Search found 107 matches

by Pito
Fri Jan 24, 2020 8:25 pm
Forum: Custom design boards
Topic: Bluepill F4 board, anyone still working on it?
Replies: 83
Views: 114812

Re: Bluepill F4 board, anyone still working on it?

And here is the Whetstone.c used above - it has to be ported to stm32duino..
by Pito
Fri Jan 24, 2020 7:55 pm
Forum: Custom design boards
Topic: Bluepill F4 board, anyone still working on it?
Replies: 83
Views: 114812

Re: Bluepill F4 board, anyone still working on it?

For example - here we did Whetstone with 80MHz pic32, double precision, no FPU, running off ram

http://retrobsd.org/viewtopic.php?f=6&t=4065

1.2Mflops DP.

With single and FPU it will not be more that 25Mflops. With M4F @96Mhz -> 30Mflops max.

Here

http://www.roylongbottom.org.uk/Raspberry ...
by Pito
Fri Jan 24, 2020 7:22 pm
Forum: Custom design boards
Topic: Bluepill F4 board, anyone still working on it?
Replies: 83
Views: 114812

Re: Bluepill F4 board, anyone still working on it?

Guys, I wouldn't want to spoil your happiness with the 401/411 boards.. :D

1. there is only 1 FPU on the chip, single precision
2. the MFLOPs we get are numbers which are pretty off the real life, imho.

The Whetstone benchmark has to be reviewed and put into sync with results you may find on some ...
by Pito
Fri Jan 10, 2020 3:37 pm
Forum: General discussion
Topic: ART accelerator
Replies: 25
Views: 28187

Re: ART accelerator

Whetstone results.PNG
Whetstone results.PNG (64.85 KiB) Viewed 6373 times
by Pito
Fri Jan 10, 2020 3:05 pm
Forum: General discussion
Topic: ART accelerator
Replies: 25
Views: 28187

Re: ART accelerator

Double is aprox 2x slower than single (both hw and sw).
I think we need to consolidate the Whetstone results with reality.

You CANNOT get 237 MegaFLOPS with 168MegaHz Cortex M4 FPU.

Based on the above web page with Whetstone results the ~240 MegaFLOPS is a result of Pentium 4 at 2GHz clock.
by Pito
Fri Jan 10, 2020 2:41 pm
Forum: General discussion
Topic: ART accelerator
Replies: 25
Views: 28187

Re: ART accelerator

The ART somehow eliminates the flash access waitstates.
Even theoretically the stuff would run with zero WS, the floating point unit needs 1-14 (1 or 3 or 14) cpu clocks to provide a single precision fp result.
Adding the overhead etc, I would guess the MFLOPs result @168MHz cannot be better than ...
by Pito
Fri Jan 10, 2020 2:28 pm
Forum: General discussion
Topic: ART accelerator
Replies: 25
Views: 28187

Re: ART accelerator

BTW - I doubt the result we see are the Whetstone results :)

We should rather called it "Whetduinos" :P

https://www.st.com/content/ccc/resource/technical/document/application_note/10/6b/dc/ea/5b/6e/47/46/DM00047230.pdf/files/DM00047230.pdf/jcr:content/translations/en.DM00047230.pdf

http://www ...
by Pito
Fri Jan 10, 2020 1:57 pm
Forum: Maple & Maple mini etc
Topic: No COM Port - STM32F103RC8T6
Replies: 12
Views: 69292

Re: No COM Port - STM32F103RC8T6

Maple Mini works fine with Roger (under Win7 here).
The installation of those drivers is sometimes laborious (drivers conflicts, several identical drivers in the drivers store, I had to clean it up).

You also must have the bootloader installed (or use stlink), and use Serial.begin in your sketch ...
by Pito
Fri Jan 10, 2020 1:36 pm
Forum: General discussion
Topic: ART accelerator
Replies: 25
Views: 28187

Re: ART accelerator

At 96MHz you have to see 96/84=1.143 speedup..
Like 135Mflops with Art.

Try with 168MHz, I bet it works.. :) (you should see 236.97 Mflops, xpack 9 compiler, Roger's core, -O3)

Go to advanced search