-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
Dear luke, list,
As both luke and I are interested in chip fabrication I'm dropping this email with a link to a story on a SW tool that is designed to help improve designs based on how exactly a foundries process works.
Luke, the wording in this article looks very carefully chosen in places. I'm totally in the dark as to why; e.g. "...the platform that analyzes yields enables secure collaboration between the foundry and customer."
It's not free software so it might be interesting to try to find something similar offered by the community. Or maybe we'll have to roll our own eventually.
https://www.anandtech.com/show/14613/synopsys-to-accelerate-samsungs-7nm-ram...
David
On Saturday, July 6, 2019, David Niklas doark@mail.com wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
Dear luke, list,
As both luke and I are interested in chip fabrication I'm dropping this email with a link to a story on a SW tool that is designed to help improve designs based on how exactly a foundries process works.
Luke, the wording in this article looks very carefully chosen in places. I'm totally in the dark as to why; e.g. "...the platform that analyzes yields enables secure collaboration between the foundry and customer."
The Foundries absolutely flatly refuse to trust their customers.
They flatly refuse to provide them with the "cells" for even basic things like "a transiator".
So the tools basically have to create a GDS file with "holes" in them, where things like "transistor" or "memory cell" or "IO Pad" go.
The fuckers won't even give information about the *size* of those "holes", or how to connect to them, without an NDA.
However from what that article is *not* saying, even the "usual" NDA method is just not enough, at the lower geometries. For 20nm and bigger, things like MOSIS "rules" are perfectly sufficient. Lay out a design, no really special knowledge that hasn't been known (realistically) for 20+ years, no problem.
10nm and below is a WHOLE new level of weird.
It looks like there are some quantum interferences as well as EM and RF issues, *and* probably some power and layout issues in the tinier geometries, all of which the Foundries absolutely do not want the customers to know about, because it constitues "reverse engineerable knowledge" about how the Foundry lays out the chips, and a competitor Foundry could get hold of that and start their own multi billion dollar money spinner.
So they did a deal with Synopsis, where they would tell *them* how to avoid those pitfalls, as long as Synopsis promised to hide the information in such a way that, whilst the customer got the layout advice they need to get a working ASIC, they would in no way be able to know *how* that ASIC actually got manufactured.
This is of course all inferred guesswork. Welcome to my world of low probability logical deduction aka reverse engineering.
Bottom line is, we're literally decades and hundreds of millions of dollars away from libre foundries. I am probably out on those estimates by 1 to 2 orders of magnitude. Luckily, DARPA recognises the problem and put up USD 150m to create fully libre automated ASIC layout software. It's a start.
In the meantime I am tracking what lip6.fr are doing (the team behind alliance / coriolis2). They will be doing a 40nm tapeout using FreePDK, ported to their cell library system.
Once successful it will be possible to follow in their footsteps and do a 40nm Libre RISCV tapeout.
The layout software produces *linearly* scaleable designs, so hypothetically, as long as RF EM is ok (on chip) scaling to 28nm should be feasible. Just not straight away.
L.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Fri, 5 Jul 2019 19:42:46 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Saturday, July 6, 2019, David Niklas doark@mail.com wrote:
Dear luke, list,
As both luke and I are interested in chip fabrication I'm dropping this email with a link to a story on a SW tool that is designed to help improve designs based on how exactly a foundries process works.
Luke, the wording in this article looks very carefully chosen in places. I'm totally in the dark as to why; e.g. "...the platform that analyzes yields enables secure collaboration between the foundry and customer."
The Foundries absolutely flatly refuse to trust their customers.
They flatly refuse to provide them with the "cells" for even basic things like "a transiator".
So the tools basically have to create a GDS file with "holes" in them, where things like "transistor" or "memory cell" or "IO Pad" go.
The f*ckers won't even give information about the *size* of those "holes", or how to connect to them, without an NDA.
