[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [avr-gcc-list] Whetstone Benchmark
From: |
Weddington, Eric |
Subject: |
RE: [avr-gcc-list] Whetstone Benchmark |
Date: |
Sun, 20 Jan 2008 18:00:06 -0700 |
> -----Original Message-----
> From:
> address@hidden
> [mailto:address@hidden
> org] On Behalf Of Dmitry K.
> Sent: Sunday, January 20, 2008 3:12 PM
> To: address@hidden
> Subject: [avr-gcc-list] Whetstone Benchmark
>
> Hi.
>
> A table from 'TI Competitive Benchmark': slaa205b.pdf (July 2006).
> Whetstone Code Size (in bytes) and Cycle Counts.
> IAR C Compiler used (4.12A for AVR), full optimization.
>
> MCU Code size Cycles
> ------------ --------- ------
> MSP430FG4619 6544 107040
> MSP430F149 6524 109837
> PIC24FJ128GA 4605 108619
> 8051 8723 291836
> H8/300H 4656 205910
> MaxQ20 5392 158945
> ARM7TDMI/Thumb 10532 60444
> HCS12 7370 787635
> ATmega8 4694 270991
>
> avr-gcc 4.1.2 & avr-libc CVS HEAD (Jan 2008) gives:
> MCU Code size Cycles Optimization
> ------------ --------- ------ ------------
> ATmega8 4020 88563 -Os -morder1
> ATmega8 4030 88565 -Os
> ATmega8 5310 71743 -O3 -morder1
> ATmega8 5332 71751 -O3
>
> The dummy functions are added to prevent GCC optimization.
>
> I am not sure in TI's accuracy. Results for both MSP430-es
> are practicaly equal, despite the fact thar FG4619 has
> considerable enhanced core. Also run time for ATmega8 is
> too big. Can everbody to make any measurements with
> IAR compiler?
This is fantastic! Thank you so much for doing this!
A couple of quick comments/suggestions:
- Please drop -morder1. We generally don't recommend this to users, and
it looks like it really doesn't affect the results that much
- Can you also try with avr-gc 4.2.2 and avr-libc 1.6.1? These are the
latest released versions.
Thanks,
Eric Weddington