| Poster | Thread |
 AmigaBlitter
|  |
Re: Cell roadmap
Posted on 2-Nov-2006 16:04:23
| | [ #1 ] |
|
|
 |
Elite Member
 |
Joined: 26-Sep-2005
Posts: 3514
From: Unknown | | |
|
| 
What about an Amiga Cell machine?
_________________
retired
|
|
| Status: Offline |
|
|
Fransexy
| |
Re: Cell roadmap
Posted on 2-Nov-2006 16:11:07
| | [ #2 ] |
|
|
 |
Elite Member
|
Joined: 8-Jun-2004
Posts: 2334
From: Elche (Alicante), spain | | |
|
| Quote:
| What about an Amiga Cell machine? |
I love the idea of multicore Cell Cpu but then....  it wouldn´t be a cell anymore, it would be a "tissue" CPU 
_________________
No PowerPC, No Fun
Make Amiga Great Again
|
|
| Status: Offline |
|
|
elwood
| |
Re: Cell roadmap
Posted on 2-Nov-2006 18:47:44
| | [ #3 ] |
|
|
 |
Elite Member
|
Joined: 17-Sep-2003
Posts: 3428
From: Lyon, France | | |
|
| @fransexy
LOL
_________________
Philippe 'Elwood' Ferrucci
Sam460 1.10 Ghz
AmigaOS 4 betatester
Amiga Translator Organisation
|
|
| Status: Offline |
|
|
ne_one
| |
Re: Cell roadmap
Posted on 2-Nov-2006 21:23:19
| | [ #4 ] |
|
|
 |
Cult Member
 |
Joined: 13-Jun-2005
Posts: 905
From: Unknown | | |
|
| Anyone have any idea how the Cell architecture differs from the PPC? Is it a completely different technology or is some of the PPC core retained?
And... for general-purpose computing how does it compare? |
|
| Status: Offline |
|
|
yak
| |
Re: Cell roadmap
Posted on 3-Nov-2006 1:15:50
| | [ #5 ] |
|
|
 |
Regular Member
 |
Joined: 15-Mar-2006
Posts: 322
From: Bochum, Germany | | |
|
| @ ne_one
According to Wikipedia:
"In a simple analysis the Cell processor can be split into four components: external input and output structures, the main processor called the Power Processing Element (PPE) (a two-way SMT multithreaded Power 970 architecture compliant core), eight fully-functional co-processors called the Synergystic Processing Elements or SPEs and a specialised high-bandwidth circular data bus connecting the PPE, input/output elements and the SPEs, called the Element Interconnect Bus or EIB."
The "problem" is that the main source of the power of a Cell processor are the SPEs, so the code has to be optimised for Cell's architecture to use its full potential. |
|
| Status: Offline |
|
|
ne_one
| |
Re: Cell roadmap
Posted on 3-Nov-2006 5:05:11
| | [ #6 ] |
|
|
|
Cult Member
|
Joined: 13-Jun-2005
Posts: 905
From: Unknown | | |
|
| Interesting... I vaguely remember someone mentioning the Friedens had a take on what would be involved to support the chip.
Could be interesting. |
|
| Status: Offline |
|
|
Fransexy
| |
Re: Cell roadmap
Posted on 4-Nov-2006 10:59:41
| | [ #7 ] |
|
|
|
Elite Member
|
Joined: 8-Jun-2004
Posts: 2334
From: Elche (Alicante), spain | | |
|
| Quote:
| The "problem" is that the main source of the power of a Cell processor are the SPEs, so the code has to be optimised for Cell's architecture to use its full potential |
And what is the diference to Altivec unit, MMX, SSE..............? if your code is not optimised to use these "extra units" you don´t use the full potentialLast edited by Fransexy on 04-Nov-2006 at 11:04 AM.
_________________
No PowerPC, No Fun
Make Amiga Great Again
|
|
| Status: Offline |
|
|
28 crawler(s) on-line.
125 guest(s) on-line.
0 member(s) on-line.
