Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.forth > #15795 > unrolled thread

Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)

Started by"B.Person" <bruce.person@gmail.com>
First post2012-09-28 07:16 -0700
Last post2012-10-27 08:54 -0700
Articles 20 on this page of 97 — 19 participants

Back to article view | Back to comp.lang.forth


Contents

  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 1 of 5  [1] 2 3 4 5  Next page →


#15795 — Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)

From"B.Person" <bruce.person@gmail.com>
Date2012-09-28 07:16 -0700
SubjectYour next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)
Message-ID<76fe0697-97b3-4844-9b18-b2b86e028d8c@googlegroups.com>
Posted 16 hours ago.  Time to get into the ARM world; IMHO.

http://arstechnica.com/information-technology/2012/09/99-raspberry-pi-sized-supercomputer-touted-in-kickstarter-project/

[toc] | [next] | [standalone]


#15801

Fromjacko <jackokring@gmail.com>
Date2012-09-28 10:44 -0700
Message-ID<a34f030b-63ba-4753-8a48-869dfee506f1@googlegroups.com>
In reply to#15795
Doesn't sound too bad. I have written some Java code for a soft implementation of a DSP core with 1KB memory. I wonder if this idea is extended with some memory exchange interconnect using a time division multiplex bus an external co-ordination processor using the memory, could calculate the TDM flux, and then build the correct software in each 1KB cell. With 1GB memory sticks common, this is 1M * 20 cycle processors as an estimate. Neural net simulation? I assume silicon density can exceed neuron density by quite a margin.

https://github.com/jackokring/jarvar/blob/master/jarvar-base/src/uk/co/peopleandroid/jarvar/internals/SoftDSP.java

Cheers Jacko

[toc] | [prev] | [next] | [standalone]


#15802 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)

Fromjacko <jackokring@gmail.com>
Date2012-09-28 10:45 -0700
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)
Message-ID<ef3c731e-e4a7-423b-8de8-3eec13caef94@googlegroups.com>
In reply to#15801
Sky Net is upon us...

[toc] | [prev] | [next] | [standalone]


#15808 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)

FromMark Wills <forthfreak@gmail.com>
Date2012-09-28 11:42 -0700
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip. (link supplied)
Message-ID<1677708b-c824-4476-8480-5dba1e90a1e9@hj1g2000vbb.googlegroups.com>
In reply to#15801
On Sep 28, 6:44 pm, jacko <jackokr...@gmail.com> wrote:
> Doesn't sound too bad. I have written some Java code for a soft implementation of a DSP core with 1KB memory. I wonder if this idea is extended with some memory exchange interconnect using a time division multiplex bus an external co-ordination processor using the memory, could calculate the TDM flux, and then build the correct software in each 1KB cell. With 1GB memory sticks common, this is 1M * 20 cycle processors as an estimate. Neural net simulation? I assume silicon density can exceed neuron density by quite a margin.
>
> https://github.com/jackokring/jarvar/blob/master/jarvar-base/src/uk/c...
>
> Cheers Jacko

It's shared memory with up to 128K per node.


    Complete multicore solution featuring a high performance
microprocessor ISA, Network-On-Chip, and distributed memory system
    Fully-featured ANSI-C programmable GNU/Eclipse based tool chain
    Scalable to 1000’s of cores and TFLOPS of performance on a single
chip
    1GHz superscalar RISC processor cores
    IEEE Floating Point Instruction Set
    Shared memory architecture with up to 128KB memory at each
processor node
    Zero startup-cost messaging passing
    Vector Interrupt Controller
    Distributed Multicore Multidimensional DMAs
    32 GB/sec local memory bandwidth per core
    8GB/sec per processor network bandwidth
    72 GFLOPS/Watt energy efficiency
    Processor tile size of 0.5mm^2 at 65nm, 0.128mm^2 at 28nm


From http://www.adapteva.com/products/epiphany-ip/epiphany-architecture-ip/

Awesome.

[toc] | [prev] | [next] | [standalone]


#15807

FromJecel <jecel@merlintec.com>
Date2012-09-28 11:39 -0700
Message-ID<959f4162-731d-4f94-b53a-86f14da4b686@googlegroups.com>
In reply to#15795
On Friday, September 28, 2012 11:16:25 AM UTC-3, B.Person wrote:
> Posted 16 hours ago.  Time to get into the ARM world; IMHO.

