Are they even talking about lima? Or are they giving us the wool-over-eyes treatment?
https://www.kickstarter.com/projects/librecomputer/la-frite-open-source-frie...
Searching....
Nope, I can't find anything for or against lima.
Can anyone else find out?
I'm going to try and ask them directly, but evidence would be better.
Thanks!
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
On Wed, Oct 17, 2018 at 3:45 AM doark@mail.com wrote:
<Quote name=libre.computer>
$1,000,000 Goal Reward: Reverse Engineering and Upstreaming a Open Source Mali Utgard
The Mali-450 GPU currently uses a blob for X11/Wayland/DRM. This is less than ideal and it is one of the sticking points in terms of open source facing ARM adoption on multiple platforms. If we reach this amount, we commit the necessary resources to upstream Mali Utgard GPU family support in the Linux kernel and Mesa. This effort will take approximately a year.
We will announce more stretch goals as we move farther along the campaign.
</Quote>
Yes! But it is a BIG goal.
Thanks!
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On October 17, 2018 4:43:29 PM GMT+09:00, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
That would be a way better use of $1 million USD IMO. Would much rather have an open GPU than constantly be playing catchup reversing proprietary GPUs.
On Wed, Oct 17, 2018 at 8:49 AM Adam Van Ymeren adam@vany.ca wrote:
On October 17, 2018 4:43:29 PM GMT+09:00, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
That would be a way better use of $1 million USD IMO. Would much rather have an open GPU than constantly be playing catchup reversing proprietary GPUs.
some people would say it's better for users to at least have something. others would say that after reverse-engineering is completed, ODMs are likely (eventually) to actually distribute the reverse-engineered drivers.
ARM of course is quite likely to continue to release proprietary drivers, with newer proprietary versions of MALI never being compatible, thus always ensuring that end-users are left trapped.
l.
On Wed, 17 Oct 2018 09:04:54 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Wed, Oct 17, 2018 at 8:49 AM Adam Van Ymeren adam@vany.ca wrote:
On October 17, 2018 4:43:29 PM GMT+09:00, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
That would be a way better use of $1 million USD IMO. Would much rather have an open GPU than constantly be playing catchup reversing proprietary GPUs.
Ya, that's what I was thinking...
some people would say it's better for users to at least have something. others would say that after reverse-engineering is completed, ODMs are likely (eventually) to actually distribute the reverse-engineered drivers.
Again, my line of reasoning.
ARM of course is quite likely to continue to release proprietary drivers, with newer proprietary versions of MALI never being compatible, thus always ensuring that end-users are left trapped.
l.
We'll never know till we try :) I'm talking green branch and such.
Sincerely, David
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D graphics *without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
Seriously, when I get BIG money (which will be a while), I'm planning on calling you on this one and I don't want a tripled price tag.
Where is the extra 750K going to?
Thanks!
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Wed, Oct 17, 2018 at 2:58 PM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D graphics *without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
if only wanting to reach a target of say producing the design files (HDL), that's a different target entirely as well, and if someone else can take over at that point (and make the actual money selling product), then that's one way to ensure that the goal's reached [but not guarantee financial benefit from it].
so it was a throwaway comment written in about 30 seconds with no strict analysis done. last time i collaborated with an associate to come up with a proper figure it was around USD $5m (for a client) and it took us around five days to put that together.
however that particular deal excluded certain resources and had specific requirements.
Seriously, when I get BIG money (which will be a while),
cool
I'm planning on calling you on this one
excellent. well, when that happens, let's talk (on or off list), you let me know what you can raise and i'll put the time in - which will be several days not a few moments - to work out what can be done, and what side-deals will be needed, if necessary.
l.
On Wednesday 17. October 2018 16.30.17 Luke Kenneth Casson Leighton wrote:
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
Some interesting perspectives on this can be read about here:
https://chips4makers.io/blog/startup-costs-and-low-volume-manufacturing.html
The author is currently running a campaign on Crowd Supply:
https://www.crowdsupply.com/chips4makers/retro-uc
Maybe you might have spoken to the people involved, either at FOSDEM or via Crowd Supply, but I suppose there might be some potential for collaboration or discussion.
The campaign itself won't get funded, it would seem, but like various others that went the same way, this is more a consequence of a number of different factors rather than it being a statement on the merits of making genuinely free and open silicon. FPGAs and the implemented CPUs are widely available and already in use amongst the target audiences, for instance.
However, pursuing such a campaign with a product that really needs to be done with "proper" silicon - needing higher frequencies and power/performance benefits, perhaps - might get more support, albeit from not quite the same audiences as the ones targeted in this campaign.
Paul
P.S. Still no progress from Crowd Supply on the latest update?
