ConFoo 2016: Parallella, Supercomputing, & Ruby

Presented in Montreal, Canada at ConFoo 2016. Includes new material on WebSockets and field programmable gate arrays (FPGAs).

Curious about the Internet of Things? Check out http://WindyCityThings.com.

Transcript

1. Confoo.CA Montreal, QB, Canada February 24-26, 2016 RayHightower.com Supercomputing & Ruby
2. •Why Parallelism? •Concurrency vs Parallelism •Parallella vs MacBook Pro •Cool & Useful •Q & A Our Journey…
3. RISC ARM 18 cores: 2 ARM + 16 RISC
4. RayHightower.com
5. Why?
6. Faster is better.
7. Faster is better.
8. Moore’s Law: 2x every 18 months
9. Moore’s Law: 2x every 18 months 1993 http://www.washingtonpost.com/blogs/innovations/wp/ 2015/04/14/10-images-that-explain-the-incredible-power-of- moores-law/ 2013
10. Parallelism
11. Concurrency is not Parallelism. https://www.youtube.com/watch?v=cN_DpYBzKso&list=PLOnWKC1gI_OPU8SDIBnCLHsgzNLSbnPJQ&index=3 -Rob Pike, Go
12. Concurrency At least two threads are making progress. Parallelism At least two threads are executing simultaneously. Oracle Multithreaded Programming Guide http://docs.oracle.com/cd/E19455-01/806-5257/6je9h032b/index.html vs.
14. If one ox could not do the job they did not try to grow a bigger ox, but used two oxen. When we need greater computer power, the answer is not to get a bigger computer, but to build systems of computers and operate them in parallel. -Grace Hopper
15. RJ-45 Power μUSB μHDMI μSD
19. Watts & Dollars
21. 17.8 million watts $17.8 million per year
22. 5 watts for Parallella?
23. http://rayhightower.com/blog/2014/09/09/solar-powered-parallella/
24. http://rayhightower.com/blog/2014/09/09/solar-powered-parallella/ 5 volts 1 amp 5 watts Solar!
25. Reduced Instruction Set Computer RISC
26. Advanced (or Acorn) RISC Machine ARM
27. Field Programmable Gate Array FPGA
28. RISC ARM + FPGA 18 cores: 2 ARM + 16 RISC
30. Find all primes up to 16,000,000. Serial on Parallella.
31. 1 #include <stdio.h> 2 #include <stdlib.h> 3 #include <math.h> 4 #include <time.h> 5 6 #define DEFAULT_MAX_TESTS 16000000 7 8 inline int isprime(unsigned long number) 9 { 10 unsigned long i; 11 unsigned long s = sqrt(number); 12 for(i=3;i<=s;i+=2) 13 { 14 if(number % i == 0) 15 return 0; 16 } 17 return 1; 18 } /* Copyright (c) Adapteva, contributed by M. Thompson with modifications by T. Malthouse. */
32. 8 inline int isprime(unsigned long number) 9 { 10 unsigned long i; 11 unsigned long s = sqrt(number); 12 for(i=3;i<=s;i+=2) 13 { 14 if(number % i == 0) 15 return 0; 16 } 17 return 1; 18 } /* Copyright (c) Adapteva, contributed by M. Thompson with modifications by T. Malthouse. */
34. 237.1 sec
35. Find all primes up to 16,000,000. Serial on Mac OS X.
36. Same serial code, written in C. Build it on OS X.
38. 14.4 sec
39. Find all primes up to 16,000,000. Parallel on Parallella.
40. 27 #include <e-hal.h> 28 29 // Default max number of prime tests per core 30 // Used if a limit it not provided in argv[1] 31 #define DEFAULT_MAX_TESTS 500000 32 33 int main(int argc, char *argv[]) 34 { 35 unsigned row, col, coreid, i, j; 36 e_platform_t platform; 37 e_epiphany_t dev; 38 /* Copyright (c) Adapteva, contributed by M. Thompson with modifications by T. Malthouse.*/
42. 18.6 sec
43. Summary: Finding Primes 0" 50" 100" 150" 200" 250" Parallel Parallella Serial Mac Serial Parallella 18.6 sec 14.4 sec 237.1 sec ($2,000.00 Apple MacBook Pro) ($150.00 Parallella)
44. Embarrassingly parallel problem.
45. Mandelbrot Set
47. Why?
48. https://www.e-education.psu.edu/worldofweather/node/2029 Grid spacing influences accuracy. Weather Prediction
49. Where can Ruby (or Python, Node.js, Erlang, etc.) fit?
50. Ruby
 +
 Sinatra
 +
 WebSockets
51. Original example by The Hybrid Group and Engine Yard. Modified by WisdomGroup.
52. Field Programmable Gate Array FPGA
54. AB 00 01 10 11 Q 1 1 1 0
55. Parallella vs GPU
56. 384 cores MacBook Pro Video
57. Specialized vs. General
58. Something new for you… Pine A64 4 cores 64-bit $15.00
60. Merci! RayHightower.com