Re: OT: Wheeeeee! New PSpice Benchmarks
From: Keith Williams (krw_at_att.bizzzz)
Date: 08/24/04
- Next message: Andrew Tweddle: "Re: RF2126 model"
- Previous message: Active8: "Re: Wheeeeee! New PSpice Benchmarks"
- In reply to: Ken Smith: "Re: OT: Wheeeeee! New PSpice Benchmarks"
- Next in thread: Jim Thompson: "Re: OT: Wheeeeee! New PSpice Benchmarks - LTSpice Data Added"
- Messages sorted by: [ date ] [ thread ]
Date: Tue, 24 Aug 2004 13:24:07 -0400
In article <cgfl2e$6k5$1@blue.rahul.net>, kensmith@green.rahul.net
says...
> In article <mn.c1f97d4831c16dbd.15428@spamyourselfagainstawall>,
> /* frank */ <__frank__@despammed.com> wrote:
> >Dopo dura riflessione, Jim Thompson ha scritto :
> >
> >> 2.175 times faster !!!
> >
> >Faster time - surely - isn't due to 64 bit:
> >nor WinXP neither pspice.exe are 64 bit ;)
>
> You may be wrong. A 64 bit bus can transfer more instructions per cycle
> and move a "double" in one stroke. Since the software has to do a lot of
> both, it could be the 64 bit nature of the machine that helped to gain the
> speed.
Even the lowly Pentium-1 (P5) had a 64bit bus. Memory can only be
referenced a cache-line/sector at a time. You want one bit? You're
getting 255 more where that one came from. ;-)
> It could also be like the new navigation system we just installed. They
> got a 25% increase in performance just by changing to a synthetic oil on
> the relative bearings.
The difference is quite likely do to the integrated DRAM controllers on
the Athlon64/Opteron. The memory latency is significantly less than it
would be with a northbridge attached memory system.
-- Keith
- Next message: Andrew Tweddle: "Re: RF2126 model"
- Previous message: Active8: "Re: Wheeeeee! New PSpice Benchmarks"
- In reply to: Ken Smith: "Re: OT: Wheeeeee! New PSpice Benchmarks"
- Next in thread: Jim Thompson: "Re: OT: Wheeeeee! New PSpice Benchmarks - LTSpice Data Added"
- Messages sorted by: [ date ] [ thread ]