On Wed, Oct 17, 2018 at 5:29 PM Paul Boddie paul@boddie.org.uk wrote:
On Wednesday 17. October 2018 16.30.17 Luke Kenneth Casson Leighton wrote:
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
Some interesting perspectives on this can be read about here:
https://chips4makers.io/blog/startup-costs-and-low-volume-manufacturing.html
yes, so, it actually costs, in the UK, about $600 per 180nm wafer. a full mask set however are still USD $250,000. so even 180nm is still mad.
However, pursuing such a campaign with a product that really needs to be done with "proper" silicon - needing higher frequencies and power/performance benefits, perhaps - might get more support, albeit from not quite the same audiences as the ones targeted in this campaign.
indeed, and that's great.
Paul
P.S. Still no progress from Crowd Supply on the latest update?
getting there
On Thursday 18. October 2018 06.11.42 Luke Kenneth Casson Leighton wrote:
On Wed, Oct 17, 2018 at 5:29 PM Paul Boddie paul@boddie.org.uk wrote:
P.S. Still no progress from Crowd Supply on the latest update?
getting there
Thanks for posting this, with it being published yesterday:
https://www.crowdsupply.com/eoma68/micro-desktop/updates/test-run-of-100-a20...
I am sorry to hear about the hassle with the contracting and the disappointing experience that seemed to result. It is unfortunate that organisations are often incapable of being persuaded to function in a way that would actually serve their best interests.
In my earlier consulting days, the mantra "appropriate quality" would be used as an equivalent to the rather familiar "don't spend too much time on it". What results from this is that people then don't do things properly, store up trouble for the future, and then hope that the customer will pay again to make things right. I guess it is an indication of how much of human society operates: put things off, collect the immediate rewards, and leave the mess for others to fix at their expense.
A note about Liberapay, though: it seems that they do support other payment providers now. If you can tolerate PayPal (I cannot) then they are supported. The other current option is Stripe.
In any case, I hope the test run works out and can pave the way for production and the release of more funds.
Paul
On Wed, Oct 31, 2018 at 3:32 PM Paul Boddie paul@boddie.org.uk wrote:
In any case, I hope the test run works out and can pave the way for production and the release of more funds.
appreciated, paul.
i am also grateful for luks efforts, read the update as soon as i got the notification email. It was a shame the company was a disappointment to work for. I’d say your quite a valuable person :).
i am happy to patently wait till the time comes for production. :)
Ever faithful.
Thank you
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Fri, Nov 2, 2018 at 12:33 AM Alexander Ross maillist_arm-netbook@aross.me wrote:
i am also grateful for luks efforts, read the update as soon as i got the notification email. It was a shame the company was a disappointment to work for. I’d say your quite a valuable person :).
i am happy to patently wait till the time comes for production. :)
thanks alexander.
On Wed, 17 Oct 2018 16:30:17 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Wed, Oct 17, 2018 at 2:58 PM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D graphics *without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
My statement was based on: 1M - (3 x 150K == 450K [for the HDL]) = 550K for the silicon, and that's huge! What is the cost here? Silicon as an element is inexpensive and even 32nm foundries should have their HW paid for by now. It's been what? 8 years since 32nm and the half node, 28nm started? And the big buyers (AMD, Intel, Nvidia) were on it for 3+ years.
It's not like an OSH project can purchase a license for closed source HW something-or-another libraries, which I've read as being the main expense from anandtech (No, I'm not in the field).
if only wanting to reach a target of say producing the design files (HDL), that's a different target entirely as well, and if someone else can take over at that point (and make the actual money selling product), then that's one way to ensure that the goal's reached [but not guarantee financial benefit from it].
I assumed at the time, you were talking HDL and with an FPGA for hardware backing with basic OpenGL, OpenCL, and 2D driver. No video encode/decode HW or Vulkan driver. Not mass produced, just a few prototypes. At about the same speed as a good Intel GPU, say a high end HD630 series. If you have a more concrete or different idea I'd be glad to hear it.
so it was a throwaway comment written in about 30 seconds with no strict analysis done.
Really? You sounded so serious! Contrary to popular belief, sound does travel across the Internet, it's just that no body has managed to tie the Internet into the speakers yet. ;-)
last time i collaborated with an associate to come up with a proper figure it was around USD $5m (for a client) and it took us around five days to put that together. however that particular deal excluded certain resources and had specific requirements.
:( I was aiming for 2M minimum. And that is BIG to me.
Seriously, when I get BIG money (which will be a while),
cool
Let me be frank, luke, I am under no delusions. No lottery. No gambling. No get rich quick schemes. But it will take time, as in ~5 years. Therefore, be excited about the OSH, not the money which may or may never be there.