However from what that article is *not* saying, even the "usual" NDA method is just not enough, at the lower geometries. For 20nm and bigger, things like MOSIS "rules" are perfectly sufficient. Lay out a design, no really special knowledge that hasn't been known (realistically) for 20+ years, no problem.
10nm and below is a WHOLE new level of weird.
It looks like there are some quantum interferences as well as EM and RF issues, *and* probably some power and layout issues in the tinier geometries, all of which the Foundries absolutely do not want the customers to know about, because it constitues "reverse engineerable knowledge" about how the Foundry lays out the chips, and a competitor Foundry could get hold of that and start their own multi billion dollar money spinner.
Here's where closed source IP really confuses me: If a "money spinner" tried to do that wouldn't they be sued, pay royalties and regret it for the rest of their existences? Therefore, this "they understand the science and specs of our implementation," paranoia is unfounded, right?
So they did a deal with Synopsis, where they would tell *them* how to avoid those pitfalls, as long as Synopsis promised to hide the information in such a way that, whilst the customer got the layout advice they need to get a working ASIC, they would in no way be able to know *how* that ASIC actually got manufactured.
This is of course all inferred guesswork. Welcome to my world of low probability logical deduction aka reverse engineering.
Bottom line is, we're literally decades and hundreds of millions of dollars away from libre foundries. I am probably out on those estimates by 1 to 2 orders of magnitude.
Are we talking any libre foundry, or some particular nm size (not that a nm is actually used to describe a nm anymore)?
Luckily, DARPA recognises the problem and put up USD 150m to create fully libre automated ASIC layout software. It's a start.
Interesting. For posterity, here's a link (with HTML garbage removed): https://www.fbo.gov/index?id=a32e37cfad63edcba7cfd5d997422d93
Though it looks to be pre-alpha.
In the meantime I am tracking what lip6.fr are doing (the team behind alliance / coriolis2). They will be doing a 40nm tapeout using FreePDK, ported to their cell library system.
I was able to locate the coriolis2 docs here: https://www-soc.lip6.fr/sesi-docs/coriolis2-docs/coriolis2/en/html/users-gui... and the alliance project has a web page with installation instructions here: https://www-soc.lip6.fr/equipe-cian/logiciels/alliance/
It's all very interesting.
Once successful it will be possible to follow in their footsteps and do a 40nm Libre RISCV tapeout.
The layout software produces *linearly* scaleable designs, so hypothetically, as long as RF EM is ok (on chip) scaling to 28nm should be feasible. Just not straight away.
L.
Yay!
Thanks for the info, David
On Sat, Jul 6, 2019 at 3:49 AM David Niklas doark@mail.com wrote:
It looks like there are some quantum interferences as well as EM and RF issues, *and* probably some power and layout issues in the tinier geometries, all of which the Foundries absolutely do not want the customers to know about, because it constitues "reverse engineerable knowledge" about how the Foundry lays out the chips, and a competitor Foundry could get hold of that and start their own multi billion dollar money spinner.
Here's where closed source IP really confuses me: If a "money spinner" tried to do that wouldn't they be sued, pay royalties and regret it for the rest of their existences?
in the meantime, whilst such a court case is underway, they're losing literally billions due to the upstart having "stolen" their knowledge. a CSMC employee actually did that to TSMC: worked on TSMC's 28nm line, went back to China and started CSMC's 28nm line with the knowledge.
it took years for the court case to go through, as it's in a different international jurisdiction.
Bottom line is, we're literally decades and hundreds of millions of dollars away from libre foundries. I am probably out on those estimates by 1 to 2 orders of magnitude.
Are we talking any libre foundry, or some particular nm size (not that a nm is actually used to describe a nm anymore)?
outside of my ability to say.
Luckily, DARPA recognises the problem and put up USD 150m to create fully libre automated ASIC layout software. It's a start.
