Supercomputing for Everyone

Presented at RubyConf Colombia on October 16, 2015. Why does parallel computing matter? How can developers use parallel computing to deliver better results for clients? Let’s explore these questions together.

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

Transcript

1. RubyConf Colombia
 Medellín, Colombia October 16-17, 2015 Supercomputing for Everyone
2. RISC ARM 18 cores: 2 ARM + 16 RISC
3. RayHightower.com
4. IBM Blue Gene https://commons.wikimedia.org Cray X-MP
5. Concurrency is not Parallelism. https://www.youtube.com/watch?v=cN_DpYBzKso&list=PLOnWKC1gI_OPU8SDIBnCLHsgzNLSbnPJQ&index=3 -Rob Pike, Go
6. 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
7. RJ-45 Power μUSB μHDMI μSD
8. Moore’s Law: 2x every 18 months
9. Moore’s Law: 2x every 18 months Moore’s Law 1993 - 2013 http://www.washingtonpost.com/blogs/innovations/wp/2015/04/14/10-images-that-explain-the- incredible-power-of-moores-law/ https://www.parallella.org/board/
13. Watts & Dollars
15. 17.8 million watts $17.8 million per year
16. 5 watts for Parallella?
18. http://rayhightower.com/blog/2014/09/09/solar-powered-parallella/
19. http://rayhightower.com/blog/2014/09/09/solar-powered-parallella/
20. http://rayhightower.com/blog/2014/09/09/solar-powered-parallella/ 5 volts 1 amp 5 watts Solar!
21. Reduced Instruction Set Computer RISC
22. ARM A9
23. RISC ARM 18 cores: 2 ARM + 16 RISC
25. Hello Epiphany
26. host.c epiphany.c
28. Find all primes up to 16,000,000. Serial on Parallella.
29. 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. */
30. 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. Find all primes up to 16,000,000. Serial on Mac OS X.
33. Same serial code, written in C. Build it on OS X.
35. Find all primes up to 16,000,000. Parallel on Parallella.
36. 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.*/
38. Serial/Parallella: Serial/Mac: Parallel/Parallella: 237.1 sec 14.4 sec 18.6 sec Summary: Finding Primes 18 cores (using one). Up to 1GHz. $150.00 4 cores (using one). 2.7GHz. $2,000.00 18 cores. Up to 1GHz. $150.00
39. Embarrassingly Parallel Problems
40. Mandelbrot Set
42. Why?
43. Finite Element Analysis http://www.ce.ncsu.edu/news/article/21550/making-bridges-more-robust-to-earthquakes/
44. break? wear out? work? Finite Element Analysis Will it {
45. Free Body Diagram F F applied friction F gravity F normal
46. Finite Element Analysis http://www.ce.ncsu.edu/news/article/21550/making-bridges-more-robust-to-earthquakes/
47. https://www.e-education.psu.edu/worldofweather/node/2029 Grid spacing influences accuracy. Weather Prediction
50. Parallella vs GPU
51. 384 cores MacBook Pro Video
52. Specialized vs. General
54. Thanks! RayHightower.com