Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.forth > #15795 > unrolled thread
| Started by | "B.Person" <bruce.person@gmail.com> |
|---|---|
| First post | 2012-09-28 07:16 -0700 |
| Last post | 2012-10-27 08:54 -0700 |
| Articles | 20 on this page of 97 — 19 participants |
Back to article view | Back to comp.lang.forth
Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "B.Person" <bruce.person@gmail.com> - 2012-09-28 07:16 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) jacko <jackokring@gmail.com> - 2012-09-28 10:44 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) jacko <jackokring@gmail.com> - 2012-09-28 10:45 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Mark Wills <forthfreak@gmail.com> - 2012-09-28 11:42 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Jecel <jecel@merlintec.com> - 2012-09-28 11:39 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-09-28 11:44 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-09-29 14:55 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-09-29 12:13 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-09-29 23:13 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-09-29 17:30 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Elizabeth D. Rather" <erather@forth.com> - 2012-09-29 17:59 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Anonymous <nobody@remailer.paranoici.org> - 2012-09-30 13:50 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Ed" <invalid@nospam.com> - 2012-10-02 23:53 +1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Fritz Wuehler <fritz@spamexpire-201210.rodent.frell.theremailer.net> - 2012-10-03 19:10 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-03 07:41 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. vandys@vsta.org - 2012-10-03 18:06 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-03 20:22 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-04 08:39 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-04 07:51 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-04 14:05 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-04 09:26 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-04 19:23 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. vandys@vsta.org - 2012-10-05 00:41 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-05 15:43 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. vandys@vsta.org - 2012-10-05 21:14 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-07 17:58 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Ed" <invalid@nospam.com> - 2012-10-04 13:40 +1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-03 21:34 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Anonymous <nobody@remailer.paranoici.org> - 2012-10-04 09:49 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Ed" <invalid@nospam.com> - 2012-10-10 00:10 +1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-09 16:42 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-09 14:01 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Paul Rubin <no.email@nospam.invalid> - 2012-10-09 11:43 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-09 09:02 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-09 15:34 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-09 09:53 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-09 16:06 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-09 17:00 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-10 20:54 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-11 19:26 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-12 23:52 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Elizabeth D. Rather" <erather@forth.com> - 2012-10-13 07:56 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-13 15:06 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-10 01:29 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. rickman <gnuarm@gmail.com> - 2012-10-09 15:31 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Paul Rubin <no.email@nospam.invalid> - 2012-10-09 14:14 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-09 21:37 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Paul Rubin <no.email@nospam.invalid> - 2012-10-09 15:26 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-10 02:21 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. Paul Rubin <no.email@nospam.invalid> - 2012-10-09 19:40 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-04 17:07 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-09-29 22:12 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Elizabeth D. Rather" <erather@forth.com> - 2012-09-29 18:38 -1000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-09-29 23:38 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Jecel <jecel@merlintec.com> - 2012-09-30 11:47 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-02 00:19 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-02 16:06 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) gavino_himself <visploveslisp@gmail.com> - 2012-10-05 14:50 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-06 03:14 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-09-30 16:34 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-01 02:31 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-10-01 01:27 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-01 16:57 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-02 16:51 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-03 02:48 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-04 17:47 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-05 00:39 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-06 18:47 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Paul Rubin <no.email@nospam.invalid> - 2012-10-02 19:23 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) gavino_himself <visploveslisp@gmail.com> - 2012-09-30 17:05 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cmm> - 2012-10-01 02:26 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) albert@spenarnc.xs4all.nl (Albert van der Horst) - 2012-10-05 19:45 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) albert@spenarnc.xs4all.nl (Albert van der Horst) - 2012-10-04 13:22 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) awegel@arcor.de (Alex Wegel) - 2012-10-06 02:37 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Mark Wills <forthfreak@gmail.com> - 2012-10-08 01:38 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-08 20:43 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-08 20:56 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-08 16:00 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-08 21:05 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-08 16:40 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-08 23:16 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Mark Wills <forthfreak@gmail.com> - 2012-10-09 00:19 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-09 18:43 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) albert@spenarnc.xs4all.nl (Albert van der Horst) - 2012-10-09 09:59 +0000
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Chris Hinsley <chris.hinsley@gmail.com> - 2012-10-09 18:46 +0100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) marko <marko@marko.marko> - 2012-10-10 07:29 +1100
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-09 17:11 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-10 01:11 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-09 20:46 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) Bernd Paysan <bernd.paysan@gmx.de> - 2012-10-10 14:48 +0200
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) gavino_himself <visploveslisp@gmail.com> - 2012-10-12 13:35 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-13 00:01 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "Rod Pemberton" <do_not_have@notemailnotz.cnm> - 2012-10-13 20:51 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) rickman <gnuarm@gmail.com> - 2012-10-14 17:36 -0400
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) gavino_himself <visploveslisp@gmail.com> - 2012-10-12 13:34 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "B.Person" <bruce.person@gmail.com> - 2012-10-26 13:35 -0700
Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) "B.Person" <bruce.person@gmail.com> - 2012-10-27 08:54 -0700
Page 3 of 5 — ← Prev page 1 2 [3] 4 5 Next page →
| From | rickman <gnuarm@gmail.com> |
|---|---|
| Date | 2012-10-12 23:52 -0400 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <k5c8qk$otl$1@dont-email.me> |
| In reply to | #16207 |
On 10/12/2012 1:26 AM, Elizabeth D. Rather wrote: > On 10/10/12 2:54 PM, rickman wrote: >> On 10/9/2012 11:00 PM, Elizabeth D. Rather wrote: >>> On 10/9/12 10:06 AM, rickman wrote: > ... >>>> As to cost, Forth may be similar in cost to other code development >>>> tools, but in this context it is a superfluous expense since the other >>>> tools already exist in the company. At least that is how managers view >>>> it. >>> >>> So, the next opportunity would be when you're addressing a new target >>> for which you don't have a compiler :-) >> >> I don't work for one of these large companies anymore. The next >> opportunity is the next time I get paying work. In the mean time I will >> take a look at the eval copies. I seem to recall the target eval >> systems are somewhat limited the last time I looked, but that has been a >> while so I'll do a refresh. > > There's a feature comparison chart here: > http://www.forth.com/embedded/swiftx-embedded-systems-13.html I have looked at this before, but perhaps it has been updated. The problem is I don't understand what the list describes. I am offline now, but I'll look into this when I am back online. (Now is when I am writing this, not when it is sent...). Rick
[toc] | [prev] | [next] | [standalone]
| From | "Elizabeth D. Rather" <erather@forth.com> |
|---|---|
| Date | 2012-10-13 07:56 -1000 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <Mu6dncGiicf_NuTNnZ2dnUVZ_oudnZ2d@supernews.com> |
| In reply to | #16235 |
On 10/12/12 5:52 PM, rickman wrote: > On 10/12/2012 1:26 AM, Elizabeth D. Rather wrote: >> On 10/10/12 2:54 PM, rickman wrote: >>> On 10/9/2012 11:00 PM, Elizabeth D. Rather wrote: >>>> On 10/9/12 10:06 AM, rickman wrote: >> ... >>>>> As to cost, Forth may be similar in cost to other code development >>>>> tools, but in this context it is a superfluous expense since the other >>>>> tools already exist in the company. At least that is how managers view >>>>> it. >>>> >>>> So, the next opportunity would be when you're addressing a new target >>>> for which you don't have a compiler :-) >>> >>> I don't work for one of these large companies anymore. The next >>> opportunity is the next time I get paying work. In the mean time I will >>> take a look at the eval copies. I seem to recall the target eval >>> systems are somewhat limited the last time I looked, but that has been a >>> while so I'll do a refresh. >> >> There's a feature comparison chart here: >> http://www.forth.com/embedded/swiftx-embedded-systems-13.html > > I have looked at this before, but perhaps it has been updated. The > problem is I don't understand what the list describes. I am offline > now, but I'll look into this when I am back online. (Now is when I am > writing this, not when it is sent...). The various features are described in the pages linked in the right-hand panel. Cheers, Elizabeth -- ================================================== Elizabeth D. Rather (US & Canada) 800-55-FORTH FORTH Inc. +1 310.999.6784 5959 West Century Blvd. Suite 700 Los Angeles, CA 90045 http://www.forth.com "Forth-based products and Services for real-time applications since 1973." ==================================================
[toc] | [prev] | [next] | [standalone]
| From | rickman <gnuarm@gmail.com> |
|---|---|
| Date | 2012-10-13 15:06 -0400 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <k5ce4n$pis$1@dont-email.me> |
| In reply to | #16242 |
On 10/13/2012 1:56 PM, Elizabeth D. Rather wrote: > On 10/12/12 5:52 PM, rickman wrote: >> On 10/12/2012 1:26 AM, Elizabeth D. Rather wrote: >>> On 10/10/12 2:54 PM, rickman wrote: >>>> On 10/9/2012 11:00 PM, Elizabeth D. Rather wrote: >>>>> On 10/9/12 10:06 AM, rickman wrote: >>> ... >>>>>> As to cost, Forth may be similar in cost to other code development >>>>>> tools, but in this context it is a superfluous expense since the >>>>>> other >>>>>> tools already exist in the company. At least that is how managers >>>>>> view >>>>>> it. >>>>> >>>>> So, the next opportunity would be when you're addressing a new target >>>>> for which you don't have a compiler :-) >>>> >>>> I don't work for one of these large companies anymore. The next >>>> opportunity is the next time I get paying work. In the mean time I will >>>> take a look at the eval copies. I seem to recall the target eval >>>> systems are somewhat limited the last time I looked, but that has >>>> been a >>>> while so I'll do a refresh. >>> >>> There's a feature comparison chart here: >>> http://www.forth.com/embedded/swiftx-embedded-systems-13.html >> >> I have looked at this before, but perhaps it has been updated. The >> problem is I don't understand what the list describes. I am offline >> now, but I'll look into this when I am back online. (Now is when I am >> writing this, not when it is sent...). > > The various features are described in the pages linked in the right-hand > panel. Actually the feature list is pretty clear, it just isn't complete. It doesn't mention any of the limitations in the eval version that you told me about. Also, just so you know, the link you gave is to a page where the box on the right has few working links. If I go to the hardware page it has the same box and the links work ok. I assume that when you say I can't save the object, you mean I have to download to the target every time I want to test the code, so it has to fit in the RAM. That means I can't burn it to Flash? These sorts of limitations may seem necessary to a company, but it can make it hard to evaluate a product. I often don't have time to spend just evaluating a tool that doesn't contribute directly to a project. At this time that's not true, so maybe I'll take a few days and look at this. Is there any aspect of these tools that require an Internet connection during use? I've seen tools that only provide help via the Internet for example. I don't have an Internet connection full time. Rick
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-10-10 01:29 +0200 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <3360703.kdtp696y0j@sunwukong.fritz.box> |
| In reply to | #16121 |
rickman wrote: > How does the engineer get the code working when money has to be spent > on a commercial Forth package? If I didn't need to spend money, I > would have had Forth on my lab bench many times over. There are two reasons to use a commercial Forth over a free one: * You want to produce a binary for customers and not hand it out in a GPL compatible way * You need support, which is typically the business model of Forth vendors, and support for a free system highly depends on how nicely you ask the developers. If you are just doing lab tests (which are "inhouse use", and therefore don't result in any problems with the GPL), and you feel comfortable enough with the system so that you don't need much support, there's no reason not to go with a free Forth like Gforth. The people in our Forth group in Munich had no problem to use Forth in quite a number of locations. Even free Forth systems. One time, they ported Gforth to a smart card at G&D, a highly secretive and paranoid company, and they even were allowed to take their modifications to Gforth out (as required by the GPL)! That never happened at G&D before or after - but sometimes, even these things are possible. Ob-Gavino: What they implemented on the smard card was a tiny web server, including SSL, as secure way to access a smard card from a browser. G&D decided to implement that in Java for a commercial product instead, but for years they had nothing to show but the Forth demo, and finally, they stopped the Java project. Java is fine on a big machine, but on a small smartcard, it's not possible to do a web server in Java. It is possible in Forth. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
| From | rickman <gnuarm@gmail.com> |
|---|---|