Interesting. For posterity, here's a link (with HTML garbage removed): https://www.fbo.gov/index?id=a32e37cfad63edcba7cfd5d997422d93
that is (was) a session link, now invalid. what keywords did you use?
l.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Sat, 6 Jul 2019 07:47:07 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Sat, Jul 6, 2019 at 3:49 AM David Niklas doark@mail.com wrote:
It looks like there are some quantum interferences as well as EM and RF issues, *and* probably some power and layout issues in the tinier geometries, all of which the Foundries absolutely do not want the customers to know about, because it constitues "reverse engineerable knowledge" about how the Foundry lays out the chips, and a competitor Foundry could get hold of that and start their own multi billion dollar money spinner.
Here's where closed source IP really confuses me: If a "money spinner" tried to do that wouldn't they be sued, pay royalties and regret it for the rest of their existences?
in the meantime, whilst such a court case is underway, they're losing literally billions due to the upstart having "stolen" their knowledge. a CSMC employee actually did that to TSMC: worked on TSMC's 28nm line, went back to China and started CSMC's 28nm line with the knowledge.
It took years for the court case to go through, as it's in a different international jurisdiction.
Ah, here the story is: https://web.archive.org/web/20150225094201/http://english.cw.com.tw/article....
It seems to me that there will always be thieves in the world. Hiding science does more harm to the scientific community then the thieves. Think long term though, luke, as chip fabrication plants are prone to do, at least for process design and improvement. If I'm a priest and I'll give away copies of the Bible and your a chip fabrication guru and you don't ever tell anyone anything about chip fabrication what will the population learn about?! Isn't it obvious that science will ultimately die off or become some super diluted idiocy (old wives tales)?! If people are willing to die for God(s) (real or not), but not even lift a finger, but instead purposefully restrict knowledge and understanding, to teach their fellow man the sciences will the sciences not be out competed many to one?! Am I the only one who sees this (our/your/the chip fabs), fate?
I digress, this cannot be my life. It's fundamentally not for me. I'm a fool for wanting to learn -- anything. I'm wedded to suicide. I'd be equally productive considering the long term, in watching the clouds go by. I'm going to do things differently... much, much differently, if I have any say so at all in the matter!
Bottom line is, we're literally decades and hundreds of millions of dollars away from libre foundries. I am probably out on those estimates by 1 to 2 orders of magnitude.
Are we talking any libre foundry, or some particular nm size (not that a nm is actually used to describe a nm anymore)?
Outside of my ability to say.
You know what happens when you can't answer one of my questions, right? :) tap tap tap tap ...... tap tap tap ... Aha! How's this, a 28/22nm chip fabrication plant for $4.2 billion? Not that GlobalFounderies ever had a winning process... https://www.anandtech.com/show/2814/3 And, if you read further down you can see that you can get a smaller plant with a College campus for $4 billion.
I guess this means to buy one we'll have to empty our piggy banks. ;-)
Luckily, DARPA recognises the problem and put up USD 150m to create fully libre automated ASIC layout software. It's a start.
Interesting. For posterity, here's a link (with HTML garbage removed): https://www.fbo.gov/index?id=a32e37cfad63edcba7cfd5d997422d93
That is (was) a session link, now invalid. What keywords did you use?
l.
I got the link from the bottom of this page: https://www.militaryaerospace.com/computers/article/16722067/darpa-asks-indu...
There's an actual absolute path on the webpage which seems to not be a session link: https://www.fbo.gov/spg/ODA/DARPA/CMO/HR001119S0037/listing.html
David
Ah, here the story is: https://web.archive.org/web/20150225094201/http://english.cw.com.tw/article....
It seems to me that there will always be thieves in the world. Hiding science does more harm to the scientific community then the thieves. Think long term though, luke, as chip fabrication plants are prone to do, at least for process design and improvement. If I'm a priest and I'll give away copies of the Bible and your a chip fabrication guru and you don't ever tell anyone anything about chip fabrication what will the population learn about?! Isn't it obvious that science will ultimately die off or become some super diluted idiocy (old wives tales)?! If people are willing to die for God(s) (real or not), but not even lift a finger, but instead purposefully restrict knowledge and understanding, to teach their fellow man the sciences will the sciences not be out competed many to one?! Am I the only one who sees this (our/your/the chip fabs), fate?
