 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : BASIC Speed Benchmark Tests
  Page 2 of 6   |
|||||
| Author | Message | ||||
MicroBlocks Guru  Joined: 12/05/2012 Location: ThailandPosts: 2209 |
And when using cfunctions.:) (Probably need an extra decimal place.) Microblocks. Build with logic. |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 8573 |
Results for benchmark on 168MHz STM32F407VG running MMBasic BM1 0.011 BM2 0.075 BM3 0.119 BM4 0.128 BM5 0.211 BM6 0.32 BM7 0.505 BM8 0.165 |
||||
darthmite Senior Member  Joined: 20/11/2011 Location: FrancePosts: 240 |
Well , i just try this benchmark on stm32f429 ... BM1 BM2 BM3 BM4 BM5 BM6 BM7 BM8 ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 C-Language (code see above): Arduino UNO 0.010 0.010 0.058 0.043 0.043 0.043 0.045 0.284 Arduino DUE 0.003 0.003 0.006 0.007 0.007 0.007 0.106 0.014 stm32f429 bm1 0.000 022 "C-Language" bm2 0.000 022 bm3 0.000 050 bm4 0.000 022 bm5 0.000 023 bm6 0.000 089 bm7 0.000 334 bm8 0.001 188 So , it appear that the chip is very fast  Theory is when we know everything but nothing work ... Practice is when everything work but no one know why ;) |
||||
| WhiteWizzard Guru Joined: 05/04/2013 Location: United KingdomPosts: 2794 |
Wow impressive stuff- are you 100% sure of these figures Fabrice? Also can you let me know how much current the basic circuit pulls when achieving these times! Thanks WW For everything Micromite visit micromite.org Direct Email: whitewizzard@micromite.o |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 3650 |
The C compiler may have optimised some or almost all of the code away. Maybe worth posting the C? Someone may try RPi, A20, etc... John |
||||
| darthmite Senior Member Joined: 20/11/2011 Location: FrancePosts: 240 |
for the 100% sure ... look at the picture
Now how much current it draw ... no clue ... here i have a TFT , Wii joystick , sd card , ps2 keyboard + mouse connected  ...
for the sin , cos function i use the FPU , it ... help  ...
Cheers. Theory is when we know everything but nothing work ... Practice is when everything work but no one know why ;) |
||||
| darthmite Senior Member Joined: 20/11/2011 Location: FrancePosts: 240 |
Hi , i just compiled for st429 with the matherp last code (little change). Here just the frequency adjusted and put it on a st429 disco board. BM1 BM2 BM3 BM4 BM5 BM6 BM7 BM8 ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 st429 180Mhz 0.012 0.073 0.117 0.125 0.207 0.315 0.497 0.162 Now , why is it not faster ? ... he he ... The FPU is not active , no sdram , dma2d ... nothing , just like the st407 for this test
Cheers. Theory is when we know everything but nothing work ... Practice is when everything work but no one know why ;) |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 8573 |
Just done a bit of performance tweaking on the 168MHz STM32F407 New times added below 
BM1 BM2 BM3 BM4 BM5 BM6 BM7 BM8 ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 st429 180Mhz 0.012 0.073 0.117 0.125 0.207 0.315 0.497 0.162 st407 168Mhz 0.010 0.067 0.104 0.111 0.195 0.290 0.443 0.146 |
||||
BobD Guru Joined: 07/12/2011 Location: AustraliaPosts: 935 |
I notice that one of the benchmarks above is with line numbers. If the Micromite MMBasic still has line numbers then, surely, it is about time to drop them. We have been far less pragmatic about maintaining other features and compatibility. Line numbers are inefficient, lead to very poor coding practices, and "hard to follow" code. Space seems to be always at a premium in the flash. If they still exist then this may be a way get gain a bit of space. Thoughts anyone? |
||||
Grogster Admin Group  Joined: 31/12/2012 Location: New ZealandPosts: 9059 |
As a previous user of line-numbers in a big way, let me say that I agree with Bob on this one. I have not really used line-numbers for quite a while now, and no line-numbers is actually easier to program in then with line-numbers, IMHO. Once you get the hang of structured programming and labels, line-numbers become un-necessary. Again - MHO only. 