I don't always agree with you, but what you write I do pay attention to. And I have a terrible tendency of believing you. :) I had assumed that some company would opensource or support the OSS Mali drivers, eventually. Google has android running off of Mali, Samsung and LG have Smart TVs running off of Mali, the list goes on... But when you wrote about creating an OSH GPU yourself I realized that is spite of every Mali customer, Google et. al. would not support opensourcing anything if they could help it. Chromium's binary blob problems are even further proof. This became painfully clear when RISC-V came out. I expected *at least* the cache to be OSH... :CRY:
I decided that if I could, it was up to me to pull in the funds for OSH, and "I think I can"(TM).
I'm planning on calling you on this one
excellent. well, when that happens, let's talk (on or off list), you let me know what you can raise and i'll put the time in - which will be several days not a few moments - to work out what can be done, and what side-deals will be needed, if necessary.
l.
I will tell you, but keep in mind that there are other OSH products to be developed that I have my eyes on.
Finally, I'm uncertain that I want to know, but what is a "side deal"?
Thanks!
On Wed, 17 Oct 2018 19:44:05 -0400 David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 16:30:17 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
<snip>
It's not like an OSH project can purchase a license for closed source HW something-or-another libraries, which I've read as being the main expense from anandtech (No, I'm not in the field).
if only wanting to reach a target of say producing the design files (HDL), that's a different target entirely as well, and if someone else can take over at that point (and make the actual money selling product), then that's one way to ensure that the goal's reached [but not guarantee financial benefit from it].
I assumed at the time, you were talking HDL and with an FPGA for hardware backing with basic OpenGL, OpenCL, and 2D driver. No video encode/decode HW or Vulkan driver. Not mass produced, just a few prototypes. At about the same speed as a good Intel GPU, say a high end HD630 series. If you have a more concrete or different idea I'd be glad to hear it.
<snip>
But can it run Crysis? :-) I *had* to say it!
PS: luke, I aught to have mentioned in my previous email that an Intel high end HD630 series is 1/3rd as fast as an rx550 and Nvidia 1050 ti, it is also 1/2 as fast as the 2200G AMD Raven Ridge APU.
Happy returns, David
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Thu, Oct 18, 2018 at 12:44 AM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 16:30:17 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Wed, Oct 17, 2018 at 2:58 PM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D graphics *without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
My statement was based on: 1M - (3 x 150K == 450K [for the HDL]) = 550K for the silicon, and that's huge!
relatively speaking it's tiny.
What is the cost here?
it was a throwaway comment, david, written in about 30 seconds, with about 5 seconds thought. the focus was more "a much higher bang per buck can be achieved with that kind of money" than "here is a detailed statement of work".
it would take me a *lot* more time to specifically answer that very general question without specific information.
Silicon as an element is inexpensive and even 32nm foundries should have their HW paid for by now. It's been what? 8 years since 32nm and the half node, 28nm started? And the big buyers (AMD, Intel, Nvidia) were on it for 3+ years.
28nm mask charges - the lithographic stencils onto glass - for 28nm are i think USD $2m. it's an exponential curve. 45nm is $1m. it drops to around $250,000 for 180nm.
It's not like an OSH project can purchase a license for closed source HW something-or-another libraries,
the industry term is "HDL" - i think it's just "hardware definition language" or something.
which I've read as being the main expense from anandtech (No, I'm not in the field).
so that's why i spent several months tracking down libre equivalents, and/or tracking down people in open hardware willing to *create* libre-licensed HDL
if only wanting to reach a target of say producing the design files (HDL), that's a different target entirely as well, and if someone else can take over at that point (and make the actual money selling product), then that's one way to ensure that the goal's reached [but not guarantee financial benefit from it].
I assumed at the time, you were talking HDL and with an FPGA for hardware backing with basic OpenGL, OpenCL, and 2D driver. No video encode/decode HW or Vulkan driver. Not mass produced, just a few prototypes. At about the same speed as a good Intel GPU, say a high end HD630 series. If you have a more concrete or different idea I'd be glad to hear it.
so it was a throwaway comment written in about 30 seconds with no strict analysis done.
Really? You sounded so serious! Contrary to popular belief, sound does travel across the Internet, it's just that no body has managed to tie the Internet into the speakers yet. ;-)
:)
last time i collaborated with an associate to come up with a proper figure it was around USD $5m (for a client) and it took us around five days to put that together. however that particular deal excluded certain resources and had specific requirements.
:( I was aiming for 2M minimum. And that is BIG to me.
if aiming for 28nm (without additional investors) that would barely cover the production mask charges. hence why i am interested to work with IIT Madras.
if requiring for example a modern DDR4 memory controller, over 50% of that budget would be taken up. hence why i am interested in HyperRAM (upcoming JEDEC xSPI).
if not using a back-end team such as the people that IIT Madras have access to, that USD $2m would be entirely eaten by proprietary layout tool licensing from Mentor Graphics and the engineers who would need to be hired to do the work. hence why i am interested in Magic, alliance2 / coriolis, and libresilicon, all of which are developing open ASIC layout tools.
there is a *lot* i simply have not had time to talk about, here, david.