To be honest, I don't really understand why more people don't support freedom of knowledge more, one of the basic beliefs of serving God was supposed to be freedom to be the best you, you can be. Yet at the same time, there are people out there who *claim to believe in god* who think excercising freedom means hoarding knowledge or rights for yourself only and restricting others. The irony of that is that you cannot serve both money or God. Another irony though is that sharing is universally a good thing and in many ways is Godly. As long as it isn't about sharing people... aka romance...
I also believe though we need to take better care of our world. Stupid people who don't believe in any form of science or believe in psuedo science(s) should not be our rulers, etc... Extremism, as always is the main issue of life. People like to be all or nothing a lot, but I believe that all or nothing thinking leads to calamities. There is a lot of gray out there. Some darker, some lighter. Emulation is a perfect example of this, as companies genuinely seem to want to make you have to pay for multiple copies of the same thing if one breaks. This is complete bs, digital restrictions management and all rights reserved licenses will always be the wrong method due to the issue of not being allowed to modify or share or reveal what is in it.
PS, I don't believe science and God disprove each other. There can be unfamothable parallels.
Many paths, many sorrows, one hope only.
I digress, this cannot be my life. It's fundamentally not for me. I'm a fool for wanting to learn -- anything. I'm wedded to suicide. I'd be equally productive considering the long term, in watching the clouds go by. I'm going to do things differently... much, much differently, if I have any say so at all in the matter!
As for this, suicide is not the answer. IF someone is truly evil on a nazi level, I think its less bad, but still bad nonetheless.
I wish you the best on finding hope in life. That being said, evil is very hard to avoid in our pit of a world.
PS, anyone reading this may agree to disagree if they choose. I don't want to force people to be a certain way. Everyone has a path to carve, choose wisely.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
According to claws-mail, this email was lost. If this is a resend then it is also a re-write with a little additional material at the end. So, reply to this one.
On Tue, 9 Jul 2019 00:00:06 -0400 zap zapper@disroot.org wrote:
Ah, here the story is: https://web.archive.org/web/20150225094201/http://english.cw.com.tw/article....
It seems to me that there will always be thieves in the world. Hiding science does more harm to the scientific community then the thieves. Think long term though, luke, as chip fabrication plants are prone to do, at least for process design and improvement. If I'm a priest and I'll give away copies of the Bible and your a chip fabrication guru and you don't ever tell anyone anything about chip fabrication what will the population learn about?! Isn't it obvious that science will ultimately die off or become some super diluted idiocy (old wives tales)?! If people are willing to die for God(s) (real or not), but not even lift a finger, but instead purposefully restrict knowledge and understanding, to teach their fellow man the sciences will the sciences not be out competed many to one?! Am I the only one who sees this (our/your/the chip fabs), fate?
To be honest, I don't really understand why more people don't support freedom of knowledge more, one of the basic beliefs of serving God was supposed to be freedom to be the best you, you can be. Yet at the same time, there are people out there who *claim to believe in god* who think excercising freedom means hoarding knowledge or rights for yourself only and restricting others. The irony of that is that you cannot serve both money or God. Another irony though is that sharing is universally a good thing and in many ways is Godly. As long as it isn't about sharing people... aka romance...
I also believe though we need to take better care of our world. Stupid people who don't believe in any form of science or believe in psuedo science(s) should not be our rulers, etc... Extremism, as always is the main issue of life. People like to be all or nothing a lot, but I believe that all or nothing thinking leads to calamities. There is a lot of gray out there. Some darker, some lighter. Emulation is a perfect example of this, as companies genuinely seem to want to make you have to pay for multiple copies of the same thing if one breaks. This is complete bs, digital restrictions management and all rights reserved licenses will always be the wrong method due to the issue of not being allowed to modify or share or reveal what is in it.