| Date | 2012-10-09 15:31 -0400 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <k51u2g$9e3$1@dont-email.me> |
| In reply to | #16110 |
On 10/9/2012 2:43 PM, Paul Rubin wrote: > rickman<gnuarm@gmail.com> writes: >>> I haven't encountered such an employer yet. Even the ones I chose to >>> leave allowed me to use Forth. >> I have yet to find an employer who is ok with using Forth. > > We're shaking out a new piece of hardware where I work. If the hardware > guys happened to be Forthers and wanted to use Forth as a benchtop > debugging tool to poke at registers during development, it would > probably be fine. There's no way the software guys (e.g. me) would ever > be allowed to put Forth into the actual end product. The last time I tried to use Forth at work was exactly that, to test hardware designs (since I'm a hardware person). I got it into the initial budget, but when it came time to spend it, the program manager refused. You can make all the arguments in the world, but the PMs consider it to be their money. I got to pound sand. > Outside of small embedded devices that have no long-term code base > (develop, test, finished--hopefully no updates ever), since programmers > come and go, any code inside a product has to expect to be maintained by > more than one person. That means a decision to let one person use Forth > is also a decision to require other people to maintain that Forth, which > in turn creates the need to find or train more Forthers. Avoiding that > headache is a perfectly sound decision unless you're already running a > Forth shop (there are of course a few of these that are doing ok). > > It's less a matter of imposing conditions on the would-be Forth user, > than avoiding imposing on the non-Forth-users who are going to have to > deal with the code later. Trust me, I understand all of this completely. Heck, I've worked places before where I couldn't even talk anyone into letting me have newsgroup access. That was when I started using Google Groups, when it was clean and simple even if it had all sorts of other problems. The problem is that Forth is not mainstream and that means managers don't want to hear about it. I can make all the productivity arguments I want and they won't cut loose the money. Rick
[toc] | [prev] | [next] | [standalone]
| From | Paul Rubin <no.email@nospam.invalid> |
|---|---|
| Date | 2012-10-09 14:14 -0700 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <7x1uh7tkjf.fsf@ruckus.brouhaha.com> |
| In reply to | #16120 |
rickman <gnuarm@gmail.com> writes: >> If the hardware guys happened to be Forthers and wanted to use Forth >> as a benchtop debugging tool to poke at registers during development, >> it would probably be fine. ... > The last time I tried to use Forth at work was exactly that, to test > hardware designs (since I'm a hardware person). I got it into the > initial budget, but when it came time to spend it, the program manager > refused. Maybe I should have said explicitly: I couldn't imagine my management approving spending even 1 cent for purchasing Forth software. I just meant they probably wouldn't have interfered if the hardware guys had downloaded a free Forth from someplace and used it in the lab for one-off debugging.
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-10-09 21:37 +0200 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <1788394.NIAO6PipmW@sunwukong.fritz.box> |
| In reply to | #16110 |
Paul Rubin wrote: > Outside of small embedded devices that have no long-term code base > (develop, test, finished--hopefully no updates ever), since > programmers come and go, any code inside a product has to expect to be > maintained by > more than one person. That means a decision to let one person use > Forth is also a decision to require other people to maintain that > Forth, which > in turn creates the need to find or train more Forthers. Avoiding > that headache is a perfectly sound decision unless you're already > running a Forth shop (there are of course a few of these that are > doing ok). IMHO this is nonsense. The last company I left (because they weren't actually innovative and had many "replacement" concerns, and led slip some "We don't know what we can do with you in the long run" pretty early) needed a replacement for me quickly, when I left. The engineer that had to take over my work didn't take that long to learn b16 Forth - he managed to ask all his questions before I actually left. Well, it's only a small subset of Forth, but it still *is* Forth. If you assume that your programmers can't learn another language quickly, you should question that assumption. In the job I had at that company, I was confronted with Verilog (as part of the digital design), Lua (a measurement equipment used that as embedded language), Tcl (as scripting language in most Cadence products), Skill (scripting language in the integrated analog/mixed signal design system from Cadence), LabView, and maybe some I forgot. If you can only do one language, you can do nothing. One part of the project - a GUI to operate the prototype, was done by the intern, who had one week to a) learn Forth, and b) complete the task. He had some similar LabView experience (for which he had two months), and concluded that Forth+MINOS was much easier to use than LabView. My boss told me I rather should have written the program in C so that he could understand it. My boss didn't have any software coding experience, he didn't even understand the textual description of the algorithm, any formula was "too confusing" and all the diagrams as well. This was a 500 lines program... What he wanted was an employee that didn't challenge his position. Some elbonian pigs might barely qualify for *that part* of the job description, though he was bald, fat, and could only grunt. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
| From | Paul Rubin <no.email@nospam.invalid> |
|---|---|
| Date | 2012-10-09 15:26 -0700 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <7xr4p7s2ms.fsf@ruckus.brouhaha.com> |
| In reply to | #16122 |
Bernd Paysan <bernd.paysan@gmx.de> writes: > The engineer that had to take over my work didn't take that long to > learn b16 Forth ... Well, it's only a small subset of Forth, but it > still *is* Forth. I'd consider it to be a slightly weird machine language, like the F18's. As such, someone who's programmed other devices in assembler can probably deal with it, by treating the Forth as assembly code. Of course there's a big productivity hit compared to writing in a HLL. > If you assume that your programmers can't learn another language > quickly, you should question that assumption. It's not just languages but the body of tools and techniques that the languages come with. Assembly language is conceptually very simple, but writing big assembly programs without creating a nest of bugs requires careful technique that takes a while to develop. It's much easier with HLL's. > In the job I had at that company, I was confronted with Verilog (as > part of the digital design), Lua (a measurement equipment used that as > embedded language), Tcl (as scripting language in most Cadence > products), Skill (scripting language in the integrated analog/mixed > signal design system from Cadence), LabView, and maybe some I forgot. Tcl is pretty trivial and Lua is straightforward to someone who's Lisp/Python/Javascript/etc. I don't know about Skill or Labview. I've been interested in Verilog for a while, and maybe I'm overestimating its difficulty or maybe I'm just not all that bright, but I currently think of learning it (to the point of being able to implement nontrivial circuits that aren't full of bugs or bad performance mistakes) to be a fairly major undertaking. Forth as a HLL is something like that too, from what I can tell. > My boss told me I rather should have written the program in C ... > This was a 500 lines program... C is itself a clunky legacy language, still useful for some things, but often not really what you're competing against. If you're writing a GUI to externally control some lab gizmo, you should compare the effort of writing your 500 lines of Forth to implementing in (e.g.) Python, or as a HTML web app running on a PC.
