Of interest, eventually? (IBM opening ppc h/w design)
Here is IBM's announcement:
http://www-306.ibm.com/chips/news/2004/0331_power.html
and Wired's blurb on it:
http://www.wired.com/news/infostructure/0,1377,62885,00.html?tw=wn_tophead_2
What would you do if you could design ideal h/w support for python or a python VM? For memory allocation/garbage collection? Monitoring/detecting dynamic conditions triggering recompilation/re-optimization etc? Position-independent code and structures? Threading/context switching/timeouts/atomic queue stuff/etc? Profiling support at the hardware level? Crypto stuff? Sandboxing/capability support/etc?
How would pypy adapt-to/have-an-advantage-in-exploiting the existence of such a configurable h/w platform?
Anyway, just thought I'd pass on an interesting blurb maybe to keep in mind.
Regards, Bengt Richter
On 2004 Apr 01, at 00:07, Bengt Richter wrote:
Here is IBM's announcement:
Wow.
How would pypy adapt-to/have-an-advantage-in-exploiting the existence of such a configurable h/w platform?
First of all, it seems to me, by having a generator for PPC machine code -- which, it seems, was one of the things dropped from our projects due to reduced funding:-(. With PPC (in the form of plain unadorned G5) pulsing at the core of supercomputers ("Big Mac" at Virginia Tech), IBM mainframes, Apple 1U servers, down to high-end Apple PCs, and soon, probably, laptops, and also a serious contender for the next generation of gaming consoles (apparently both Sony and Microsoft are considering it), _and_ now even potentially "extensible" in such ways, I think that if I had to nominate just one chip architecture to generate machine code for it would have to be PPC (the fact that I'm a recent and enthusiastic Mac owner is not really related, except in stemming partly from the same reason -- I've revisited PPC architecture in the wake of all the recent hoopla about G5 and it appears that there _is_ a lot to like about it). Ah well -- I do realize that, with our limited resources, the near ubiquitousness of the [expletive deleted] descendants of Intel 80386 makes it most likely impractical and impolitical to eschew supporting their bedraggled complexity in favour of PPC's clean and powerful extensibility:-(.
Alex
On Apr 1, 2004, at 1:53 PM, Alex Martelli wrote:
On 2004 Apr 01, at 00:07, Bengt Richter wrote:
Here is IBM's announcement:
Wow.
How would pypy adapt-to/have-an-advantage-in-exploiting the existence of such a configurable h/w platform?
First of all, it seems to me, by having a generator for PPC machine code -- which, it seems, was one of the things dropped from our projects due to reduced funding:-(. With PPC (in the form of plain unadorned G5) pulsing at the core of supercomputers ("Big Mac" at Virginia Tech), IBM mainframes, Apple 1U servers, down to high-end Apple PCs, and soon, probably, laptops, and also a serious contender for the next generation of gaming consoles (apparently both Sony and Microsoft are considering it), _and_ now even potentially "extensible" in such ways, I think that if I had to nominate just one chip architecture to generate machine code for it would have to be PPC (the fact that I'm a recent and enthusiastic Mac owner is not really related, except in stemming partly from the same reason -- I've revisited PPC architecture in the wake of all the recent hoopla about G5 and it appears that there _is_ a lot to like about it). Ah well -- I do realize that, with our limited resources, the near ubiquitousness of the [expletive deleted] descendants of Intel 80386 makes it most likely impractical and impolitical to eschew supporting their bedraggled complexity in favour of PPC's clean and powerful extensibility:-(.
I'm sure a PPC code generator will get done eventually, even if I have to do it myself on my own time :)
-bob
Bob Ippolito bob@redivi.com writes:
I'm sure a PPC code generator will get done eventually, even if I have to do it myself on my own time :)
You might get some help with that...
Cheers, mwh
participants (4)
-
Alex Martelli -
Bengt Richter -
Bob Ippolito -
Michael Hudson