consume, like it or lump it.  We can do far better for ourselves (and even for adventurous non-techies who like to follow along), but it will not happen if we do nothing other than consume.

It's a total misapprehension of OSHW that our role is just to buy the commercial hardware that we're so generously offered and that its openness is purely informational.  Again I'll point to Arduino and RepRap as OSHW communities that use commercial components but are not limited to commercial systems.

selsinork wrote:

 

The inevitable conclusion to that becomes something like today's x86 PC where the smaller replaceable pieces are built to some set of agreed common interfaces,

No need to redefine the component interfaces.  There are enough standard interfaces like pinouts, connectors, internal and external buses, etc, that OSHW can happily pick whatever suits it best without needing any changes.  Using existing interfaces doesn't mean that you have to stick to existing system designs.

Not sure it's worked out well for the PC though.

The PC has been the greatest successful application of modular systems based on standard interfaces of the last several decades.  It is aversion to financial risk that has prevented companies from evolving the PC architecture, but OSHW doesn't have that problem as long as it actually (re)designs and evolves hardware systems rather than just consuming commercial ones.

Morgaine Dinova wrote:

 

PS. Guzunty Pi is a wonderful example of OSHW modularity encouraging easy add-on projects without major investment, notwithstanding that Pi itself is not OSHW.  Just imagine if the entire computer were made of such small replaceable pieces --- there would be continuous evolution in the community.

Guzunty Pi is wonderful because it uses the only CPLD I know of that's still available in a PLCC, which in turn has a pin grid array socket with 0.1" centers.  Every other FPGA/CPLD that's available is fine-pitch SMT, so prototyping with them is challenging.  Fortunately, a lot of reasonably-priced FPGA/CPLD modules with interfaces suitable for prototyping have come out, so the situation is improving.  If you have a bigger budget, you can get some very technically attractive PLCC units from HuMANDATA Ltd (Japan).