| 1 |
On Friday 13 January 2006 07:45, Francesco Riosa wrote: |
| 2 |
> Tom Smith wrote: |
| 3 |
> > Well, if they're /not/ mutually exclusive, another question that comes |
| 4 |
> > up is... |
| 5 |
> > |
| 6 |
> > If a program is compiled with sse or sse2 support on a Pentium II, will |
| 7 |
> > the program run slower than it otherwise would? (Some of the programs I |
| 8 |
> > have are compiled and then distributed to servers with different |
| 9 |
> > CPUs--P-IIs and P-IVs, mainly.) |
| 10 |
> |
| 11 |
> speaking of manually added options to CFLAGS*, not of use flags |
| 12 |
> |
| 13 |
> The only place where mathematics count on a server is encryption ? |
| 14 |
> (notice the question mark) |
| 15 |
> Mayor part of server software use integer math that are not so enhanced |
| 16 |
> by optimizations. |
| 17 |
> The code produced is less stable, and difficult to debug, this bring to the |
| 18 |
> question: why take the risk ? |
| 19 |
actually, mmx (MultiMedia eXtensions) , sse and sse2 instructions are designed |
| 20 |
primarily for multimedia and gaming type applications, which _do_ use |
| 21 |
floating-point math, and AFAIK, encryption is going to be all-integer too |
| 22 |
(floating-point math is not perfectly precise) |
| 23 |
|
| 24 |
And, like I said earlier, if you put a program with an sse or sse2 instruction |
| 25 |
on a PII, the program will most likely spontaneously abort when it tries to |
| 26 |
execute the unsupported instruction. |
| 27 |
-- |
| 28 |
# |
| 29 |
# electronerd, the electronerdian from electronerdia |
| 30 |
# |
Archives