The 16 cores are not ARM, though they do talk to an ARM processor.

-- Jecel

[toc] | [prev] | [next] | [standalone]


#15809

FromPaul Rubin <no.email@nospam.invalid>
Date2012-09-28 11:44 -0700
Message-ID<7xwqzekmwk.fsf@ruckus.brouhaha.com>
In reply to#15795
"B.Person" <bruce.person@gmail.com> writes:
> Posted 16 hours ago.  Time to get into the ARM world; IMHO.

I looked at that.  It looked interesting at first, but meh.

I don't think they claimed to have ARM cores on the array chip.  They
are some unidentified 32-bit risc architecture, maybe ARM, maybe not.
(There is an architecture manual but I didn't look at it).  The board
does have an ARM running Linux and you might be thinking of that.  The
array chip is a peripheral to the ARM chip, operating as a coprocesssor
or accelerator.  Basically you are looking at a Raspberry Pi with an
external, programmable GPU.  

The array chip has limited memory and i/o bandwidth, the usual
bottleneck of these things.  Its total computational performance is
nowhere near that of a desktop GPU, though of course its power
consumption is much lower, making it perhaps of interest for mobile
products.  And instead of SIMD, it's a bunch of semi-independent
processors sort of like the recently announced Xeon Phi.  For the
16-core chip there is 0.5MB of on-chip NUMA ram, I guess 32k "owned" by
each core, with lower speed access possible from other cores. 

[toc] | [prev] | [next] | [standalone]


#15815

From"Rod Pemberton" <do_not_have@notemailnotz.cmm>
Date2012-09-29 14:55 -0400
Message-ID<k47g06$ntp$1@speranza.aioe.org>
In reply to#15795
"B.Person" <bruce.person@gmail.com> wrote in message
news:76fe0697-97b3-4844-9b18-b2b86e028d8c@googlegroups.com...

And, another OT thread ...

> Posted 16 hours ago.  Time to get into the ARM world; IMHO.
>

Why?

a) The project doesn't have any funding, yet.
b) There is no markete demand for it, yet.
c) It's not being produced by a major semiconductor company, i.e., it's got
the "kiss of death" already ...
d) It'll require some custom hardware, i.e., an additional expense.
e) There is no software for it.

> [link to proposed many core ARM]

I saw a slightly different link for it:
http://hackaday.com/2012/09/28/massively-parallel-64-core-computer-costs-99/

My thoughts when I saw that were:
"Just what advantage does this design have over the failed Transputer?"
http://en.wikipedia.org/wiki/Transputer

Not to burst anyone's bubble, but I saw nothing of any merit that would
offer this project _any_ chance at real success.

They've given themselves a year to produce the 16-core version.  That
seems a bit unrealistic.


Rod Pemberton

[toc] | [prev] | [next] | [standalone]


#15817

FromPaul Rubin <no.email@nospam.invalid>
Date2012-09-29 12:13 -0700
Message-ID<7xmx08hcbz.fsf@ruckus.brouhaha.com>
In reply to#15815
"Rod Pemberton" <do_not_have@notemailnotz.cmm> writes:
> a) The project doesn't have any funding, yet.

True, and they may be asking a bit much.

> b) There is no markete demand for it, yet.

Yeah, they are too hopeful and enamored of their design, it seems to me.

> 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.

> d) It'll require some custom hardware, i.e., an additional expense.

Just a Raspberry Pi-like PC board (plus the 64 core cpu if they do that,
I thought).

> 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.

> They've given themselves a year to produce the 16-core version.  That
> seems a bit unrealistic.

I thought they already had the 16-core silicon, and they were looking to
build a board and a 64-core chip.  Maybe I'm mistaken about the 16-core
silicon.  

[toc] | [prev] | [next] | [standalone]


#15818

FromBernd Paysan <bernd.paysan@gmx.de>
Date2012-09-29 23:13 +0200
Message-ID<2207084.uDy9Jnq5yQ@sunwukong.fritz.box>
In reply to#15817
Paul Rubin wrote:

> "Rod Pemberton" <do_not_have@notemailnotz.cmm> writes:
>> a) The project doesn't have any funding, yet.
> 
> True, and they may be asking a bit much.

AFAIK they already had a first round of founding for their 16 cores 
unit, with IIRC $2 million.

>> b) There is no markete demand for it, yet.
> 
> Yeah, they are too hopeful and enamored of their design, it seems to
> me.

