[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
PPRAM Consortium Foundation and role of 1355
Dear Peter Thompson,
As I told you before my marketing strategy plans for 1355 are
gradually working.
PPRAM (Parallel Processing Random Access Memory) Consortium Foundation
was officially announced on 20th of January collaborating with many
universities and governmental institute.
Here I attached a brief explanation of this consortium.
A simple definition of PPRAM is the parallel/distributed architecture
of integrating:-
PPRAM := { PROCESSOR + DRAM + PPRAM-Link }
PROCESSOR := DSP | GRAPHICS | MPU | MPU++ | MPEG2 | whatever
with using shared memory and network technology.
Beside several small chips (maybe 2-4) can be coexisted inside the
same chip if needed.
A goal of this design is to realize a high performance system with
using communicating architecture. So the key issues are to find the
low cost and high speed serial link technology.
This seems to become a transputer-like architecture when occam is
ported. Or maybe T9000 must be reproduced from Japanese company !! I
discussed the possibility of porting occam to one Professor at Tokyo
Institute of Technology. He had a strong interest of re-trargeting
occam to PPRAM chip. But it takes more time to decide which
programming language is suitable for PPRAM.
Now the consortium has started to look at several STDs such as IEEE
1394, 1596(SCI) and 1355(HIC). But they already know what are the
benefits of using 1355 instead of 1394, 1596. So most of consortium
people are focusing to 1355 technology as a PPRAM-Link. (But some are
still sticking on SCI.)
As I am requested from Dr. Murakami who is a leader of this project
(at Kyushu University) to support 1355 technology when designing PPRAM
chip, I think I can lead these industrial people to the right way.
Probably we need to have a joint international conference of HIC and
Parallel Prpcessing to stimulate this consortium sometime in the
future.
This consortium continues to work on finalizing specifications of
PPRAM-Link and API(application interface) by the end of this year. And
final decision of making chip will be made by themselves not me.
Consortium only gives a guideline of PPRAM architecture to all
members.
Targeted marketing area is consumer area but not super computer area.
The other thing I must give appreciation to Paul Walker is that he
will kindly submit a general article of 1355 technology for Nikkei
Electronics. (I think deadline of an article is the end of January.)
This article probably will be a trigger of inspiring Japanese people.
I have already spoken to consortium people. So they are really
expecting to read his article !! I will try to get many copies when
issued for the further distribution. Then 1355 technology will be
well known more and more in Japan.
Regards,
Kazuto Matsui
----------------------------------------------------------------------
Currently 11 companies are joining. In future this number will be
increased more.
<< Industry >>
1. NEC Corp.
2. OKI Electric Industry Co.(Multimedia LSI Business Development
Centre)
3. TOSHIBA (System ULSI Engineering Laboratory)
4. SONY (Media Processing Lab)
5. Fuji Xerox (Advanced Technology & Engineering Centre)
6. Matsushita Electric Co. (Semiconductor Research Lab)
7. Mitsubishi Electric Co.(System LSI Development Lab)
8. Fujitsu (System LSI Development Lab)
9. NTT (System Electronics Co.)
10. Hitachi
11. Samsung Electronics Co.
<< Academia >>
1. Keio University
2. Tohoku University
3. Waseda University
4. Kyushu Institute of Technology
5. University of Tokyo
6. Kyoto University
7. University of Electro-Communications
8. Ochanomizu University
9. University of Tsukuba
10. Kyushu University
11. Tokyo Institute of Technology
<< Governmental Institute >>
1. RWCP
2. ISIT