Invasive Malleable Applications Sebastian Buchwald, Manuel Mohr, - - PowerPoint PPT Presentation

Invasive Malleable Applications

Sebastian Buchwald, Manuel Mohr, Andreas Zwinkau Karlsruhe Institute of Technology TCRC 89 "Invasive Computing"

ATPS 2015

No faster CPUs,

• nly more of them

More cores means slower cores

• M. B. Taylor, A Landscape of the New Dark Silicon Design Regime, Micro, IEEE , vol.33, no.5, 2013

Battery power is more important these days

Challenge: Use more cores more effjciently

Invasive Computing means Tiled Manycore Architectures

Invasive Computing means completely rewritten stack

Custom Hardware iNoC, CiC, i-Core Custom Operating System iRTSS, OctoPOS Custom Programming Language invadeX10 Applications ported/rewritten HPC, Robotics, ...

iRTSS Agent OctoPOS

Tile Tile Tile

OctoPOS OctoPOS App Agent Core Core Core Core Core Core Invasive Framework Application Code X10 Runtime

The invasive framework gives access to more OS functionality

Invasive Computing is about invade-infect-retreat

Constraints Claim i n v a d e r e i n v a d e r e t r e a t i n f e c t i-let

v a l i l e t = ( i d : I n c a r n a t i

• n

I D ) = > { C

• n

s

• l

e . O U T . p r i n t l n ( " H e l l

• W
• r

l d " ) ; } v a l c

• n

s t r a i n t s = n e w P E Q u a n t i t y ( 4 , 1 ) & & n e w S c a l a b i l i t y H i n t ( s p e e d u p C u r v e ) ; v a l c l a i m = C l a i m . i n v a d e ( c

• n

s t r a i n t s ) ; c l a i m . i n f e c t ( i l e t ) ; c l a i m . r e t r e a t ( ) ;

Invasive Computing uses X10

• D. G. Feitelson, L. Rudolph; Towards convergence in job schedulers

for parallel supercomputers, IPPS 1996

• n submission

at runtime User decides

rigid evolving

System decides

moldable malleable

Invasive Applications are Malleable

Example: Heat Dissipation with Multigrid Approach

• M. Schreiber, A. Zwinkau, et al. Invasive computing in HPC with X10, X10 Workshop 2013

Multigrid lets resource idle

idle!

Invasive Apps exchange cores

Asynchronously Malleable: „System decides any time at runtime“

Master Slave Master Slave Slave

Works nicely for Master-Slave Applications

Asynchronously Malleable is more effjcient

Not idle!

Integration is async-malleable

f(x) = sin(x²)

Sorting is async-malleable

Patrick Flick, Peter Sanders, Jochen Speck Malleable Sorting IPDPS 2013

v a l i l e t = ( i d : I n c a r n a t i

• n

I D ) = > { f

• r

( j

• b

i n q u e u e ) { i f ( q u e u e . c h e c k T e r m i n a t i

• n

( i d ) ) b r e a k ; j

• b

. d

• (

) ; } } v a l r e s i z e H a n d l e r = ( a d d : L i s t [ P E ] , r e m

• v

e : L i s t [ P E ] ) = > { f

• r

( p e i n a d d ) q u e u e . a d d W

• r

k e r ( p e , i l e t ) ; q u e u e . a d a p t ( ) ; f

• r

( p e i n r e m

• v

e ) q u e u e . s i g n a l T e r m i n a t i

• n

( p e ) ; } v a l c

• n

s t r a i n t s = n e w P E Q u a n t i t y ( 4 , 1 ) & & n e w A s y n c M a l l e a b l e ( r e s i z e H a n d l e r ) & & n e w S c a l a b i l i t y H i n t ( s p e e d u p C u r v e ) ; v a l c l a i m = C l a i m . i n v a d e ( c