PS, I don't believe science and God disprove each other. There can be unfamothable parallels.
Many paths, many sorrows, one hope only.
I was contrasting the spread of religion with science to make a point, not that the are mutually exclusive. For if God created the world, which works on scientific principles, can God be against science?
I digress, this cannot be my life. It's fundamentally not for me. I'm a fool for wanting to learn -- anything. I'm wedded to suicide. I'd be equally productive considering the long term, in watching the clouds go by. I'm going to do things differently... much, much differently, if I have any say so at all in the matter!
As for this, suicide is not the answer. IF someone is truly evil on a nazi level, I think its less bad, but still bad nonetheless.
I wish you the best on finding hope in life. That being said, evil is very hard to avoid in our pit of a world.
PS, anyone reading this may agree to disagree if they choose. I don't want to force people to be a certain way. Everyone has a path to carve, choose wisely.
I did not mean to indicate that suicide was the answer. I was trying to say that a system that kills itself is suicidal and that in as much as I participate in it then I too will be killing my field and aught to be in another. My hope is in doing things differently in keeping with the obvious words of Christ: "Neither do men light a candle (the candle of science), and put it under a bushel, but upon a candlestick, that it may give light to all that are in the house." -- Douay Rheims
Chip fabs depend on smart well educated people to operate and design them. If they wall off the knowledge as to how they work the brains of the smart people will either take interest in another field, atrophy, or arrive at the plants will little knowledge of what to do or how they work. I've tried to find books on the subject. Look at the current material like "Silicon Processing For The Vlsi Era":
These are ebay used books: $4.57 293115019126 Volume 1: Process Technology $5.60 143211754813 Volume 2: Process Integration (ExLib) $6.08 143179464527 Volume 3: The Submicron MOSFET
And finally on amazon used books: $499.51 ISBN: 978-0961672171 Volume 4: Deep-Submicron Process Technology
That price for a book of 822 pages *used* is HUGE!!!! I can get a *new* book with ~2x as many pages for $130, namely, my Bible.
Read the comments (highlights mine): "This book provides a great amount of information on a field where there isn't much information to be found.[SIC] It is easy to read and very descriptive - highly recommended.
It is a 2002[SIC] book though, talking about a field which evolves too quickly. Some of the "future" things the book talks about have already happened or the industry decided to move in a different direction, but in general this is not much a problem.[Unless you want to learn how to do this.]"
Or
"This well written reference book is highly recommended to anyone interested in the technology used to manufacture deep-submicron MOSFETs, i.e., MOSFETs requiring lithography in the 1/4 to 1/8-micron[SIC] range. ..."
1/8th of a micron is 125nm! 125nm! We're at 7nm, with 6nm risk production IIRC, and 5nm is in an alpha state!
...Sigh...
Feel free to ask for clarification if needed. I'll try again. I'm not mad at you, at them, rather like Faust, I suffer from "care". And a bad case of it at that.
Sincerely, David
I did not mean to indicate that suicide was the answer. I was trying to say that a system that kills itself is suicidal and that in as much as I participate in it then I too will be killing my field and aught to be in another. My hope is in doing things differently in keeping with the obvious words of Christ: "Neither do men light a candle (the candle of science), and put it under a bushel, but upon a candlestick, that it may give light to all that are in the house." -- Douay Rheims
Ah, okay. Just making sure. Suicide is understandable to me, but its still not a good idea or right no matter the reason. Though I do sympathize with people who think they need to.
That being said, if everything we have comes from God, why on earth do we try to copyright things to death to the point where when we sell stuff, we still technically own a part of it and the other person only half owns it. It really doesn't add up.
That being said, on an unrelated note, I wonder if OpenBSD or a fork of it, will ever be put on the eoma68 laptop. (I know its a standard, but I was talking about all revisions of it.)
I only ask, because OpenBSD can be used in a libre way and because its the most secure known operating system on the planet that I know of.
@Doark, I appreciate how you attached the (s) at the end. A nice gesture of "good faith" for any pagan..