Well, whatever startup you have, if you aren't overly hopeful and 
enamored, it won't go anywhere.  You absolutely have to love what you 
are doing.

>> 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... 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.  Even AMD is now fabless.  The years where they couldn't 
fill their expensive fabs nearly killed them.

>> 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 transputer failed for being too early.  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.

Note that this thing we are discussing here has a unified, but non-
uniform memory, you communicate by reading and writing remote memories.  
This is pretty conventional, so people can use it.

>> They've given themselves a year to produce the 16-core version.  That
>> seems a bit unrealistic.
> 
> I thought they already had the 16-core silicon, and they were looking
> to build a board and a 64-core chip.  Maybe I'm mistaken about the
> 16-core silicon.

It took him (as this was a one-man-show) 6 weeks to tapeout the 16-core 
silicon after he got the design software - he did invest 2 years of 
design time prior to that; probably using FPGA tools to evaluate his 
stuff.  The design is there, all they (they are now more than one man) 
need to do is to retarget the physical layout to a 28nm process.

Shouldn't take too long.

The design software of today is really in a much better shape than it 
used to be.  AMD turned away from hand-tuned hand-layouted everything, 
and their new Steamroller design looks like they are have recovered from 
that; they are now as good as the pre-Bulldozer hand layout used to be.

I'm not sure if this will be a tremendous success.  They aren't that 
much better than a GPGPU, and the GPGPU will be included into devices 
just because you need a GPU.  Current desktop GPUs are already GPGPUs 
(with OpenCL and alike as language), and the next generation mobile GPUs 
like Mali 600 will be GPGPUs, too.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#15819

FromPaul Rubin <no.email@nospam.invalid>
Date2012-09-29 17:30 -0700
Message-ID<7xsja0z70x.fsf@ruckus.brouhaha.com>
In reply to#15818
Bernd Paysan <bernd.paysan@gmx.de> writes:
>> True, and they may be asking a bit much.
> AFAIK they already had a first round of founding for their 16 cores 
> unit, with IIRC $2 million.

Making chips is expensive.  But once the chip is made, another $750k to
make a PC board with off-the-shelf parts around it?  Or is it mostly for
software porting, or what?

>> Yeah, they are too hopeful and enamored of their design, it seems to me.
> Well, whatever startup you have, if you aren't overly hopeful and 
> enamored, it won't go anywhere.  You absolutely have to love what you 
> are doing.

It also helps to have concrete evidence that what you're doing fills a
need, instead of being a solution looking for a problem.  This echoes
similar things you've said in the past about the GA chip.

> It took him (as this was a one-man-show) 6 weeks to tapeout the 16-core 
> silicon after he got the design software - he did invest 2 years of 
> design time prior to that; probably using FPGA tools to evaluate his 
> stuff.

May I ask what is involved in 2 years of something like that?  I'm
imagining a process like "download a 32-bit RISC core from
opencores.org, replicate 16 times, plus add some stuff for the shared
memory blocks".

> I'm not sure if this will be a tremendous success.  They aren't that 
> much better than a GPGPU, and the GPGPU will be included into devices 
> just because you need a GPU.  Current desktop GPUs are already GPGPUs 
> (with OpenCL and alike as language), and the next generation mobile GPUs 
> like Mali 600 will be GPGPUs, too.

I think the chip is targeted to mobile phones with little power
available.  The 16-core part's total commputational performance is lower
than today's desktop processors, and even the 64-core is not that much
ahead, especially if you count on-chip desktop GPU's.

I'd say though that gpgpu's are not so good at algorithms involving lots
of conditional branches, so these array thingies probably have uses.
The Xeon Phi looks interesting.

[toc] | [prev] | [next] | [standalone]


#15821

From"Elizabeth D. Rather" <erather@forth.com>
Date2012-09-29 17:59 -1000
Message-ID<vv2dnRdZzq0GXvrNnZ2dnUVZ_ridnZ2d@supernews.com>
In reply to#15819
On 9/29/12 2:30 PM, Paul Rubin wrote:
> Bernd Paysan <bernd.paysan@gmx.de> writes:
>>> True, and they may be asking a bit much.
>> AFAIK they already had a first round of founding for their 16 cores
>> unit, with IIRC $2 million.
>
> Making chips is expensive.  But once the chip is made, another $750k to
> make a PC board with off-the-shelf parts around it?  Or is it mostly for
> software porting, or what?

