|
PureBytes Links
Trading Reference Links
|
Well, halleluiah!! I'm back on again... seems my domain server was a
naughty boy with some spam (hijacked was more likely) and was locked out for
a while there...
I had been trying to add to the discussion re: homebrew systems
and which CPU to use... following is a copy of that original email, which I
hope will be helpful.
Regards,
Gene Pope
> ----------------------------------
>
> Hello Jim and Ross,
>
> You raise a **very** interesting issue. I just home-built a "best of
Intel"
> machine with 2.54 gig Pent.4 with 533 frontside bus and fastest RDRAM
> possible... Now, if you read the "reviews", including MoBo reviews, this
is
> the one setup that is supposed to give AMD indigestion.
>
> Ran a performance test suite on three systems, 1) My old Athlon 1.2 gig
with
> DDR, 2) This new P4 puppy, 3) A baseline test on an Athlon 1800+ that was
> included. All with 500 megs.
>
> Results were interesting. If you only looked at CPU performance for
> add/subtract/mult etc. the P4 did kick butt.
>
> But floating point division was a different story. The P4 was slightly
> *worse* than my 1.2 gig Athlon, and the 1800 smacked it smartly.
>
> Maximum MegaFLOPS was just as lopsided. The P4 edged out the 1.2 Athlon,
but
> once again, the 1800 whalloped the best P4 on the market.
>
> As for memory read/write (where the 533 FSB and RDRAM is supposed to
shine),
> the P4 did trounce the competition in the basic Read/Write cached and
> uncached tests **but** was beaten by *both* Athlons under the "Allocate
> Small Block" memory test.
>
> Finally, I too booted up the SETI program, and was more than a little
> disappointed at the increase in speed (I see roughly a 20% increase...
> maybe). I remember that compared to a dual Xeon P2 machine, my old Athlon
> could run SETI at ***three times*** the speed... This latest "upgrade" in
> raw processing power using the P4 is completely underwhelming.
>
> Methinks all this talk about AMD getting it's tail clipped is way too
> premature. In fact, I'm beginning to suspect good 'ol Intel of tweaking
it's
> designs to only outperform in certain test cases which it can "advertise"
> about,
> with AMD cranking hard where it really counts.
>
> Reminds me of when 3D video boards started tweeking their designs to only
> test as fast as possible on Quake and Unreal, but with disappointing
results
> under real world
> conditions.
>
> And finally, the damn P4 MoBo/RDRAM combination is a LOT more expensive
than
> the fastest Athlon/DDR MoBo.
>
> Live and loin.
>
> Best regards,
>
> Gene Pope
ADDENDUM 9/1/02
I also added 1066 RDRAM to the P4 equation. What this did was to give a
significant jump to, of all things, disk access and read/write speeds...
most everything else tested the same... go figure.
>
>
> ----- Original Message -----
> From: "Ross S Bond" <ross.bond@xxxxxxxxxxxxx>
> To: "Jim" <g67j3sq245qlo@xxxxxxxxxxx>
> Cc: <omega-list@xxxxxxxxxx>
> Sent: Monday, July 22, 2002 1:32 AM
> Subject: Re: OT:Which processor?
>
>
> > Hello Jim,
> >
> > I'm a layman too, and my Athalon 1.57GB sure seems a lot quicker
> > to use than the P4 2.0Gb I have access to. (Roughly equivalent memory)
> >
> > One of my ISP's is in the SETI program where they do large off-peak
> > data-processing runs and he has a dual P4 beast - he says his single
> > Athalon machine does the SETI runs noticeably quicker.
> > (SETI - Search for Extra Terrestrial Intelligence)
> >
> > No idea of the technical reasons though, sorry.
> >
> > --
> > Best regards,
> > Ross mail to: ross.bond@xxxxxxxxxxxxx
> >
> > Sunday, July 21, 2002, 9:47:49 PM, you wrote:
> >
> > J> No doubt this is obvious to many but not to me:-
> >
> > J> Would someone briefly explain in laymans terms why it matters which
> type of
> > J> processor (pentium/celeron/athlon) is in use if their quoted speeds
are
> the
> > J> same.
> >
> > J> TIA
> >
> > J> Jim
> >
> >
>
>
>
|