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-10-10 02:21 +0200 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <4594174.cvhJYoXqy3@sunwukong.fritz.box> |
| In reply to | #16135 |
Paul Rubin wrote: > Bernd Paysan <bernd.paysan@gmx.de> writes: >> The engineer that had to take over my work didn't take that long to >> learn b16 Forth ... Well, it's only a small subset of Forth, but it >> still *is* Forth. > > I'd consider it to be a slightly weird machine language, like the > F18's. As such, someone who's programmed other devices in assembler > can > probably deal with it, by treating the Forth as assembly code. Of > course there's a big productivity hit compared to writing in a HLL. I've had other cowoerkers who went through learning b16 Forth earlier, i.e. where I could get feedback. They went through that "Forth is an assembler" stage, but after they got a bit more fluent in Forth, they didn't see it as such anymore. It may not feel as comfortable as Python or something like that to them, but it is way beyond the typical weird assembler language (think e.g. of 8051) you see in these small embedded processors. Most of these tasks can only be done reasonably in assembly language, because these deep embedded stuff is done in >9 digits units, and whatever pain it takes the programmer to make the program fit into 1mm² less memory is absolutely worth it. I've written a demo program for Forth Gesellschaft's pick&place robot in about a week for the last LinuxTag. It included some tricky trigonometic. Apart from the fact that b16 can't do an acos() function (so use a table), it wasn't particularly tricky. We showed the program to various bypassers on our booth, and one considered the code as "fairly readable" - not the mostly trivial higher level parts, but even the trigonometry stuff in the middle of the program. >> If you assume that your programmers can't learn another language >> quickly, you should question that assumption. > > It's not just languages but the body of tools and techniques that the > languages come with. Assembly language is conceptually very simple, > but writing big assembly programs without creating a nest of bugs > requires > careful technique that takes a while to develop. It's much easier > with HLL's. You don't write big programs, there is no space for big programs on a small embedded processor. You write small programs. It's impossible to write small programs in HLLs (including their run-time). > Tcl is pretty trivial and Lua is straightforward to someone who's > Lisp/Python/Javascript/etc. I don't know about Skill or Labview. Labview is painting programs with little icons and arrows, similar to Matlab/Simulink (another language you often see in these environments). Skill is another Lisp-like language. > I've been interested in Verilog for a while, and maybe I'm > overestimating its difficulty or maybe I'm just not all that bright, > but I currently think of learning it (to the point of being able to > implement nontrivial circuits that aren't full of bugs or bad > performance mistakes) to be a > fairly major undertaking. Forth as a HLL is something like that too, > from what I can tell. You probably have to unlearn a lot. Too much particular knowledge is in the way when you do Forth. Don't try to be clever. The unfactored Doug Hoffman FizzBuzzZoom is a good example. >> My boss told me I rather should have written the program in C ... >> This was a 500 lines program... > > C is itself a clunky legacy language, still useful for some things, > but > often not really what you're competing against. If you're writing a > GUI to externally control some lab gizmo, you should compare the > effort of writing your 500 lines of Forth to implementing in (e.g.) > Python, or as a HTML web app running on a PC. The 500 lines were the Forth program on the deeply embedded b16. Some parts of the management were considering buying an ARM core and use C on that, though we had evaluated that comparison earlier, and the ARM core simply was way too expensive both in terms of die size and license costs. The thinking is about "when we buy an ARM core, our costs for the chip increase by several million dollars, but we don't have to be nice to an expensive expert, but can be mean to a cheap programmer". That's definitely worth it ;-). The MINOS GUI could probably compared to a Python implementation using Qt as library, but a web app would be completely impossible (talk through an FTDI USB device in SPI mode with a web app?). The company bought me (including team), because they had no internal knowledge about battery monitoring, and I found out why (they don't want people with knowledge, because these threaten the position of the bosses, who have no knowledge at all, including no knowledge in management). This was about the most dilbertesque and grotesque employer I had, but they did still allow me to use Forth, even though they apparently were scared by me for whatever I did. Even taking a direct international flight was considered offensive (the typical plan was to have a stop- over in London, which is IMHO completely crazy). -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
| From | Paul Rubin <no.email@nospam.invalid> |
|---|---|
| Date | 2012-10-09 19:40 -0700 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <7xr4p76od8.fsf@ruckus.brouhaha.com> |
| In reply to | #16139 |
Bernd Paysan <bernd.paysan@gmx.de> writes: > You don't write big programs, there is no space for big programs on a > small embedded processor. You write small programs. It's impossible to > write small programs in HLLs (including their run-time). Fair enough. On such small processors, the advantages of HLL's are less pronounced anyway. > The 500 lines were the Forth program on the deeply embedded b16. Oh I see. > The MINOS GUI could probably compared to a Python implementation using > Qt as library, but a web app would be completely impossible (talk > through an FTDI USB device in SPI mode with a web app?). Sure--by web app I just mean a PC program speaking HTTP to a socket, not something running on a hosted server or anything like that. It would have the same USB stuff as the Qt program. The Python stdlib has an embeddable web server module, or you could write the program as a CGI behind a separate server. Using either of these is less work than programming Qt if you don't mind the limitations of a web gui, and has the advantages that multiple people can use it from their desks using browsers. I do this all the time for my own stuff where I just want simple functionality and don't care about it looking slick. > This was about the most dilbertesque and grotesque employer I had, Condolences--there is a lot of that anywhere you go.