Don't forget marketing. It's expensive, and engineers often fail to 
appreciate both its expense and its necessity. Over the years I've been 
involved in many more projects that were marketing failures than 
technical failures.

>>> Yeah, they are too hopeful and enamored of their design, it seems to me.
>> Well, whatever startup you have, if you aren't overly hopeful and
>> enamored, it won't go anywhere.  You absolutely have to love what you
>> are doing.
>
> It also helps to have concrete evidence that what you're doing fills a
> need, instead of being a solution looking for a problem.  This echoes
> similar things you've said in the past about the GA chip.

Yes, that's important, and I'm afraid I also agree it's a problem with GA.

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]


#15825 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

FromAnonymous <nobody@remailer.paranoici.org>
Date2012-09-30 13:50 +0000
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<7f64f71d6400cd5cc142ea3ad85f1da6@remailer.paranoici.org>
In reply to#15821
"Elizabeth D. Rather" <erather@forth.com> wrote:

> On 9/29/12 2:30 PM, Paul Rubin wrote:
> > Bernd Paysan <bernd.paysan@gmx.de> writes:
> >>> True, and they may be asking a bit much.
> >> AFAIK they already had a first round of founding for their 16 cores
> >> unit, with IIRC $2 million.
> >
> > Making chips is expensive.  But once the chip is made, another $750k to
> > make a PC board with off-the-shelf parts around it?  Or is it mostly for
> > software porting, or what?
> 
> Don't forget marketing. It's expensive, and engineers often fail to 
> appreciate both its expense and its necessity. Over the years I've been 
> involved in many more projects that were marketing failures than 
> technical failures.

If that isn't a good description of Forth I don't know what is!









[toc] | [prev] | [next] | [standalone]


#15859 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

From"Ed" <invalid@nospam.com>
Date2012-10-02 23:53 +1000
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<k4eriv$r0e$1@speranza.aioe.org>
In reply to#15825
Anonymous wrote:
> "Elizabeth D. Rather" <erather@forth.com> wrote:
>
> > On 9/29/12 2:30 PM, Paul Rubin wrote:
> > > Bernd Paysan <bernd.paysan@gmx.de> writes:
> > >>> True, and they may be asking a bit much.
> > >> AFAIK they already had a first round of founding for their 16 cores
> > >> unit, with IIRC $2 million.
> > >
> > > Making chips is expensive.  But once the chip is made, another $750k to
> > > make a PC board with off-the-shelf parts around it?  Or is it mostly for
> > > software porting, or what?
> >
> > Don't forget marketing. It's expensive, and engineers often fail to
> > appreciate both its expense and its necessity. Over the years I've been
> > involved in many more projects that were marketing failures than
> > technical failures.
>
> If that isn't a good description of Forth I don't know what is!

Which - a technical or marketing failure?  :)

Paul's observation applies equally to Forth:

    "It also helps to have concrete evidence that what you're doing
    fills a need, instead of being a solution looking for a problem."

Marketing generates initial interest.  For it to be sustainable there
must be ongoing need.  How many programming languages does
the world need to sustain in any given era?  Not that many.


[toc] | [prev] | [next] | [standalone]


#15883 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

FromFritz Wuehler <fritz@spamexpire-201210.rodent.frell.theremailer.net>
Date2012-10-03 19:10 +0200
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<923e0c628f2cd64d742acdea342c91e8@msgid.frell.theremailer.net>
In reply to#15859
"Ed" <invalid@nospam.com> wrote:

> Anonymous wrote:
> > "Elizabeth D. Rather" <erather@forth.com> wrote:
> >
> > > On 9/29/12 2:30 PM, Paul Rubin wrote:
> > > > Bernd Paysan <bernd.paysan@gmx.de> writes:
> > > >>> True, and they may be asking a bit much.

> > > Don't forget marketing. It's expensive, and engineers often fail to
> > > appreciate both its expense and its necessity. Over the years I've been
> > > involved in many more projects that were marketing failures than
> > > technical failures.
> >
> > If that isn't a good description of Forth I don't know what is!
> 
> Which - a technical or marketing failure?  :)

Forth seems to be a good example of a marketing failure that isn't a
technical failure. It's a simple, flexible tool that just doesn't get any
respect!

> Paul's observation applies equally to Forth:
> 
>     "It also helps to have concrete evidence that what you're doing
>     fills a need, instead of being a solution looking for a problem."