Seriously, when I get BIG money (which will be a while),
cool
Let me be frank, luke, I am under no delusions. No lottery. No gambling. No get rich quick schemes. But it will take time, as in ~5 years. Therefore, be excited about the OSH, not the money which may or may never be there.
I don't always agree with you, but what you write I do pay attention to. And I have a terrible tendency of believing you. :)
:) always never do that: i am first and foremost an ethical person, who goes, "that's not good, let's fix that", has absolutely *no idea* how to go about "fixing that", and persistently chips away (often randomly) regardless of complete lack of knowledge and expertise, until success.
it means that i really *really* need feedback and people to double-check and triple-check.
I had assumed that some company would opensource or support the OSS Mali drivers, eventually. Google has android running off of Mali, Samsung and LG have Smart TVs running off of Mali, the list goes on... But when you wrote about creating an OSH GPU yourself I realized that is spite of every Mali customer, Google et. al. would not support opensourcing anything if they could help it.
YA THINK???
Chromium's binary blob problems are even further proof. This became painfully clear when RISC-V came out. I expected *at least* the cache to be OSH... :CRY:
I decided that if I could, it was up to me to pull in the funds for OSH, and "I think I can"(TM).
awesome. well, as usual i have a parallel set of tracks being investigated, as part of a wider strategy of deals and collaborations with various people, just bear in mind that your help would be part of a much larger deal, ok?
I'm planning on calling you on this one
excellent. well, when that happens, let's talk (on or off list), you let me know what you can raise and i'll put the time in - which will be several days not a few moments - to work out what can be done, and what side-deals will be needed, if necessary.
l.
I will tell you, but keep in mind that there are other OSH products to be developed that I have my eyes on.
Finally, I'm uncertain that I want to know, but what is a "side deal"?
ok, so let's say we make a chip. it's successfully made, it's all done, and it works, right?
and then we go, "ok, who wanna buy?"
and... total absolute silence.
egg.
on.
face.
so the way to deal with that is: find **MULTIPLE** potential customers, take their requirements into account, and make damn sure that the SoC is capable of fulfilling THEIR needs times as many of such potential customers as can possibly be found.
and if even 30% of them come through and order 100,000, that's enough to actually get a serious order with a foundry.
now, some of those people might be willing to actually invest up-front, to get the SoC made.
that's what i mean by "side deals". deals "on the side" that result in everybody succeeding in getting what they want.
l.
There are dozens of arm sbs right now that are crippled mainly due to Mali graphics. A lot of them would be very useful for stuff that e.g. the rpi doesn't do like cheap nas. Just saying.
On Thu, Oct 18, 2018, 08:32 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Thu, Oct 18, 2018 at 12:44 AM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 16:30:17 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware:
https://www.crowdsupply.com/eoma68
On Wed, Oct 17, 2018 at 2:58 PM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be
damn
good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D
graphics
*without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
My statement was based on: 1M - (3 x 150K == 450K [for the HDL]) = 550K for the silicon, and that's huge!
relatively speaking it's tiny.
What is the cost here?
it was a throwaway comment, david, written in about 30 seconds, with about 5 seconds thought. the focus was more "a much higher bang per buck can be achieved with that kind of money" than "here is a detailed statement of work".
it would take me a *lot* more time to specifically answer that very general question without specific information.
Silicon as an element is inexpensive and even 32nm foundries should have their HW paid for by now. It's been what? 8 years since 32nm and the half node, 28nm started? And the big buyers (AMD, Intel, Nvidia) were on it for 3+ years.
28nm mask charges - the lithographic stencils onto glass - for 28nm are i think USD $2m. it's an exponential curve. 45nm is $1m. it drops to around $250,000 for 180nm.
It's not like an OSH project can purchase a license for closed source HW something-or-another libraries,
the industry term is "HDL" - i think it's just "hardware definition language" or something.
which I've read as being the main expense from anandtech (No, I'm not in the field).
so that's why i spent several months tracking down libre equivalents, and/or tracking down people in open hardware willing to *create* libre-licensed HDL
if only wanting to reach a target of say producing the design files (HDL), that's a different target entirely as well, and if someone else can take over at that point (and make the actual money selling product), then that's one way to ensure that the goal's reached [but not guarantee financial benefit from it].
I assumed at the time, you were talking HDL and with an FPGA for hardware backing with basic OpenGL, OpenCL, and 2D driver. No video encode/decode HW or Vulkan driver. Not mass produced, just a few prototypes. At about the same speed as a good Intel GPU, say a high end HD630 series. If you have a more concrete or different idea I'd be glad to hear it.
so it was a throwaway comment written in about 30 seconds with no strict analysis done.