But this vote to drop line-numbers is coming from someone who wrote all his code with line-numbers for years, but now I don't, and I can see the advantages now, whereas before, I would not consider a language that did NOT have line-numbers. ...oh, how I was led astray... Smoke makes things work. When the smoke gets out, it stops! |
||||
TassyJim Guru Joined: 07/08/2011 Location: AustraliaPosts: 5897 |
There are a lot of old (but good) programs with line numbers available for translation to MMBasic. Having line numbers might help the cause. In MMEdit I intend to keep the routines for removing line numbers to help the updating. The routine for adding line numbers has not kept up with the changes to MM Basic and going in that direction is less likely to be successful. Having the benchmarks that prove that no line numbers is faster will help persuade a few. Jim VK7JH MMedit MMBasic Help |
||||
geraldfryjr Regular Member  Joined: 02/03/2014 Location: United StatesPosts: 61 |
It is great to see the ARMmite added to the list of Benchmark tests!!! :D jer :) Keep on DIYin' !!!  |
||||
Geoffg Guru Joined: 06/06/2011 Location: AustraliaPosts: 3165 |
Out of interest I tried the PIC32MX470 running at 120MHz: BM1 BM2 BM3 BM4 BM5 BM6 BM7 BM8 ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 st429 180Mhz 0.012 0.073 0.117 0.125 0.207 0.315 0.497 0.162 st407 168Mhz 0.010 0.067 0.104 0.111 0.195 0.290 0.443 0.146 MX470 120MHz MMBasic 0.013 0.082 0.135 0.135 0.265 0.380 0.597 0.174 The MX470 MMBasic is an internal version of mine that is different from matherp's port (it runs at 120MHz for a start). It is interesting to note that the performance of the high speed processors like the MX470 and STM32F407 does not scale directly with the clock speed. This is because the CPU is more limited by the speed at which it can read the flash memory rather than its clock speed. Benchmarks BM3 and above depend heavily on floating point calculations and you would have expected the STM32F407 with a floating point processor to be MUCH faster than the MX470 but the difference is not that great. It must be because the overhead of the interpreter is the dominating factor rather than the speed of calculating numbers. Just why the BM5 benchmark running on a 80MHz "Maximite without line Nos" processor is faster than an ARM chip running at 180MHz is a mystery. As they say... "lies, damn lies and benchmarks". Geoff Geoff Graham - http://geoffg.net |
||||
| aargee Senior Member Joined: 21/08/2008 Location: AustraliaPosts: 255 |
Very interesting, Geoff. I'm spending way too much time playing with different variants of the 'mite that I'm not getting any projects done. (... actually, nothing new there...) 
For crying out loud, all I wanted to do was flash this blasted LED. |
||||
| darthmite Senior Member Joined: 20/11/2011 Location: FrancePosts: 240 |
Hi , As primary test of MMBasic on the STM32F7 discovery board i have want see if it was fast enough. BM1 BM2 BM3 BM4 BM5 BM6 BM7 BM8 ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 st429 180Mhz 0.012 0.073 0.117 0.125 0.207 0.315 0.497 0.162 st407 168Mhz 0.010 0.067 0.104 0.111 0.195 0.290 0.443 0.146 MX470 120MHz MMBasic 0.013 0.082 0.135 0.135 0.265 0.380 0.597 0.174 st7 200Mhz 0.005 0.041 0.06 0.065 0.11 0.161 0.254 0.103 Not bad at all
It look like if we have a original PIC32 at 200Mhz , the result will be very close to the st7 chip ... You will get more info about the alpha release to the Uwe Becker (UB) page. stm32f7 MMBasic port I just send him the timer integration ,so you will get this release soon. The project is compiled with System Workbench for stm32. here the picture from the benchmark. 
