Re: Non Intel & AMD Arch

From: Matt Taylor (para_at_tampabay.rr.com)
Date: 01/15/04 

Date: Thu, 15 Jan 2004 02:15:35 +0000 (UTC)

"Bryan Parkoff" <bryan.nospam.parkoff@nospam.com> wrote in message
news:rwiNb.23384$RV5.6712@fe2.texas.rr.com...
> Please state your opinion what you think Intel & AMD arch are better.
> What about other non Intel & AMD arch? Little Endian is faster than Big
> Endian on Intel & AMD arch. Why do other processors choose to use Big
> Endian? How can they benefit the performance?

Are you talking about specific processors or architectures (e.g. IA-32,
IA-64, AMD-64, Alpha, MIPS, etc.)? I've always been impressed with Athlon.
Athlon is like a P6-core processor with better execution bandwidth and no
essentially decode constraints.

Dealing with big endian isn't slow on x86 processors. The bswap instruction
was added to the 486 to allow x86 to switch between little & big endian.
This is one of the simplest ALU instructions, and it's fast (1-2 cycles) on
most processors.

> Can you claim which is the most popular -- Intel & AMD, PowerPC, or
> other processors?

Intel has somewhere around 90% of the PC market.

> Most modern processors use RISC, but IA32 and AMD use CISC. It looks
> like that there are only two Intel and AMD CISC processors, but there are
so
> many other processors that use RISC. Why do Intel decide to design IA64
> that uses RISC instead of CISC. It is what Intel Technician told me that
> IA64 remains at RISC only.

Modern x86 processors borrow a lot of ideas from RISC processors. They are
RISC processors with some special hardware to decompose x86 instructions
into internal RISC instructions.

IA-64 is not really RISC. It is based off of VLIW which is an extension of
RISC philosophy. VLIW code has groups ("molecules") of instructions
("atoms") which are executed simultaneously. In RISC architectures like
MIPS, SPARC, or Alpha, the CPU determines at runtime which instructions can
execute in parallel. In a VLIW architecture that information is encoded into
the instruction set. Intel calls IA-64 an EPIC design. EPIC is like VLIW,
but it allows molecules to be variable-length. I don't know of any
differences aside from that.

> Why do Intel claim that CISC is better than RISC for game compability
> because games on RISC can be much slower than CISC like 386 at 20 MHz.

Obviously non-x86 processors are not compatible with games targetted for
x86. Since the PC is based on x86, most games are targetted to x86. Other
systems have to emulate.

Also, I don't think many people play games on 20 MHz computers anymore.

-Matt

Relevant Pages

  • Re: What does it mean
    ... It's a RISC core, served by microcode decoding x86 ... instructions. ...
    (rec.games.computer.ultima.dragons)
  • Re: CISC vs RISC concepts -- from an assembly view
    ... RISC is better, hands down. ... CISC chips have become so cheap. ... original PC design and Intel took the effort to keep the x86 ... of the plethora of available software for the x86. ...
    (comp.lang.asm.x86)
  • Re: Intels 6-core processor codenamed "Dunnington" ...
    ... I read somewhere that Intel might be better at marketing than ... Pentium-4 being a RISC processor but we all know that those SIMD ... instructions into a series of RISC instructions. ... which use CISC instructions descended from the 8086 processor. ...
    (comp.os.vms)
  • Re: AMD vs Intel for video format conversions and editing
    ... Linux) compiler rather than the Intel one. ... basic x86 code rather than the fancy high-performance extended x86 ... they've spent a lot of effort in getting their basic x86 instructions ...
    (comp.sys.ibm.pc.hardware.chips)
  • Re: What does it mean
    ... It's a RISC core, served by microcode decoding x86 ... To me, if something accepts x86 instructions, it is not RISC. ...
    (rec.games.computer.ultima.dragons)