Really? You sounded so serious! Contrary to popular belief, sound does travel across the Internet, it's just that no body has managed to tie the Internet into the speakers yet. ;-)
:)
last time i collaborated with an associate to come up with a proper figure it was around USD $5m (for a client) and it took us around five days to put that together. however that particular deal excluded certain resources and had specific requirements.
:( I was aiming for 2M minimum. And that is BIG to me.
if aiming for 28nm (without additional investors) that would barely cover the production mask charges. hence why i am interested to work with IIT Madras.
if requiring for example a modern DDR4 memory controller, over 50% of that budget would be taken up. hence why i am interested in HyperRAM (upcoming JEDEC xSPI).
if not using a back-end team such as the people that IIT Madras have access to, that USD $2m would be entirely eaten by proprietary layout tool licensing from Mentor Graphics and the engineers who would need to be hired to do the work. hence why i am interested in Magic, alliance2 / coriolis, and libresilicon, all of which are developing open ASIC layout tools.
there is a *lot* i simply have not had time to talk about, here, david.
Seriously, when I get BIG money (which will be a while),
cool
Let me be frank, luke, I am under no delusions. No lottery. No gambling. No get rich quick schemes. But it will take time, as in ~5 years. Therefore, be excited about the OSH, not the money which may or may never be there.
I don't always agree with you, but what you write I do pay attention to. And I have a terrible tendency of believing you. :)
:) always never do that: i am first and foremost an ethical person, who goes, "that's not good, let's fix that", has absolutely *no idea* how to go about "fixing that", and persistently chips away (often randomly) regardless of complete lack of knowledge and expertise, until success.
it means that i really *really* need feedback and people to double-check and triple-check.
I had assumed that some company would opensource or support the OSS Mali drivers, eventually. Google has android running off of Mali, Samsung and LG have Smart TVs running off of Mali, the list goes on... But when you wrote about creating an OSH GPU yourself I realized that is spite of every Mali customer, Google et. al. would not support opensourcing anything if they could help it.
YA THINK???
Chromium's binary blob problems are even further proof. This became painfully clear when RISC-V came out. I expected *at least* the cache to be OSH... :CRY:
I decided that if I could, it was up to me to pull in the funds for OSH, and "I think I can"(TM).
awesome. well, as usual i have a parallel set of tracks being investigated, as part of a wider strategy of deals and collaborations with various people, just bear in mind that your help would be part of a much larger deal, ok?
I'm planning on calling you on this one
excellent. well, when that happens, let's talk (on or off list), you let me know what you can raise and i'll put the time in - which will be several days not a few moments - to work out what can be done, and what side-deals will be needed, if necessary.
l.
I will tell you, but keep in mind that there are other OSH products to be developed that I have my eyes on.
Finally, I'm uncertain that I want to know, but what is a "side deal"?
ok, so let's say we make a chip. it's successfully made, it's all done, and it works, right?
and then we go, "ok, who wanna buy?"
and... total absolute silence.
egg.
on.
face.
so the way to deal with that is: find **MULTIPLE** potential customers, take their requirements into account, and make damn sure that the SoC is capable of fulfilling THEIR needs times as many of such potential customers as can possibly be found.
and if even 30% of them come through and order 100,000, that's enough to actually get a serious order with a foundry.
now, some of those people might be willing to actually invest up-front, to get the SoC made.
that's what i mean by "side deals". deals "on the side" that result in everybody succeeding in getting what they want.
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Thu, Oct 18, 2018 at 6:46 AM Bill Kontos vkontogpls@gmail.com wrote:
There are dozens of arm sbs right now that are crippled mainly due to Mali graphics. A lot of them would be very useful for stuff that e.g. the rpi doesn't do like cheap nas. Just saying.
yes, i agree. the thing is... *sigh*... percentage-wise, the people you're referring to (those that would actually put arm SBCs to different purposes) are an absolutely tiny fraction of the overall market. an important *strategic* fraction, nonetheless.
the lesson there is one from HTC's products absolutely tanking after they neglected to listen to a tiny vocal minority on strategic (high-prominence) forums like xda-developers.com. when going to "buy product", anyone *not* technical of course does a google search. if the high-ranking google pages, not controlled by HTC, are filled with complaints "this product is shit! it doesn't have a keyboard like our favourite predecessor does!!!" then the non-techies aren't gonna buy...
and it's the same here. so... yes, reverse-engineering MALI will have a much larger strategic impact on purchases of ARM-based products than people realise...
and that's what deeply concerns me: by raising USD $1m, this campaign is empowering ARM to continue in their highly and deeply unethical behaviour, such as fucking with libv's livelihood.
l.