Unfortunately, new ideas tend to require isolation from existing ideas. People create more quarky software, when they have no examples to go on.
Free information AND free culture, somewhat, if not mostly depend, on people isolating their selves from information, to continue developing new perspectives.
Increasing awareness about internet addictions as well as what makes them unique, as well as discussion about filter bubbles which may lead to more rigorous libre designs, we can behold that shift however we must lockstep with internet community health.
Not to derail an already wobbly train. As for suicide, This disturbing zeitgeist causes me to feel a few axioms need said. Too many, too prestigious forums censor these moral views. (including wikipedia: https://en.wikipedia.org/wiki/Talk:Eaterjolly ) I deem giving meaning to another's suicidal ideation, (normative-ly) immoral. I deem actively (contrary to passively) preventing suicide more heinous than murder.
We could discuss for ages the caveats and reasons. If anyone would like to, I invite them to email me directly.. we could even start a group discussion. Sounds grimly fun.
More on-topic: @Luke, your 3D printer setup still impresses me.
P.S. I changed my tagline name to J.L., but you all knew me as "Jean" or "Jean-Luc" Flamelle previously (nickname, hopefully one day real name). I haven't posted in a while.
On Tue, Jul 9, 2019 at 4:28 AM doark@mail.com wrote:
You know what happens when you can't answer one of my questions, right? :) tap tap tap tap ...... tap tap tap ... Aha! How's this, a 28/22nm chip fabrication plant for $4.2 billion? Not that GlobalFounderies ever had a winning process... https://www.anandtech.com/show/2814/3 And, if you read further down you can see that you can get a smaller plant with a College campus for $4 billion.
cool!
I guess this means to buy one we'll have to empty our piggy banks. ;-)
or someone else's. always better to empty someone else's piggy bank :)
l.
On Tue, 9 Jul 2019 14:35:48 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Tue, Jul 9, 2019 at 4:28 AM doark@mail.com wrote:
You know what happens when you can't answer one of my questions, right? :) tap tap tap tap ...... tap tap tap ... Aha! How's this, a 28/22nm chip fabrication plant for $4.2 billion? Not that GlobalFounderies ever had a winning process... https://www.anandtech.com/show/2814/3 And, if you read further down you can see that you can get a smaller plant with a College campus for $4 billion.
cool!
<snip>
Please be aware that there are 3 costs involved in a foundry (I wish I could remember or find where I read this). 1. The cost of building it. Really big. 2. The cost of starting it up for the first time. AFAIR 1/2 of the cost of building it. 3. The cost of running it. Normally the customers pay for this.
So, just having the funds to build a foun doesn't mean we can start producing chips.
Sincerely, David
On 7/5/19, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
This is of course all inferred guesswork. Welcome to my world of low probability logical deduction aka reverse engineering.
Sounds as though they have a "Speak no evil; Hear no evil, then See no evil" mentality. They'd likely prefer not to consciously perceive by any readily made stretch of the imagination the moral dissonance associated with secrecy to stifle global manufacturing developments, including their own in a prisoner's dilemma type scenario.
That probably would make them more willing to admit fault or talk openly about changing the world to a close deeply-trusted associate very rarely, since exclusively thinking about such changes in safely confined contexts mitigates moral strain or moral system breakage.
Fortunately none of us have sold our souls, yet; so, we needn't worry about reinforcing that sortof self-deception to maintain our sense of identity... yet : )
Attack on Titan, when Commander Erwin commanded Levi to save Armin instead. Erwin had become a monster to fight monsters. Now, with the humans on the verge of victory against their foe, someone not a monster must take the monsters place. Can one even preserve their personhood as a titan, or do all titans eventually become monsters?
I suppose the struggle between person and beast will always exist, and stronger in titans.
arm-netbook@lists.phcomp.co.uk
-
David Niklas -
doark＠mail.com
-
J.L. -
Luke Kenneth Casson Leighton -
zap