Hi,
I was curious how the thought process of my chess program looks like.
I had some neopixels lying around so I interfaced it using an arduino to my chess program. In this youtube clip https://youtu.be/7opPMQPKO7A you see the result.
neopixels & "thought" process
Moderators: hgm, Dann Corbit, Harvey Williamson
-
PK
- Posts: 893
- Joined: Mon Jan 15, 2007 11:23 am
- Location: Warsza
-
sje
- Posts: 4675
- Joined: Mon Mar 13, 2006 7:43 pm
Re: neopixels & "thought" process
I did something much like this a few years ago before neopixels became available. I used an 8x8 grid of RGB LEDs all in a single plastic block controlled by its own Atmel CPU. The block was connected via an SPI interface to an Arduino Mega (128 KiB PROM, 8 KiB RAM) which was running my program Myopic.
I seem to recall that I got most of the parts from SparkFun.
The display was probably slower than any made from neopixels because to change any single LED, all 64 LEDs needed refreshed.
Each chessman kind had its own color. All of the White pieces were mostly blue and all of the Black pieces were mostly red.
It was interesting to watch perft() calculations as well as searches. Bun refreshing the entire grid for each move meant that the system spent most of its time waiting for data to be shipped from the Arduino to the display block.
I seem to recall that I got most of the parts from SparkFun.
The display was probably slower than any made from neopixels because to change any single LED, all 64 LEDs needed refreshed.
Each chessman kind had its own color. All of the White pieces were mostly blue and all of the Black pieces were mostly red.
It was interesting to watch perft() calculations as well as searches. Bun refreshing the entire grid for each move meant that the system spent most of its time waiting for data to be shipped from the Arduino to the display block.
-
stegemma
- Posts: 859
- Joined: Mon Aug 10, 2009 10:05 pm
- Location: Italy
- Full name: Stefano Gemma
Re: neopixels & "thought" process
But because our eyes have a persistence of about 1/10 second maybe you can get the same result just collecting the average of the "pixels" and displaying with such interval or a little less (1/20 second), as cinema does. The blinking leds at the speed of perft would becomes just fading lights, I think.sje wrote:I did something much like this a few years ago before neopixels became available. I used an 8x8 grid of RGB LEDs all in a single plastic block controlled by its own Atmel CPU. The block was connected via an SPI interface to an Arduino Mega (128 KiB PROM, 8 KiB RAM) which was running my program Myopic.
I seem to recall that I got most of the parts from SparkFun.
The display was probably slower than any made from neopixels because to change any single LED, all 64 LEDs needed refreshed.
Each chessman kind had its own color. All of the White pieces were mostly blue and all of the Black pieces were mostly red.
It was interesting to watch perft() calculations as well as searches. Bun refreshing the entire grid for each move meant that the system spent most of its time waiting for data to be shipped from the Arduino to the display block.
Author of Drago, Raffaela, Freccia, Satana, Sabrina.
http://www.linformatica.com
http://www.linformatica.com