[toc] | [prev] | [next] | [standalone]
| From | "Rod Pemberton" <do_not_have@notemailnotz.cmm> |
|---|---|
| Date | 2012-10-04 17:07 -0400 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. |
| Message-ID | <k4ktkr$hgm$1@speranza.aioe.org> |
| In reply to | #15883 |
"Fritz Wuehler" <fritz@spamexpire-201210.rodent.frell.theremailer.net> wrote in message news:923e0c628f2cd64d742acdea342c91e8@msgid.frell.theremailer.net... > "Ed" <invalid@nospam.com> wrote: ... > There are many niche tools and domain specific languages that > are not popular [n]or widely known, yet they serve a purpose > and are in active use. Sigh ... Who cares? If they're not popular nor widely known, they'll be extinct. > Really, the statement you quoted is troubling not from a Forth point of > view, but consider all the graphical crap and web programming that > probably make up 99% of all code written today...if that junk isn't a > solution looking for a problem I don't know what is. We got business > done with text interfaces for many years and I don't see any benefit > except for teenagers and H1Bs in reducing programming to flash > webpages. Doesn't anybody write code that actually *does* > something anymore? "We got business done with text interfaces for many years ..." Are you a retired COBOL programmer? It fits. > In theory that seems reasonable. In practice there seems to be no limit to > the number of programming languages in use and no end in sight when it > comes to creating new ones. Except, they all have C backends, or had them. So, why bother with them? Why not learn C? Rod Pemberton
[toc] | [prev] | [next] | [standalone]
| From | "Rod Pemberton" <do_not_have@notemailnotz.cmm> |
|---|---|
| Date | 2012-09-29 22:12 -0400 |
| Message-ID | <k489je$g36$1@speranza.aioe.org> |
| In reply to | #15818 |
"Bernd Paysan" <bernd.paysan@gmx.de> wrote in message news:2207084.uDy9Jnq5yQ@sunwukong.fritz.box... > Paul Rubin wrote: > > "Rod Pemberton" <do_not_have@notemailnotz.cmm> writes: ... > >> c) It's not being produced by a major semiconductor company, i.e., > >> it's got the "kiss of death" already ... > > > > Meh. Is ARM a "major semiconductor company"? Is Allwinner a major > > semiconductor company? What about NVidia? Heck, what about Green > > Arrays? You're at a cost disadvantage by being fabless but it's > > certainly possible to ship successful products using foundries. > > A typical Rod Pemberton argument: Completely clueless and hopelessly > outdated... Except for x86 and ARM, all the other early, major microprocessor manufacturers are dead. Some of the early processors: 6502: dead. Z80: dead. TMS9900: dead. 6800: dead. 6809: dead. etc. Every microprocessor design with great potential and great hype is dead too. Can you name one between 1974 and 2012 that isn't? Some of the great hypes: NEC V20/V30: dead. Transputer: dead. DEC Alpha: dead. PA-RISC: dead. Transmeta: dead. PowerPC: dead. Itanium: dead. etc. The microprocessor industry is a graveyard of dead processors. Three of the companies that are still alive are very old: AMD, Intel, and ARM. ARM very nearly died, in the 1990's. I was surprised when I learned it hadn't. It was declared dead. It went into bankruptcy. It never had much market share. It was a bad RISC design decades ago. It was clearly saved by advent of mobile computing devices. What's going to fuel the massively, multi-core ARM market? Low cost? That might work for the $99 version, but not the $499 version. If Zilog Z80 line of microprocessors hadn't died (converted into microcontrollers...), Zilog might have all of ARM's market, today. They had the lowest transistor counts of any microprocessor design at a given performance level. I.e., they'd be cheaper and use less power. Does NVidia produce microprocessors? No ... They produce GPU's. They're different. It's typical for Bernd to support non-facts as facts. There has been a resurgance of new, inexpensive _microcontroller_ designs, not _microprocessors_. That'd make Bernd clueless about the difference between the two. Does it matter that a modern _microcontroller_ is as powerful as an ancient _microprocessor_? AFAIK, I don't lump and dump all processing devices into one category as you do. I don't know much about Green Arrays. But, if it's like Chuck Moore's previous, older, failed, Forth microprocessor projects, it's just a design house of his ... Basically, he's become a high-level patent troll. He predicts a future for microprocessors, which turns out to be false. In the process, he pushes the limits of microprocessor design. Next, he patents what he believes will be needed in the future by other companies, like AMD or Intel. He's correct only a small percentage of the time, but enough to earn him some money or respect. > There are a few semiconductor companies left like Samsung, > TSMC, Intel, IBM, GlobalFoundries, Ti, who can produce the most recent > generation of semiconductor process. But as Moore's law has an > exponential cost in fabs, they get fewer and fewer. Therefore, fabless > is the future. Fabless is a _temporary_ solution driven by cost pressures of building a new fab. No one wants to pay the large upfront cost anymore. However, for a major manufacturer to profit, they must control their own means of production and it's associated costs. That's where the profit comes from: lower cost of manufacturing with large volumes. How do you get lower costs when you outsource? You don't. Your supplier expects a profit above their costs. Their costs are similar to what yours would've been, but typically, slightly more. If you're a major player, they're not as skilled as you in building manufacturing plants. In a vertically integrated company, their profit would've been part of your profit. You can only get modest costs if there is competition among suppliers, and your suppliers are _not_ using recently built - expensive - factories. Any supplier with a new factory can't compete on a cost basis against an old factory. A modern factory can only compete if they are a monopoly. Why would their costs be any lower than your costs? Or, the fabless manufacturer must have an exclusive contract for full use of a fab. Otherwise, a competitor could pay the fab slightly more to produce it's product, instead of producing your product. Then, you don't have any product, or reduced volumes of it. > Even AMD is now fabless. The years where they couldn't > fill their expensive fabs nearly killed them. Do they allow others to use their fabs? (No.) So, effectively the fab becomes a subsidiary via exclusive, contractual use. In exchange for the fab fronting the upfront build cost of the fab for their customers, their fab is paid for over time at a higher total cost to the customer than the upfront cost to build in-house. Your choices are: pay much now, immediately, or pay much more, very slowly. If you've got cash flow but no cash, then the second choice is your only option. > >> My thoughts when I saw that were: > >> "Just what advantage does this design have over the failed > >> Transputer?" http://en.wikipedia.org/wiki/Transputer > > > > I'm not sure the Transputer failed for technical reasons. Tilera has > > been making a 64-core(?) cpu for a while, and Intel has announced the > > Xeon Phi. > > And all the desktop