Re: P2 asm & K6 asm / execution differences

Graham TerMarsch (gtermars@home.com)
Wed, 30 Dec 1998 10:52:10 -0800

Kurt Garloff wrote:
> On Tue, Dec 29, 1998 at 02:18:44PM -0500, Aaron Tiensivu wrote:
> > This compares the decode/execute engines of the K6 & PII.
> > PII is a faster decoder and K6 is a faster executor.
> >
> > http://www.cetinc.com/hardware/articles/x86/k6-3_indepth.html
>
> The reason might be the K6 was first designed by NexGen, and I don't know,
> if, at the beginning, they headed for decoding ix86 assembly instructions or
> have their own assembly language (or any other, like m68k).
> So while the execution units have undergone very careful design, the decode
> unit might have been made with a shorter development time.

If I recall correctly, the NexGen guys had headed towards decoding x86
instructions into their own microcode and executing that. Quite a similar
design to what everyone is doing nowadays with their x86 decoders on some sort
of RISC core. NexGen had it's pros and cons though; memory access was
seriously fast on it (my NexGen-90Mhz outperformed my Intel P-120), but it did
lack in other areas. Seems that they traded off some of the time to
enhance/perfect certain components by having other components that
outperformed most everything else. Standard trade-off I guess.

-- 
Graham TerMarsch


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/