Re: a dozen cpu's on a chip

On Mon, 12 May 2008 09:33:47 -0400, krw <krw@xxxxxxxxxxxxxxxx> wrote:

In article <3iea249phks4ht3694942e80gmabvba5qa@xxxxxxx>,
jjlarkin@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx says...
On Sat, 10 May 2008 00:04:22 -0400, krw <krw@xxxxxxxxxxxxxxxx> wrote:

In article <9f2a24lb3qdd4fplttffo6oarcbgqc952v@xxxxxxx>,
jjlarkin@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx says...
On Fri, 9 May 2008 22:27:56 -0400, krw <krw@xxxxxxxxxxxxxxxx> wrote:

In article <3n6624pu6762nup9apu3crj5vh1uu6fqbn@xxxxxxx>,
jjlarkin@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx says...
On Thu, 8 May 2008 07:42:04 -0700 (PDT), MooseFET <kensmith@xxxxxxxxx>
wrote:

On May 7, 7:48 pm, John Larkin
<jjlar...@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
http://www.eetimes.com/news/latest/showArticle.jhtml;jsessionid=CESEX...

I bet we'll see 256 one of these days.

When you get to large numbers of CPUs it seems to make sense to stop
making them identical. For servers this would be doubly so. Many of
the CPUs won't need to do floating point operations.

Right. Amybe a few cpu's would have serious floating point power, or a
few separate fp engines could be assigned to cpu's as needed. Lots of
cpu's, doing stuff like file i/o or serial stuff, could be less
powerful. I suppose we'll always need special graphics hardware, but
just a few of those per chip.

Asymmetric multiprocessing makes the scheduler's life more
complicated. Since the scheduler is part of the OS, and the OS is
most often M$, this isn't a good idea, IMO. ;-) Hardware is cheap
(so cheap PowerPC is including decimal FPUs). Throw the FPU on
every node, whether its needed or not.

It also would make sense to do things like memory moves in the "Memory
Mismanagement Unit" since the values don't need to be modified on the
way through.

This will make it a lot harder to say how many CPUs are in a chip. If
there is only as much hardware as 200 full CPUs but 500 threads can be
running at the same time, do you call it 200 or 500 CPUs.

Next step is to get rid of task swapping and threads altogether. One
CPU is the OS, and one cpu gets assigned per process.

Which negates what you say above. Running a task, then getting an
exception because you don't have an instruction you thought you had
is expensive.

Why would you get an exception? If a device driver doesn't need fp
opcodes, run it on one of the many cpu's that doesn't have floating
point. And vice versa. <> rocket science.

You're making your scheduler's job more difficult and limiting
flexibility. Computer architecture is rocket surgery.

A bunch of cpu's don't need scheduling like a single-processor os
does; individual cpu's do their thing concurrently and set semaphores,
and go idle, if they finish whatever they are assigned to do.

ONLY if the CPUs have tasks assigned when the system is designed.
IOW, this may work for embedded processors but not general purpose
computing.

And
besides, the task manager cpu doesn't have anything else to do. The
scheduler will mostly set up things like memory management and
priviliges and assignments and turn them loose, rather than
frantically swapping them in real time. When everything runs
simultaneously, priorities become less important. It's a whole new way
of thinking.

No, it's really not new, rather rejected.

The IBM Cell chip is an architecture that trends in that direction.

Now you sound like Dimbulb. ;-) Note that the Cell processor is
essentially an embedded processor. The tasks it has been designed
for are quite limited.

I said it "trends in that direction", not that it was the ultimate
architecture. But why is one PPC plus six simpler integer processors
"quite limited"? It's obviously more general than the PPC alone.




Current hardware and software has been driven by Intel's silicon
process skill (and their vicious lack of ethics) and by Microsoft's
thousands of programmers (and their vicious lack of ethics) but not by
any particularly intelligent planning. Most big software apps are
spinning-out-of-control crapware with gigabyte service packs just
pushing the bugs around. It's time for a change, for the next
generation of computing, and I think it will happen when there are so
many processors on a chip that multitasking quits making sense.

No, you have it backwards. Intel has been driven by hardware for
the past thirty years.

I believe that's just what I said. They have just pushed an 8008
architecture - a dog when it was new - into nanometer silicon. Their
attempts at cutting over to more modern architectures - iapx32, Arm,
Itanic - have been expensive failures.

As many processors as we're likely to see,
there will always be more tasks/threads.

Why? What would a desktop PC need with 1024 threads?


A new language wouldn't hurt either.

Well, we can certainly agree there, though not likely on what. ;-)

Something more like Cobol, where programmers are forced to deal with
the application, rather than using the application as a platform to
show off how tricky they can be. Something without pointers. Something
that is impossible to crash.

John

.

Relevant Pages

  • Re: Reliably finding general CPU architecture on Windows XP and newer?
    ... For the actual CPU architecture, I've done some further digging through the actual DMTF docs and I think this is a problem that can be generally solved with a very thorough check of the CPU numeric ID. ... Future CPUs from AMD or Intel that are AMD64 are going to contain the AMD64 or Intel 64 string in their actual names, and Itanium architecture CPUs are almost certainly going to contain Itanium or IA64 descriptors. ... It looks like I use AddressWidth and Architecture for the hardware and PROCESSER_ARCHITECTURE for the software. ...
    (microsoft.public.win32.programmer.wmi)
  • Re: a dozen cpus on a chip
    ... ;-) Hardware is cheap. ... This will make it a lot harder to say how many CPUs are in a chip. ... Computer architecture is rocket surgery. ... priviliges and assignments and turn them loose, ...
    (sci.electronics.design)
  • Re: Reliably finding general CPU architecture on Windows XP and newer?
    ... It looks like I use AddressWidth and Architecture for the hardware and PROCESSER_ARCHITECTURE for the software. ... The problem with using Family is that although I _could_ map documented CPUs to a particular architecture, the list is always slightly obsolete - and any tool would then have no way to categorize new numbers returned by Family. ...
    (microsoft.public.win32.programmer.wmi)
  • Re: a dozen cpus on a chip
    ... When you get to large numbers of CPUs it seems to make sense to stop ... I suppose we'll always need special graphics hardware, ... priviliges and assignments and turn them loose, ... generation of computing, and I think it will happen when there are so ...
    (sci.electronics.design)
  • Re: a dozen cpus on a chip
    ... When you get to large numbers of CPUs it seems to make sense to stop ... I suppose we'll always need special graphics hardware, ... This will make it a lot harder to say how many CPUs are in a chip. ... priviliges and assignments and turn them loose, ...
    (sci.electronics.design)