On Thu, 18 Oct 2018 06:52:20 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Thu, Oct 18, 2018 at 6:46 AM Bill Kontos vkontogpls@gmail.com wrote:
There are dozens of arm sbs right now that are crippled mainly due to Mali graphics. A lot of them would be very useful for stuff that e.g. the rpi doesn't do like cheap nas. Just saying.
yes, i agree. the thing is... *sigh*... percentage-wise, the people you're referring to (those that would actually put arm SBCs to different purposes) are an absolutely tiny fraction of the overall market. an important *strategic* fraction, nonetheless.
the lesson there is one from HTC's products absolutely tanking after they neglected to listen to a tiny vocal minority on strategic (high-prominence) forums like xda-developers.com. when going to "buy product", anyone *not* technical of course does a google search. if the high-ranking google pages, not controlled by HTC, are filled with complaints "this product is shit! it doesn't have a keyboard like our favourite predecessor does!!!" then the non-techies aren't gonna buy...
and it's the same here. so... yes, reverse-engineering MALI will have a much larger strategic impact on purchases of ARM-based products than people realise...
and that's what deeply concerns me: by raising USD $1m, this campaign is empowering ARM to continue in their highly and deeply unethical behaviour, such as fucking with libv's livelihood.
The light dawns in my head!!!
Thanks luke, that *really* explained your objections to an OSS mali driver!
On Thu, 18 Oct 2018 06:30:13 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Thu, Oct 18, 2018 at 12:44 AM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 16:30:17 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Wed, Oct 17, 2018 at 2:58 PM David Niklas doark@mail.com wrote:
On Wed, 17 Oct 2018 08:43:29 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
USD $1m would fund the Libre RISC-V 3D GPU effort through to 100% completion including quite likely actual test silicon in a smaller geometry like 65 / 55nm (700mhz or so, which for a GPU would be damn good).
l.
Wait a sec! Last year you said:
On Mon, 8 May 2017 16:43:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, May 8, 2017 at 4:27 PM, ronwirring@safe-mail.net wrote:
Best knowledge is, that new intel and amd processors cannot be reverse engineered. What in regard of the latest mali gpus? If you have the money, they can be reverse engineered?
yes. about $150k would do it. but the question is, really: what would happen if you did? and, what else could you do with the same money?
well, with the same money it would be possible to make our own libre processor, with enough extensions to be able to do 3D graphics *without* paying anyone a cent. any company tries to claim patent royalties, all that happens is a search is made on their "claims", for anything similar that has prior art.
if it's another company, guess what? we notify that other company and watch the fireworks...
l.
1M - 150k = 750K!!! Therefore you're tripling your price estimate on us!
it depends on what you take into account. if someone else pays the NREs to the foundry, i.e. a university agrees to collaborate and is offered access to a foundry for either free or at reduced rates, $250k-$500k comes off that amount, straight away.
My statement was based on: 1M - (3 x 150K == 450K [for the HDL]) = 550K for the silicon, and that's huge!
relatively speaking it's tiny.
That's the equivalent of a good house over here *totally paid for*!
What is the cost here?
it was a throwaway comment, david, written in about 30 seconds, with about 5 seconds thought. the focus was more "a much higher bang per buck can be achieved with that kind of money" than "here is a detailed statement of work".
it would take me a *lot* more time to specifically answer that very general question without specific information.
Silicon as an element is inexpensive and even 32nm foundries should have their HW paid for by now. It's been what? 8 years since 32nm and the half node, 28nm started? And the big buyers (AMD, Intel, Nvidia) were on it for 3+ years.
28nm mask charges - the lithographic stencils onto glass - for 28nm are i think USD $2m. it's an exponential curve. 45nm is $1m. it drops to around $250,000 for 180nm.
I had imagined that with at least 4 people (by my memory and including me), offering you funding (in the fullness of time), you might have taken a cursory look into the matter.
<snip>
last time i collaborated with an associate to come up with a proper figure it was around USD $5m (for a client) and it took us around five days to put that together. however that particular deal excluded certain resources and had specific requirements.
:( I was aiming for 2M minimum. And that is BIG to me.
If aiming for 28nm (without additional investors) that would barely cover the production mask charges. hence why i am interested to work with IIT Madras.
If requiring, for example, a modern DDR4 memory controller, over 50% of that budget would be taken up. Hence why I am interested in HyperRAM (upcoming JEDEC xSPI).
URL? I've only heard of the upcoming DDR5 non-volatile memory.
If not using a back-end team such as the people that IIT Madras have access to, that USD $2m would be entirely eaten by proprietary layout tool licensing from Mentor Graphics and the engineers who would need to be hired to do the work. hence why I am interested in Magic, alliance2 / coriolis, and libresilicon, all of which are developing open ASIC layout tools.