GPUs are essentially many-core architectures. > The GPU tasks are also well suited to parallel processing, unlike those of a microprocessor. > The transputer failed for being too early. It failed because parallel processing was a failure back then. That was mostly because programming languages couldn't produce good parallel processing code. I doubt much has changed since. The languages weren't the problem. The task given to them was. The task weren't suited to parallelism. They still aren't. What's changed in a few decades? Nothing. > Back then, improving the speed of a single core was still the best > way to get more performance - up to about 10 years ago, parallelism > wasn't that important. Things have changed since then. > Which modern programming languages effectively implement parallel processing today? How are they going to program this thing? If each core in a massively multi-core ARM microprocessor handles a distinct task, then it's no better than the multiple co-processor design used in the old Amiga computers. http://en.wikipedia.org/wiki/Amiga Since that's probably what's going to happen, i.e., distinct tasking of each core, let's make some guesses about what each core could be used for. 1 core for single-threaded OS 1 core for sound 1 core for 2D graphics 2 cores for video Wow, we've 11 to go ... Now what? 11 idling applications? Rod Pemberton
[toc] | [prev] | [next] | [standalone]
| From | "Elizabeth D. Rather" <erather@forth.com> |
|---|---|
| Date | 2012-09-29 18:38 -1000 |
| Message-ID | <efmdnSNAa60wUfrNnZ2dnUVZ_jSdnZ2d@supernews.com> |
| In reply to | #15820 |
On 9/29/12 4:12 PM, Rod Pemberton wrote: > "Bernd Paysan" <bernd.paysan@gmx.de> wrote in message > news:2207084.uDy9Jnq5yQ@sunwukong.fritz.box... >> Paul Rubin wrote: >>> "Rod Pemberton" <do_not_have@notemailnotz.cmm> writes: > ... > >>>> c) It's not being produced by a major semiconductor company, i.e., >>>> it's got the "kiss of death" already ... >>> >>> Meh. Is ARM a "major semiconductor company"? Is Allwinner a major >>> semiconductor company? What about NVidia? Heck, what about Green >>> Arrays? You're at a cost disadvantage by being fabless but it's >>> certainly possible to ship successful products using foundries. >> >> A typical Rod Pemberton argument: Completely clueless and hopelessly >> outdated... > > Except for x86 and ARM, all the other early, major microprocessor > manufacturers are dead. > > Some of the early processors: > > 6502: dead. > Z80: dead. > TMS9900: dead. > 6800: dead. > 6809: dead. > etc. > > Every microprocessor design with great potential and great hype is dead too. > Can you name one between 1974 and 2012 that isn't? > > Some of the great hypes: > > NEC V20/V30: dead. > Transputer: dead. > DEC Alpha: dead. > PA-RISC: dead. > Transmeta: dead. > PowerPC: dead. > Itanium: dead. > etc. > > The microprocessor industry is a graveyard of dead processors. Three of the > companies that are still alive are very old: AMD, Intel, and ARM. That's a little unfair. Many of those processors had excellent careers. It's in the nature of technology to have a finite lifetime. The *companies* that survived are the ones that could keep up a pipeline of new processors to replace the ones that become obsolete. ... >> Even AMD is now fabless. The years where they couldn't >> fill their expensive fabs nearly killed them. > > Do they allow others to use their fabs? (No.) So, effectively the fab > becomes a subsidiary via exclusive, contractual use. In exchange for the > fab fronting the upfront build cost of the fab for their customers, their > fab is paid for over time at a higher total cost to the customer than the > upfront cost to build in-house. Your choices are: pay much now, > immediately, or pay much more, very slowly. If you've got cash flow > but no cash, then the second choice is your only option. ... It's a classic in-house vs. outsource decision. As fabs become more and more expensive, and the technology continues to evolve faster than new fabs can be built, the tradeoffs in that decision change. My son works at TI, in the fab that makes DSPs. I get a lot of insights from him. >> And all the desktop GPUs are essentially many-core architectures. >> > > The GPU tasks are also well suited to parallel processing, unlike those > of a microprocessor. That isn't necessarily true, but it is true that the tasks for a GPU are more well-defined than for microprocessors in general, so appropriate design decisions can be made. ... > > Which modern programming languages effectively implement parallel > processing today? How are they going to program this thing? I don't see it as a language's responsibility to sort this out, but the system architect's. > If each core in a massively multi-core ARM microprocessor handles a distinct > task, then it's no better than the multiple co-processor design used in the > old Amiga computers. > http://en.wikipedia.org/wiki/Amiga > > Since that's probably what's going to happen, i.e., distinct tasking of each > core, let's make some guesses about what each core could be used for. > > 1 core for single-threaded OS > 1 core for sound > 1 core for 2D graphics > 2 cores for video > > Wow, we've 11 to go ... Now what? 11 idling applications? Who wants a single-threaded OS? and we've already discussed multicore GPUs for the graphics. It's easy to see several cores involved in streaming video. And there are numerous other I/O tasks. The list goes on. Cheers, Elizabeth -- ================================================== Elizabeth D. Rather (US & Canada) 800-55-FORTH FORTH Inc. +1 310.999.6784 5959 West Century Blvd. Suite 700 Los Angeles, CA 90045 http://www.forth.com "Forth-based products and Services for real-time applications since 1973." ==================================================
[toc] | [prev] | [next] | [standalone]
| From | Paul Rubin <no.email@nospam.invalid> |
|---|---|
| Date | 2012-09-29 23:38 -0700 |
| Message-ID | <7xmx08dngs.fsf@ruckus.brouhaha.com> |
| In reply to | #15820 |
"Rod Pemberton" <do_not_have@notemailnotz.cmm> writes: > Except for x86 and ARM, all the other early, major microprocessor > manufacturers are dead. x86 and ARM are not manufacturers. And tons of other architectures are still doing fine, outside the general purpose 32/64-bit market. > Every microprocessor design with great potential and great hype is dead too. > Can you name one between 1974 and 2012 that isn't? MSP430, doing fine. AVR (various incarnations), doing fine. MIPS, kind of overshadowed by ARM but still doing fine (years ahead of ARM in 64-bit). PIC, doing fine. All sorts of DSP's, going like gangbusters. Etc. > If Zilog Z80 line of microprocessors hadn't died (converted into > microcontrollers...), Zilog might have all of ARM's market, today. Zilog had a 32 bit processor that wasn't very successful, iirc. > Fabless is a _temporary_ solution ... for a major manufacturer to > profit, they must control their own means of production and it's > associated costs. This is like saying that to publish a profitable magazine, you must own your own printing facilities. No not really. You just have to get people to pay enough per issue that you turn a profit. For expensive magazines (technical journals, say) the printing costs don't really figure into it.