Aside from talking about it, which is what this newsgroup is and should be
about, there are people using Forth or Elizabeth and Stephen probably
couldn't afford their internet connections and websites. Without any
additional context from your quote, I don't see Forth as a solution looking
for a problem. There are many niche tools and domain specific languages that
are not popular or widely known, yet they serve a purpose and are in active
use. 

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? 

> 
> Marketing generates initial interest.  For it to be sustainable there
> must be ongoing need.  How many programming languages does
> the world need to sustain in any given era?  Not that many.

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.

[toc] | [prev] | [next] | [standalone]


#15885 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

From"Elizabeth D. Rather" <erather@forth.com>
Date2012-10-03 07:41 -1000
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<EsKdnaBYMbFR5fHNnZ2dnUVZ_oOdnZ2d@supernews.com>
In reply to#15883
On 10/3/12 7:10 AM, Fritz Wuehler wrote:
> "Ed" <invalid@nospam.com> wrote:
>
>> Anonymous wrote:
>>> "Elizabeth D. Rather" <erather@forth.com> wrote:
...
>>>> Don't forget marketing. It's expensive, and engineers often fail to
>>>> appreciate both its expense and its necessity. Over the years I've been
>>>> involved in many more projects that were marketing failures than
>>>> technical failures.
>>>
>>> If that isn't a good description of Forth I don't know what is!
>>
>> Which - a technical or marketing failure?  :)
>
> Forth seems to be a good example of a marketing failure that isn't a
> technical failure. It's a simple, flexible tool that just doesn't get any
> respect!

I agree with that.

>> Paul's observation applies equally to Forth:
>>
>>      "It also helps to have concrete evidence that what you're doing
>>      fills a need, instead of being a solution looking for a problem."
>
> Aside from talking about it, which is what this newsgroup is and should be
> about, there are people using Forth or Elizabeth and Stephen probably
> couldn't afford their internet connections and websites. Without any
> additional context from your quote, I don't see Forth as a solution looking
> for a problem. There are many niche tools and domain specific languages that
> are not popular or widely known, yet they serve a purpose and are in active
> use.

Yes, Forth was designed from the ground up to address a particular 
real-world problem: the need for flexible programming for 
resource-constrained systems. The fact that it is still in use today, 
mostly in embedded systems, is a testimony to the fact that it is 
successful for that purpose, and that the need still exists. In the 
world of embedded processors, unit costs count a lot, and your ability 
to do more with whatever resources you have translates to success with 
your product.

The remarkable thing is that Forth has also been successful in 
environments that are not resource-constrained (e.g. PCs) because of its 
flexibility and ease of use.

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]


#15886 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

Fromvandys@vsta.org
Date2012-10-03 18:06 +0000
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<ad3d4eFm8fpU1@mid.individual.net>
In reply to#15885
> Forth seems to be a good example of a marketing failure that isn't a
> technical failure. It's a simple, flexible tool that just doesn't get any
> respect!

It's a powerful macro assembler for a stack based machine language.
The problem is that it can take your $0.23 problem and do it for $0.17--
a 25% savings!  But everybody has pockets stuffed full of $20 bills,
so nobody cares.  I think Chuck Moore was smart to head towards where
the pennies still matter.

-- 
Andy Valencia
Home page: http://www.vsta.org/andy/
To contact me: http://www.vsta.org/contact/andy.html

[toc] | [prev] | [next] | [standalone]


#15887 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

FromBernd Paysan <bernd.paysan@gmx.de>
Date2012-10-03 20:22 +0200
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<53974477.bJxicJU2s3@sunwukong.fritz.box>
In reply to#15886
vandys@vsta.org wrote:

>> Forth seems to be a good example of a marketing failure that isn't a
>> technical failure. It's a simple, flexible tool that just doesn't get
>> any respect!
> 
> It's a powerful macro assembler for a stack based machine language.

This is the low-level view of Forth.  Systems like my b16 cross compiler 
could fit into that description.  You miss the interactive part, and you 
miss the extensibility part - yes, maybe it's hidden in the word 
"powerful", but usually, I would describe other macro assemblers as 
"powerful", too, even though they are neither interactive nor allow me 
to extend them.

> The problem is that it can take your $0.23 problem and do it for
> $0.17--
> a 25% savings!  But everybody has pockets stuffed full of $20 bills,
> so nobody cares.