There is a *lot* i simply have not had time to talk about, here, david.
Open source HW is going to require lots of talking, luke, esp. as each part of the work gets closer to fruition. I don't deny your wisdom in discerning your own path, but please consider at least attempting to take on a liaison/spokesmen if you cannot keep up.
Remember The Foul Mouth of Sauron from LOTR? :)
<snip>
I don't always agree with you, but what you write I do pay attention to. And I have a terrible tendency of believing you. :)
:) always never do that:
Double self canceling absolute. Grammar error. SEGFAULT. Core dumped. :)
i am first and foremost an ethical person,
Me too(TM).
who goes, "that's not good, let's fix that", has absolutely *no idea* how to go about "fixing that", and persistently chips away (often randomly) regardless of complete lack of knowledge and expertise, until success.
That is *amazing*! You really sound like you know what you're doing.
it means that i really *really* need feedback and people to double-check and triple-check.
Don't look at me, I'm almost the same way. :)
I had assumed that some company would opensource or support the OSS Mali drivers, eventually. Google has android running off of Mali, Samsung and LG have Smart TVs running off of Mali, the list goes on... But when you wrote about creating an OSH GPU yourself I realized that is spite of every Mali customer, Google et. al. would not support opensourcing anything if they could help it.
YA THINK???
This is an important difference between us, I see people as having a choice, where any day the CEOs of these companies could wake up and say, "We have been using FLOSS for years and have not given a fraction of what we have made back; I shall be the opensource monetary sponsor and pathfinder for their endeavors."
Or, how about using this line of reasoning, "We treat the FLOSS devs as expendable laborers. Like of old, in Egypt, we are their cruel taskmasters, our thirst for their sweat and blood never satisfied. Truly, they donate their lives to the cause, and we have only beads and trinkets to offer them in return. Indeed, I shall reward their efforts, and see that every devotee receives his due."
Stallmen and ESR both thought that FLOSS just needed to prove itself to catch on. I bought into that strategy. I gave it up after 6 years. There is a lot more morals involved than that. Even AMD's FLOSS diver is more about the economics than anything else.
Did I mention that mixed source was returning under the guise of FLOSS through chromium/android et. al.?
We need a leader for the OSH, someone who will stand up to the vendor lock in/greedy bullies of out time. I thought RMS, then luke would be it, but not so. You're a great person, luke, but you don't seem the type.
Chromium's binary blob problems are even further proof. This became painfully clear when RISC-V came out. I expected *at least* the cache to be OSH... :CRY:
I decided that if I could, it was up to me to pull in the funds for OSH, and "I think I can"(TM).
awesome. well, as usual i have a parallel set of tracks being investigated, as part of a wider strategy of deals and collaborations with various people, just bear in mind that your help would be part of a much larger deal, ok?
Ok, but I must confess I have been thinking and planning this for a long time and have quite a few good ideas about the implementation.
I also have a thick head and no idea what I'm talking about. :)
I'm planning on calling you on this one
excellent. well, when that happens, let's talk (on or off list), you let me know what you can raise and i'll put the time in - which will be several days not a few moments - to work out what can be done, and what side-deals will be needed, if necessary.
l.
I will tell you, but keep in mind that there are other OSH products to be developed that I have my eyes on.
Finally, I'm uncertain that I want to know, but what is a "side deal"?
ok, so let's say we make a chip. it's successfully made, it's all done, and it works, right?
and then we go, "ok, who wanna buy?"
and... total absolute silence.
egg.
on.
face.
Eventually we will need normal people to be our user base. I know a bit about marketing, but estimating demand or creating a sense of needing non-propriety HW? *I have totally no idea*.
At all.
Ponder that for me luke, it is the one mystery that *totally evades* me.
The best I can do where I live is get blank looks and, "I want my icons!" and "What I have now works." statements. Really!? That is all?! You, the end user could have SO MUCH MORE!
They simply don't know what wonders and powers they're missing out on.
And how badly they are used and abused.
And then stuck through the meat grinder again and again.
They say/act like, "We owe [insert favorite company/politician] everything." "Yes," I respond, "that is what they want, *Everything*!"
so the way to deal with that is: find **MULTIPLE** potential customers, take their requirements into account, and make damn sure that the SoC is capable of fulfilling THEIR needs times as many of such potential customers as can possibly be found.
and if even 30% of them come through and order 100,000, that's enough to actually get a serious order with a foundry.
now, some of those people might be willing to actually invest up-front, to get the SoC made.
that's what i mean by "side deals". deals "on the side" that result in everybody succeeding in getting what they want.
l.
That's good. I thought you might have us eternally relying on a college or some other deal where a slip of some money from [insert enemy company] would lead to the projects all going down the drain. And yes, if we *really* succeed the competition will band together and do that to us. If not the 3 letter US government agencies.
