Motorola 5200 MIPS

Motorola 5200 MIPS

[Mark Chambers]

I've gotten the linux 2.4.24 5200 build running on the Motorola Lite5200
evaluation board.  Linux reports 263.78 BogoMIPS.  Motorola advertises 760
MIPs for the 5200.  Is this just mucho-bogo-promo on Motorola's part or am I
missing something?

Mark Chambers


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Motorola 5200 MIPS

[Gary Thomas]

On Fri, 2004-04-30 at 16:25, Mark Chambers wrote:
> I've gotten the linux 2.4.24 5200 build running on the Motorola Lite5200
> evaluation board.  Linux reports 263.78 BogoMIPS.  Motorola advertises 760
> MIPs for the 5200.  Is this just mucho-bogo-promo on Motorola's part or am I
> missing something?

BogoMIPS != CPU MIPS

--
Gary Thomas <gary@mlbassoc.com>
MLB Associates


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Motorola 5200 MIPS

[Gerrit Van de Velde]

Don't know, but you're not comparing the right numbers anyway. Try
running a Dhrystone benchmark. It reported something around 500 MIPS.
Bogomips is something different: it's the number of times a processor
can execute "nothing", hence that name: bogo, from bogus :)

> Is this just mucho-bogo-promo on Motorola's part or am I
>missing something?
>
>Mark Chambers
>
>
>
>
>.
>
>
>


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Motorola 5200 MIPS

[Mark Chambers]

> Don't know, but you're not comparing the right numbers anyway. Try
> running a Dhrystone benchmark. It reported something around 500 MIPS.

I suspect 500MIPS is probably pretty close, judging by some of the tests
I've run.

OK, I know I'm anal-retentive, but I can't just accept "it has 'bogo' in the
name, so therefore it means nothing."  We've run linux on several 8xx family
parts and the bogomips closely tracked the expected performance of 1
instruction/clock.  Now all of a sudden it doesn't mean anything?  So I
tracked it down, and the bogoMIPS calculation all boils down to a "bdnz"
instruction (on a PPC).  So on pipelined processors branches are the worst
case scenario.  I think it boils down to taking 2 processor cycles/bdnz
instruction.  So, it's a worst case scenario, but then again 760MIPS is a
best case scenario, assuming you can always feed two instructions into the
CPU every clock cycle (and never have to branch!).  Not going to happen, so
you've got to assume you can reasonably expect 1-point-something times the
processor clock.

Mark Chambers
Microfirst, Inc.

P.S. I'm a fan of Motorola processors, hope I didn't sound too ugly.  What
would the world be like if Bill had chosen the 68000???


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Motorola 5200 MIPS

[Wolfgang Denk]

In message <002101c4311c$6ca7d440$0301a8c0@chuck2> you wrote:
>
> OK, I know I'm anal-retentive, but I can't just accept "it has 'bogo' in the
> name, so therefore it means nothing."  We've run linux on several 8xx family

Then go back and read the archives. This has been  discussed  several
times before.

> parts and the bogomips closely tracked the expected performance of 1
> instruction/clock.  Now all of a sudden it doesn't mean anything?  So I

It never meant anything.

You can compare values  within  the  same  processor  family,  i.  e.
MPC8xx, but not across different families which may use - for example
- a completely different bus interface.

> P.S. I'm a fan of Motorola processors, hope I didn't sound too ugly.  What
> would the world be like if Bill had chosen the 68000???

There would be a so-called "operating system" which  was  to  be  68k
compatible and full of bugs ;-)

Best regards,

Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
The ideal situation is to have massive computing power right at home.
Something that dims the streetlights and shrinks the picture  on  the
neighbours' TVs when you boot it up.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Motorola 5200 MIPS

[Mark Chambers]

> Then go back and read the archives. This has been  discussed  several
> times before.
>

An excellent idea.  Why didn't I think of that?  And to all of you fellow
nerds back there who also like to know how things work and why - bless your
heart.  Actually, I don't fully follow the explanations, but clearly the 603
core is not able to execute a single looped dbnz once per clock, so 263
BogoMIPS is a reasonable number for a 396MHz 5200.  The number does not have
anything to do with bus speed, which is part of what was bothering me (2 x
132MHz, you know).  I won't pester you all with this any more but if anybody
else wants to discuss further I'd welcome your emails  (not discuss BogoMIPS
necessarily, but details of 5200 performance - I've got some very
challenging requirements to hit and any insights will be welcome).

Mark Chambers


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/