The perception I have of my 512MB 800MHz single core phone is that of a 
very resource constrained environment - when I run normal Dalvik 
programs.  Of course it isn't a resource constrained environment for a 
Forth system, it is like heaven - there is plenty of RAM, there is a 
fast CPU and GPU, etc.  But apparently, people programming in Java can 
make this overengineered thing very slow ang sluggish.  I *do* care.

I've heard that with 2GB and a 1.6GHz quadcore, Android will feel a lot 
better. m(

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]


#15912 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

Fromrickman <gnuarm@gmail.com>
Date2012-10-04 08:39 -0400
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<k4k033$43q$1@dont-email.me>
In reply to#15886
On 10/3/2012 2:06 PM, vandys@vsta.org wrote:
>> Forth seems to be a good example of a marketing failure that isn't a
>> technical failure. It's a simple, flexible tool that just doesn't get any
>> respect!
>
> It's a powerful macro assembler for a stack based machine language.
> The problem is that it can take your $0.23 problem and do it for $0.17--
> a 25% savings!  But everybody has pockets stuffed full of $20 bills,
> so nobody cares.  I think Chuck Moore was smart to head towards where
> the pennies still matter.

And exactly where would that be?

Rick

[toc] | [prev] | [next] | [standalone]


#15919 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

From"Elizabeth D. Rather" <erather@forth.com>
Date2012-10-04 07:51 -1000
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<1vCdnRlaMek9UfDNnZ2dnUVZ_rCdnZ2d@supernews.com>
In reply to#15912
On 10/4/12 2:39 AM, rickman wrote:
> On 10/3/2012 2:06 PM, vandys@vsta.org wrote:
>>> Forth seems to be a good example of a marketing failure that isn't a
>>> technical failure. It's a simple, flexible tool that just doesn't get
>>> any
>>> respect!
>>
>> It's a powerful macro assembler for a stack based machine language.
>> The problem is that it can take your $0.23 problem and do it for $0.17--
>> a 25% savings!  But everybody has pockets stuffed full of $20 bills,
>> so nobody cares.  I think Chuck Moore was smart to head towards where
>> the pennies still matter.
>
> And exactly where would that be?

In embedded devices that will be used in enormous volumes, such that 
tiny savings in unit costs are multiplied by volume to the extent that 
it's worth extra development costs to achieve them, and in applications 
where good power management can vastly extend the usefulness (and value) 
of a device.

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]


#15920 — Re: Your next supercomputer! Just $99 for 16 ARM cores on a chip.

Fromrickman <gnuarm@gmail.com>
Date2012-10-04 14:05 -0400
SubjectRe: Your next supercomputer! Just $99 for 16 ARM cores on a chip.
Message-ID<k4kj57$k2$1@dont-email.me>
In reply to#15919
On 10/4/2012 1:51 PM, Elizabeth D. Rather wrote:
> On 10/4/12 2:39 AM, rickman wrote:
>> On 10/3/2012 2:06 PM, vandys@vsta.org wrote:
>>>> Forth seems to be a good example of a marketing failure that isn't a
>>>> technical failure. It's a simple, flexible tool that just doesn't get
>>>> any
>>>> respect!
>>>
>>> It's a powerful macro assembler for a stack based machine language.
>>> The problem is that it can take your $0.23 problem and do it for $0.17--
>>> a 25% savings! But everybody has pockets stuffed full of $20 bills,
>>> so nobody cares. I think Chuck Moore was smart to head towards where
>>> the pennies still matter.
>>
>> And exactly where would that be?
>
> In embedded devices that will be used in enormous volumes, such that
> tiny savings in unit costs are multiplied by volume to the extent that
> it's worth extra development costs to achieve them, and in applications
> where good power management can vastly extend the usefulness (and value)
> of a device.

I asked the question not because I don't understand the concept, but 
because I don't see how a $15 CPU chip which requires some $10+ of 
supporting chips is a viable contender in the "$0.23" arena.  I realize 
that number is just a rhetorical figure, but with the current selling 
price of the GA144, I don't see it being suited to high volume markets 
except perhaps an extremely small niche.  I still am asking the question,

     Exactly where would that be?

Rick

[toc] | [prev] | [next] | [standalone]


Page 1 of 5  [1] 2 3 4 5  Next page →

Back to top | Article view | comp.lang.forth


csiph-web