[toc] | [prev] | [next] | [standalone]
| From | Jecel <jecel@merlintec.com> |
|---|---|
| Date | 2012-09-30 11:47 -0700 |
| Message-ID | <c817b644-bc32-4c6f-8ccc-a6dea064a7b4@googlegroups.com> |
| In reply to | #15823 |
On Sunday, September 30, 2012 3:38:43 AM UTC-3, Paul Rubin wrote: > "Rod Pemberton" writes: > > Every microprocessor design with great potential and great hype is dead too. > > Can you name one between 1974 and 2012 that isn't? > > MSP430, doing fine. AVR (various incarnations), doing fine. MIPS, kind > of overshadowed by ARM but still doing fine (years ahead of ARM in > 64-bit). PIC, doing fine. All sorts of DSP's, going like gangbusters. > Etc. Exactly. And the PowerPC has 100% of the current generation videogame console market, besides still having a niche in embedded applications. A new version of the 6502 came out this year and the Z80 is also still being developed and sold. There are variations of the 68000 that also have their niche. And we shouldn't forget that though the 8051 has recently been surpassed by the ARM, it is still way more popular than the x86. About the Transputer, I would say that though the T9000 was a major setback it was Inmos that failed and not the Transputer itself. That survived for a very long time as a ST product. And while Acorn had its share of problems, I don't remember ARM Limited having anything but success after its 1990 foundation. -- Jecel
[toc] | [prev] | [next] | [standalone]
| From | rickman <gnuarm@gmail.com> |
|---|---|
| Date | 2012-10-02 00:19 -0400 |
| Message-ID | <k4dq29$g6u$1@dont-email.me> |
| In reply to | #15823 |
On 9/30/2012 2:38 AM, Paul Rubin wrote: > "Rod Pemberton"<do_not_have@notemailnotz.cmm> writes: >> Every microprocessor design with great potential and great hype is dead too. >> Can you name one between 1974 and 2012 that isn't? > > MSP430, doing fine. AVR (various incarnations), doing fine. MIPS, kind > of overshadowed by ARM but still doing fine (years ahead of ARM in > 64-bit). PIC, doing fine. All sorts of DSP's, going like gangbusters. > Etc. Really, MIPS has been around almost as long as I have been. They are far from dead, but not growing much from what I hear, while ARM is going like gang busters. >> If Zilog Z80 line of microprocessors hadn't died (converted into >> microcontrollers...), Zilog might have all of ARM's market, today. > > Zilog had a 32 bit processor that wasn't very successful, iirc. I think it came at a time when nearly all of the makers were going the 8 to 16 to 32 bit route and some who didn't even have CPUs were coming out with one, like National. Too much competition and in the end it came down to firstest with the mostest, Motorola and Intel. >> Fabless is a _temporary_ solution ... for a major manufacturer to >> profit, they must control their own means of production and it's >> associated costs. > > This is like saying that to publish a profitable magazine, you must own > your own printing facilities. No not really. You just have to get > people to pay enough per issue that you turn a profit. For expensive > magazines (technical journals, say) the printing costs don't really > figure into it. I don't think there is much comparison between the publishing industry and semiconductors. Magazines can be printed by nearly anyone. In semiconductors there is a very important link between having a great fab process and having a profitable product line. This is paramount when you are pumping huge volumes of chips out with low profit margins like SDRAM and Flash. When you are making high profit margin devices like desktop and laptop CPUs you can tolerate the lost margin of a middleman if it gets you a leading edge process. That is the ONLY way AMD can hope to survive in Intel's market. AMD is only around at this point because they got well over a billion dollars from Intel a couple of years ago. Since then they have steadily gone down hill from what I have seen. I don't expect them to be around in five years. They will end up being sold off, likely in pieces and Intel will have to worry about the FTC. Other companies don't necessarily need their own fabs. Fabless seems to be doing well for Xilinx. Have they built any fabs yet? Isn't Altera also fabless these days? I'm not a close observer of semicoductor companies these days, but I don't think the FPGA industry spends a lot of time worrying about building their own fabs. They just find a fab house with good processes and work with them to get what they need. Rick
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-10-02 16:06 +0200 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) |
| Message-ID | <3062314.ndAJHGLEtv@sunwukong.fritz.box> |
| In reply to | #15855 |
rickman wrote: > That is the ONLY way AMD can hope to survive in Intel's market. AMD > is only around at this point because they got well over a billion > dollars from Intel a couple of years ago. It's the other way round. Intel's unfair practises did cost AMD a fortune (in the early Athlon64 days, AMD had the better processor, the better price, the better ISA, the better everything - and yet, Intel managed to keep their market share), and the one billion dollar compensation was "next to nothing". > Since then they have steadily gone > down hill from what I have seen. I don't expect them to be around in > five years. They will end up being sold off, likely in pieces and > Intel will have to worry about the FTC. They already have sold parts - they have sold their fab. And they had to bite the bullet and buy ATI, which was also quite expensive, but absolutely necessary (Intel shows how difficult it is to make a good GPGPU architecture - they can't do it. They will probably have to buy Nvidia to be competitive). At the moment, AMD is doing better than before, they almost make a profit. > Other companies don't necessarily need their own fabs. Fabless seems > to > be doing well for Xilinx. Have they built any fabs yet? Isn't Altera > also fabless these days? I'm not a close observer of semicoductor > companies these days, but I don't think the FPGA industry spends a lot > of time worrying about building their own fabs. They just find a fab > house with good processes and work with them to get what they need. All of them are fabless. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
| From | gavino_himself <visploveslisp@gmail.com> |
|---|---|
| Date | 2012-10-05 14:50 -0700 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) |
| Message-ID | <aed3d233-f112-4162-bc8e-deb72ee9d952@googlegroups.com> |
| In reply to | #15860 |