Thanks
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Fri, Oct 19, 2018 at 4:30 AM David Niklas doark@mail.com wrote:
relatively speaking it's tiny.
That's the equivalent of a good house over here *totally paid for*!
and about 10-20 homes in say scotland... yes.
the difficulty level and risks associated with silicon are so high that the profits to be made are as equally enormous.
What is the cost here?
it was a throwaway comment, david, written in about 30 seconds, with about 5 seconds thought. the focus was more "a much higher bang per buck can be achieved with that kind of money" than "here is a detailed statement of work".
it would take me a *lot* more time to specifically answer that very general question without specific information.
Silicon as an element is inexpensive and even 32nm foundries should have their HW paid for by now. It's been what? 8 years since 32nm and the half node, 28nm started? And the big buyers (AMD, Intel, Nvidia) were on it for 3+ years.
28nm mask charges - the lithographic stencils onto glass - for 28nm are i think USD $2m. it's an exponential curve. 45nm is $1m. it drops to around $250,000 for 180nm.
I had imagined that with at least 4 people (by my memory and including me), offering you funding (in the fullness of time), you might have taken a cursory look into the matter.
indeed. "paying some engineers time to get the libre-licensed HDL developed", i have a vaaaague feeling that that was the focus of our discussions some months back.
If requiring, for example, a modern DDR4 memory controller, over 50% of that budget would be taken up. Hence why I am interested in HyperRAM (upcoming JEDEC xSPI).
URL?
google it, as that's exactly and precisely what i would do anyway. only cursory information is available as JEDEC runs on an ITU-style (closed, proprietary) basis.
HyperRAM on the other hand is extremely common. it's basically Quad SPI extended to 8 bit and DDR.
If not using a back-end team such as the people that IIT Madras have access to, that USD $2m would be entirely eaten by proprietary layout tool licensing from Mentor Graphics and the engineers who would need to be hired to do the work. hence why I am interested in Magic, alliance2 / coriolis, and libresilicon, all of which are developing open ASIC layout tools.
There is a *lot* i simply have not had time to talk about, here, david.
Open source HW is going to require lots of talking, luke, esp. as each part of the work gets closer to fruition.
yup.
I don't deny your wisdom in discerning your own path, but please consider at least attempting to take on a liaison/spokesmen if you cannot keep up.
if there's funds available to pay them... of course.
who goes, "that's not good, let's fix that", has absolutely *no idea* how to go about "fixing that", and persistently chips away (often randomly) regardless of complete lack of knowledge and expertise, until success.
That is *amazing*! You really sound like you know what you're doing.
i really don't: i just have a fast enough corrective loop that it may *look* that way. there are massive holes however.
Did I mention that mixed source was returning under the guise of FLOSS through chromium/android et. al.?
don't start. i'm keenly aware of the damage that google has done by using the apache2 license.
We need a leader for the OSH, someone who will stand up to the vendor lock in/greedy bullies of out time. I thought RMS, then luke would be it, but not so. You're a great person, luke, but you don't seem the type.
i'm tackling it differently by going further and further up the hardware chain. *that* is physical items that are required to be bought. that's where software cannot be "controlled", if you will. anyone can download libre-licensed source and completely ignore their ethical and moral obligation to fund the developers who created it. that *cannot* be done with hardware.
Chromium's binary blob problems are even further proof. This became painfully clear when RISC-V came out. I expected *at least* the cache to be OSH... :CRY:
I decided that if I could, it was up to me to pull in the funds for OSH, and "I think I can"(TM).
awesome. well, as usual i have a parallel set of tracks being investigated, as part of a wider strategy of deals and collaborations with various people, just bear in mind that your help would be part of a much larger deal, ok?
Ok, but I must confess I have been thinking and planning this for a long time and have quite a few good ideas about the implementation.
cool.
ok, so let's say we make a chip. it's successfully made, it's all done, and it works, right?
and then we go, "ok, who wanna buy?"
and... total absolute silence.
egg.
on.
face.
Eventually we will need normal people to be our user base. I know a bit about marketing, but estimating demand or creating a sense of needing non-propriety HW? *I have totally no idea*.
i have a potential client who will order 100,000 units in the first year, if certain power and functionaliy requirements are met. it's a self-contained market for their product so the fact that it's RISC-V is completely irrelevant to them.
then there is the india smartphone / netbook / chromebook market.
That's good. I thought you might have us eternally relying on a college or some other deal where a slip of some money from [insert enemy company] would lead to the projects all going down the drain.
nope. flexibility and multi-pronged strategy.
arm-netbook@lists.phcomp.co.uk
-
Adam Van Ymeren -
Alexander Ross -
Bill Kontos -
David Niklas -
Luke Kenneth Casson Leighton -
Paul Boddie