For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports

Discussion of chess software programming and technical issues.

Moderator: Ras

User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports

Post by sje »

For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports
http://hackaday.com/2015/09/05/orange-is-the-new-15-pi/

Included: 100 MHz Ethernet, three USB 2 ports, HDMI video output, audio I/O, composite video output, IR port, camera input, etc.

Not included: wall wart power supply, micro SD card for mass storage, and Ethernet cabling.

Some disappointingly, the ARM A7 CPU architecture is 32 bits, not 64 bits.

From single core systems to multiple core systems to multiple systems of multiple cores; this is the direction of progress for mainstream chess processing power.

Anyone planning to build a world class chess program must now think in terms of distributed system targets.
User avatar
cdani
Posts: 2204
Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra

Re: For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports

Post by cdani »

Do you know if this mini computer will be burned on continuous 100% usage? Probably need some type of refrigeration.
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Re: For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports

Post by sje »

cdani wrote:Do you know if this mini computer will be burned on continuous 100% usage? Probably need some type of refrigeration.
I've had my 900 MHz Raspberry Pi quad core running at full load for days and passive air cooling seems to work without problems.
brtzsnr
Posts: 433
Joined: Fri Jan 16, 2015 4:02 pm

Re: For US$15: 1.6 GHz quad core CPU + 1 GiB RAM+ ports

Post by brtzsnr »

I use a few ODROID-U3 for testing. They are really nice, a bit more powerful than OrangePI+ but also about 3x more expensive. You'll need to factor in sd-card / network cable / power cable so it'll probably be around 40-50$. Still a steal.

You might want to check ODROID-XU4. 64bit and 8 (non-uniform ARM) cores for 70$. A bit better since you'll need fewer of them to build a large cluster.