[rescue] SMP on intel wasteful?

[Brian Hechinger]

On Tue, Jun 25, 2002 at 10:12:06AM -0700, David Passmore wrote:
> 
> > A took a <$2,000 PeeCee running Linux (a machine that would now be
> > chipped according to the standards of Mr. McGuire's recycling friend)
> > and a fully populated Sun E450.  I ran a number of EDA tools from
> > Synopsys, most notably VCS and Design Compiler.  The PeeCee averaged
> 
> [snip]
> 
> Saw that. How you're drawing a generalization of 'PCs whip UNIX boxes' ass',
> I don't know.

i wasn't going to reply to this, but since i want to comment on something else
you said, i may as well take the opourtunity.

he compares a <$2000 PC, to an E450 doing compute work.  first of all, the E450
was not designed to do compute work.  it's got big beefy IO balls for doing
nice database style beefy IO.  you do *NOT* need big Mhz CPUs to do that job,
no matter what Intel tells you.

second of all, how old is the E450 now?  quite old.  why doesn't he try
comparing apples to apples and get a current generation UltraSPARC-III machine?

since we're at it, why don't i make a sweeping generalization myself.  "PCs
whip PDP-11 boxes' asses!!!!"

> Can you produce uptime numbers?

i can produce big silly uptime numbers on sparc hardware no sweat. ;)

> Okay, I'll stick my penis out there; I've worked with F5, with Excite at Home,
> with Digex, and now with AOL. There are lots of dedicated UNIX boxes behind
> load-balancers, but in the enterprise, that's the /only/ place you'll find
> PC hardware.

but the only reason they are behind load balancers is not due to reliability.
it's due mostly to the way the application scales.  some applications don't
scale well on big SMP boxes, but due to stateless nature scale very well on
farms.

> Then why the thread pushing PCs as the end-all be-all of computing? I can
> only assume you like to piss people off unnecessarily.

he's been lurking on this list he says.  if he has been, and has truely been
paying attention, he knew EXACTLY what reaction his comments would get.  so
he's either clueless or he's a troll.

> It is exactly the slower components that will produce the most contention on
> a bus, no matter how fast it is. When you have a single bus, when a slow
> device is transmitting data, it hogs the bus. With a circuit switched
> architecture, other devices are free to transmit at the same time.

i thought Sun dumped circuit switching in favor of packet switching?  like a
million years ago even.  i could be wrong though, i am fuzzy on this point
after all.

(yes, this is all i wanted to comment on, heh)

-brian
-- 
"I mean Twinkies are good but getting shot really hurts."
				-- http://www.thisisatastyburger.com/ --