• n

s t r a i n t s ) ; q u e u e . a d a p t T

• (

c l a i m ) ; c l a i m . i n f e c t ( i l e t ) ; c l a i m . r e t r e a t ( ) ;

v a l i l e t = ( i d : I n c a r n a t i

• n

I D ) = > { f

• r

( j

• b

i n q u e u e ) { i f ( q u e u e . c h e c k T e r m i n a t i

• n

( i d ) ) b r e a k ; j

• b

. d

• (

) ; } } v a l r e s i z e H a n d l e r = ( a d d : L i s t [ P E ] , r e m

• v

e : L i s t [ P E ] ) = > { f

• r

( p e i n a d d ) q u e u e . a d d W

• r

k e r ( p e , i l e t ) ; q u e u e . a d a p t ( ) ; f

• r

( p e i n r e m

• v

e ) q u e u e . s i g n a l T e r m i n a t i

• n

( p e ) ; } v a l c

• n

s t r a i n t s = n e w P E Q u a n t i t y ( 4 , 1 ) & & n e w A s y n c M a l l e a b l e ( r e s i z e H a n d l e r ) & & n e w S c a l a b i l i t y H i n t ( s p e e d u p C u r v e ) ; v a l c l a i m = C l a i m . i n v a d e ( c

• n

s t r a i n t s ) ; q u e u e . a d a p t T

• (

c l a i m ) ; c l a i m . i n f e c t ( i l e t ) ; c l a i m . r e t r e a t ( ) ;

v a l i l e t = ( i d : I n c a r n a t i

• n

I D ) = > { f

• r

( j

• b

i n q u e u e ) { i f ( q u e u e . c h e c k T e r m i n a t i

• n

( i d ) ) b r e a k ; j

• b

. d

• (

) ; } } v a l r e s i z e H a n d l e r = ( a d d : L i s t [ P E ] , r e m

• v

e : L i s t [ P E ] ) = > { f

• r

( p e i n a d d ) q u e u e . a d d W

• r

k e r ( p e , i l e t ) ; q u e u e . a d a p t ( ) ; f

• r

( p e i n r e m

• v

e ) q u e u e . s i g n a l T e r m i n a t i

• n

( p e ) ; } v a l c

• n

s t r a i n t s = n e w P E Q u a n t i t y ( 4 , 1 ) & & n e w M a l l e a b l e ( r e s i z e H a n d l e r ) & & n e w S c a l a b i l i t y H i n t ( s p e e d u p C u r v e ) ; v a l c l a i m = C l a i m . i n v a d e ( c

• n

s t r a i n t s ) ; q u e u e . a d a p t T

• (

c l a i m ) ; c l a i m . i n f e c t ( i l e t ) ; c l a i m . r e t r e a t ( ) ;

Invasive Malleable Applications

Appendix

Case Study: Multigrid

• M. Schreiber, A. Zwinkau, et al. Invasive computing in HPC with X10

X10 Workshop 2013

v a l i l e t = ( i d : I n c a r n a t i

• n

I D ) = > { f

• r

( j

• b

i n q u e u e ) { i f ( q u e u e . c h e c k T e r m i n a t i

• n

( i d ) ) b r e a k ; j

• b

. d

• (

) ; } } v a l r e s i z e H a n d l e r = ( a d d : L i s t [ P E ] , r e m

• v

e : L i s t [ P E ] ) = > { f

• r

( p e i n a d d ) q u e u e . a d d W

• r

k e r ( p e , i l e t ) ; q u e u e . a d a p t ( ) ; f

• r

( p e i n r e m

• v

e ) q u e u e . s i g n a l T e r m i n a t i

• n

( p e ) ; } v a l c

• n

s t r a i n t s = n e w P E Q u a n t i t y ( 4 , 1 ) & & n e w M a l l e a b l e ( r e s i z e H a n d l e r ) & & n e w S c a l a b i l i t y H i n t ( s p e e d u p C u r v e ) ; v a l c l a i m = C l a i m . i n v a d e ( c

• n

s t r a i n t s ) ; q u e u e . a d a p t T

• (

c l a i m ) ; c l a i m . i n f e c t ( i l e t ) ; c l a i m . r e t r e a t ( ) ;