On Tuesday, October 2, 2012 7:06:22 AM UTC-7, Bernd Paysan wrote: > rickman wrote: > > > That is the ONLY way AMD can hope to survive in Intel's market. AMD > > > is only around at this point because they got well over a billion > > > dollars from Intel a couple of years ago. > > > > It's the other way round. Intel's unfair practises did cost AMD a > > fortune (in the early Athlon64 days, AMD had the better processor, the > > better price, the better ISA, the better everything - and yet, Intel > > managed to keep their market share), and the one billion dollar > > compensation was "next to nothing". > > > > > Since then they have steadily gone > > > down hill from what I have seen. I don't expect them to be around in > > > five years. They will end up being sold off, likely in pieces and > > > Intel will have to worry about the FTC. > > > > They already have sold parts - they have sold their fab. And they had > > to bite the bullet and buy ATI, which was also quite expensive, but > > absolutely necessary (Intel shows how difficult it is to make a good > > GPGPU architecture - they can't do it. They will probably have to buy > > Nvidia to be competitive). At the moment, AMD is doing better than > > before, they almost make a profit. > > > > > Other companies don't necessarily need their own fabs. Fabless seems > > > to > > > be doing well for Xilinx. Have they built any fabs yet? Isn't Altera > > > also fabless these days? I'm not a close observer of semicoductor > > > companies these days, but I don't think the FPGA industry spends a lot > > > of time worrying about building their own fabs. They just find a fab > > > house with good processes and work with them to get what they need. > > > > All of them are fabless. > > > > -- > > Bernd Paysan > > "If you want it done right, you have to do it yourself" > > http://bernd-paysan.de/ I love truth!! intel are bastards viva amd bernd when will your cpu be running a pc like machine and run firefox? for cheap? I will buy one!
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-10-06 03:14 +0200 |
| Subject | Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied) |
| Message-ID | <4737962.07d4RoZhpt@sunwukong.fritz.box> |
| In reply to | #15958 |
gavino_himself wrote: > bernd when will your cpu be running a pc like machine and run firefox? > for cheap? I will buy one! When you will understand the difference between a deeply embedded microcontroller and a PC CPU, then you will be able to buy b16 for cheap. Either that or you will understand why it doesn't run Firefox. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
| From | Bernd Paysan <bernd.paysan@gmx.de> |
|---|---|
| Date | 2012-09-30 16:34 +0200 |
| Message-ID | <10709389.OQkU8GVx6R@sunwukong.fritz.box> |
| In reply to | #15820 |
Rod Pemberton wrote: > Does NVidia produce microprocessors? No ... They produce GPU's. > They're different. And what's the Tegra line? Never heard of it? And FYI: We are discussion a project that is mostly competing with GPGPUs, not at all with a typical general purpose microprocessor. It's also not using 16 ARM cores. Rod, it would be less of a waste of time if you just followed the link the thread-starter provided. They also explain what they want their money for. They actually have two goals, and yes, I think their funding request is a bit high. > It's typical for Bernd to support non-facts as facts. It's a typical Rod to stick to his weird understanding of the world and not let any facts intrude it. Dismissing everything outside your tiny little box as non-fact, etc. I know why Hugh pushes his hate into clf, but I don't know why you do. Can you get a real live? Go out for a walk and step into some dog droppings to have a real reason for hating the world? > AFAIK, I don't lump and dump all processing devices into one category > as you do. Yes, I know, you twist words into different meanings so that you can be "right" in a discussion by providing your twisted definition of words so that everybody else must be wrong. You deliberately do the best you can to misunderstand people. >> But as Moore's law has an >> exponential cost in fabs, they get fewer and fewer. Therefore, >> fabless is the future. > > Fabless is a _temporary_ solution driven by cost pressures of building > a new fab. No one wants to pay the large upfront cost anymore. But fabs will become more expensive with every generation. How is that going ot be "temporary"? Do you expect fabs to become cheap again? >> Even AMD is now fabless. The years where they couldn't >> fill their expensive fabs nearly killed them. > > Do they allow others to use their fabs? (No.) Yes, they do. They sold it to GlobalFoundries, and it's free for everyone. > So, effectively the fab > becomes a subsidiary via exclusive, contractual use. In exchange for > the fab fronting the upfront build cost of the fab for their > customers, their fab is paid for over time at a higher total cost to > the customer than the upfront cost to build in-house. If you can sell enough of your chips to fill your fab, it is worth the cost. AMD can't - fab costs are essentially up-front costs, so you have to utilize it fully. Even Intel now started to let others use their fab; these projects are still tiny, and probably more about learning how to become a fab than to utilizing the fabs completely. Intel will face this problem in future, too. The output per fab grows faster than the total output Intel produces, and once they only need one fab for producing all their stuff, they will have to open up. They better start learning now how to do that; Intel is not stupid, they won't risk their business by doing it too late. Back to parallel computing: The fact that people can't write programs that utilize massive parallel computing has only changed slightly. We now can use that for supercomputing (number crunching) and graphics rendering. There are still a lot of programs which don't scale well. But the tasks that do scale well are important enough to sell massive parallel GPGPUs in high volume. And they don't only sell into gamer PCs. I'm *not* sceptical about the usefulness of GPGPUs. Not at all. They sell, they are useful, they do awesome stuff, people sort-of understand how to program them (in OpenCL and GLSL, which makes use of the vast amount of parallelism). They have somewhat balanced memory access (the main problem is to get the data from the host processor to the GPGPU, the GPGPU itself is more balanced). They have enough internal memory. They are a compromise, as they are optimized for gaming, not for computing. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://bernd-paysan.de/
[toc] | [prev] | [next] | [standalone]
Page 3 of 5 — ← Prev page 1 2 [3] 4 5 Next page →
Back to top | Article view | comp.lang.forth
csiph-web