Cheers. Theory is when we know everything but nothing work ... Practice is when everything work but no one know why ;) |
||||
| isochronic Guru Joined: 21/01/2012 Location: AustraliaPosts: 689 |
There is a newish pic32MZ "EF" variant , with a FPU ... don't know much bout it yet but the datasheet is : datasheet pic3MZxxEF |
||||
| darthmite Senior Member Joined: 20/11/2011 Location: FrancePosts: 240 |
Hi
Graphics with stm32f7 dma2d work great in MMBasic. Resolution is 480 x 272 in 16bits color (The normal stm32f7 4.3" discovery screen) The original random number was a little slow , i then use now the integrated RNG that work a little faster.But as you will see in the benchmark , without random we get double pixel / second ... New instruction i want made in future are the double and triple buffer. With dma2d you can copy one buffer (480x272) to another in 1.7mS and will be useful if you like to program games
Here are the preliminary results : 1000 Clear screen with random color Benchmark 1751 mS : 571.102 per seconds 10000 random Empty Circle Benchmark 1134 mS : 8818.34 per seconds 10000 random Filled Circle Benchmark 3874 mS : 2581.31 per seconds 10000 random Empty Box Benchmark 1411 mS : 7087.17 per seconds 10000 random Filled Box Benchmark 3113 mS : 3212.34 per seconds 10000 random Line Benchmark 1231 mS : 8123.48 per seconds 10000 random Pixel Benchmark 636 mS : 15723.3 per seconds No random used now 1000 Clear screen 1738 mS : 575.374 per seconds 10000 Empty Circle Benchmark 586 mS : 17064.8 per seconds 10000 Filled Circle Benchmark 2803 mS : 3567.61 per seconds 10000 Empty Box Benchmark 912 mS : 10964.9 per seconds 10000 Filled Box Benchmark 2607 mS : 3835.83 per seconds 10000 Line Benchmark 718 mS : 13927.6 per seconds 10000 Pixel Benchmark 319 mS : 31348 per seconds As usual , you will get more info on the UB page : stm32f7 maximite port. Cheers. Theory is when we know everything but nothing work ... Practice is when everything work but no one know why ;) |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 8573 |
I've updated the chart with the data from the Micromite 5.2 code running on a PIC32MZ2048EFG064. As you can see it is faster than everything apart from the ST7 and that is despite calculating in double precision floating points rather than single. The ST chips have hardware floating point like the PIC32MZ ABC 800 single precision 0.9 1.8 6.0 5.9 6.3 11.6 19.6 2.9 ABC 800 double precision 1.2 2.2 10.0 10.6 11.0 17.8 26.4 14.4 IBM PC 1.5 5.2 12.1 12.6 13.6 23.5 37.4 3.5 Apple III 1.7 7.2 13.5 14.5 16.0 27.0 42.5 7.5 VIC-20 1.4 8.3 15.5 17.1 18.3 27.2 42.7 9.9 ZX81 in "fast mode" 4.5 6.9 16.4 15.8 18.6 49.7 68.5 22.9 Maximite 0.016 0.144 0.196 0.205 0.354 0.512 0.721 0.310 Maximite with line Nos 0.016 0.131 0.193 0.194 0.245 0.393 0.582 0.241 Maximite without line Nos 0.016 0.111 0.173 0.173 0.192 0.336 0.525 0.220 MicroMite 40MHz 0.028 0.18 0.285 0.289 0.644 0.892 1.346 0.376 MicroMite 48MHz 0.023 0.15 0.237 0.24 0.536 0.744 1.121 0.313 st429 180Mhz 0.012 0.073 0.117 0.125 0.207 0.315 0.497 0.162 st407 168Mhz 0.010 0.067 0.104 0.111 0.195 0.290 0.443 0.146 MX470 120MHz MMBasic 0.013 0.082 0.135 0.135 0.265 0.380 0.597 0.174 st7 200Mhz 0.005 0.041 0.06 0.065 0.11 0.161 0.254 0.103 Micromite EXTREME 198MHz 0.006 0.045 0.07 0.08 0.141 0.201 0.287 0.126 |
||||
kiiid Guru Joined: 11/05/2013 Location: United KingdomPosts: 671 |
Can someone give me this benchmark program so I can run it on my port, please? http://rittle.org -------------- |
||||
| twofingers Guru Joined: 02/06/2014 Location: GermanyPosts: 1133 |
See page 1 regards Michael |
||||
| Page 2 of 6 |
|||||
 Print this page |
