24 Oct
2005
24 Oct
'05
11:10 a.m.
Rumor has it that Chuck Guzis may have mentioned these words:
On 10/23/2005 at 9:07 PM John Allain wrote:
Fast. Very fast. Well, at least in the x86 world, as long as the x86 code
is compiled 'just right' for it, but that's true for any CPU.
I have a Crusoe powered laptop, and Linux from Scratch runs like a raped
ape on the thang...
933Mhz Crusoe, runs Winders like a P3-650/700 or so, runs RedHat Enterprise
Linux 4 AS like a P3-600, runs LFS like a P3-1.3GHz. It just takes a month
to compile Linux w/everything one needs. ;-)
My simple answer...
On a 200Mhz Pentium, SimH running VMS feels like a MicroVAX II
(close to a 750). At 700Mhz, it feels like a MicroVax 3500. Basically
a directly proportional increase.
Okay, then how fast would one suppose that a Transmeta VLIW implementation
would run (replace the "X86 Code Morphing" ROM with your own version of a
"VAX emulation" ROM)? Would it most likely run faster than an FPGA
implmentation?
Most definitely. Think gigahertz VAX. That would be *kewl*. ;-)
If so, why bother with the FPGA?
Erm... because Transmeta's not making CPUs anymore? They found that they
can make more money with less overhead licensing their LongRun technology
to other chipmakers, than try to make and sell their chips against
(primarily) Intel and (secondarily) AMD.
And that's sad, because I *love* my Lifebook. (P2120, if ya gotta know. ;-)
Too bad it's YAIOT (Yet Another Intel-Orphaned Technology.)
Given that there was no
single VAX implemenation by DEC, how could this even offend the purists?
Depends on your definition of "impure." Needless to say (but I will anyway
:-) everyone's definition of that will be different.
Laterz,
Roger "Merch" Merchberger
--
Roger "Merch" Merchberger | "Profile, don't speculate."
SysAdmin, Iceberg Computers | Daniel J. Bernstein
zmerch at 30below.com |