|
PureBytes Links
Trading Reference Links
|
It's basically the same old formula. Unless the application is
compiled using a new 64-bit capable compiler, you're not going to
see much of a speed improvement. This is where the Athlon has
the advantage. I don't know how many apps will be
recompiled/rewritten into 64-bits, but I doubt there will be many
(at least for the next few years). Most software vendors don't
like producing/warehousing different versions of the same
program, unless it's for different O/S's. Will MSFT have
separate 64-bit versions of its Office Suites? Will Equis write
a separate 64-bit version of MetaStock? I doubt it will be
anytime soon. Video editing and drawing titles will certainly
have 64-bit versions, but I don't see the multitude of companies
jumping into this. Perhaps they will have both 32-bit and 64-bit
versions on the same CD and it will autoinstall the correct one.
Still, unless there's a huge performance advantage, and ***I have
read it isn't that much***, companies aren't going to pay the
large programming labor cost to write 2 different versions.
Daniel.
"A.J. Maas" wrote:
> > I understand that using 64 bits versus the current 32 bits
> will not increase performance much. Performance improvements by
> just using the old 512 kb Pipe Line Burst ofcourse will be very
> noticible, butaccording to the Industry, much larger than this
> type of Burst Line systems (eg into the MB's are "build andon
> their way"). Windows2000 is build for 64-bit technology and is
> also downside compatible with 'older' 32-bit systems.Overall,
> it will be much like the utmost advantage "32-bit programs" had
> over the (very memory consuming andvery irritating + error
> bugged) "16-bit programs". > I doubt Itanium will match
> Athlon's speed and will probably be more expensive to boot.
> Also, the newer INTC CPU's require MUCH more expensive
> RDRAM.Intel processors handle ALL of the Windows (+Win2000 or
> Linux) OS "requests", something that cannot truely be saidfor
> the AMD's. You can ask any modem/printer/scanner
> manufacturers.(not much use having the horsepower aboard if it
> cannot handle half of its instructions, eg doesn't know what to
> do with it)Intels high-speed RAMBUS Direct RAM chips are
> focused on supporting(and are therefore in return also
> supported by)the "Super Fast" 667 and 700 MHz versions
> (code-named "Coppermine") that will debut in mid-October. Both
> designsfeature a 0.18-micron process that utilizes Intel's new
> 133 MHz system bus and the 820 chipset.> BTW, AMD will be
> releasing 1,000 Mhz Athlons in Q3 of 2000. Incredible isn't
> it? Running a 1,000 Mhz CPU will soon be possibleMid-2000 the
> first Intels 64-bits Itaniums are due for commercial release
> (ranging from 700 MHz to 1000MHz (eg 1 GHz))and currently a
> PIII-800 MHz (for 32-bits systems) is in the RC1-stages, a
> version prior to RC2 (usualy THE release candidate)and due for
> commercial release early next year. This Mid-October should see
> the final release of the 667 and 700 MHz versions(code-named
> "Coppermine", see above).
> Regards,
> Ton Maas
> ms-irb@xxxxxxxxxxxxxxxx
> Dismiss the ".nospam" bit (including the dot) when replying and
>
> note the new address change. Also for my Homepage
> http://home.planet.nl/~anthmaas
>
> ----- Original Message -----
> From:Daniel Martinez
> To: metastock@xxxxxxxxxxxxx
> Sent: vrijdag 8 oktober 1999 19:17
> Subject: Re: Windows 2000 + Intel's 64-bit Itanium
> (was Merced)
> Hello,
> I understand that using 64 bits versus the current 32
> bits will not increase performance much. Because
> Athlon is a true seventh generation CPU, if you want
> superior performance, wait until next year when the
> second generation Athlon motherboards come out and
> get this combo. I doubt Itanium will match Athlon's
> speed and will probably be more expensive to boot.
> Also, the newer INTC CPU's require MUCH more
> expensive RDRAM.
>
> BTW, AMD will be releasing 1,000 Mhz Athlons in Q3 of
> 2000. Incredible isn't it? Running a 1,000 Mhz CPU
> will soon be possible.
>
> Daniel.
>
|