%PDF-1.5
%
1 0 obj
<>stream
application/pdfIEEE2019 IEEE/ACM International Symposium on Code Generation and Optimization (CGO);2019; ; ; Automatic Generation of Warp-Level Primitives and Atomic Instructions for Fast and Portable Parallel Reduction on GPUsSimon Garcia De GonzaloSitao HuangJuan Gomez-LunaSimon HammondOnur MutluWen-mei Hwu
2019 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)73 Feb. 201984
endstream
endobj
2 0 obj
<>
endobj
3 0 obj
<>stream
h\Vy\Gff[12iӍ`ATC9̌(D eQ0W@Ѡ8
G(r(1b}MjF~U]{W8&6pX[ZN5!&*:21(`0̰...Sw+21&Jm9YlWE?$6Ɖ0+d`-9,as1ü0|03,ð)8L갟~?|RD#E7mKI$^B[W#CI-M2lǰKᇆ5bi/@5\'G rEQhM{z[:E9I:Vx;+q?υԤWdCnFمv^|얂_:PClO 6L8zrc-ykGfOTS#Ң㷰I kSRރqkf*3%,$;+Ԥ?!CpF
Ǭ_-hC48hnI#wKIyhrJCh?uFn2i'G{!L[ ~d6.+eOkP ZzT5Or҃c}"8Цd,SЄ픂U8b_i0"ƹr?p^ޜظX_ff.!rWx"ʡc˧G7<7RDq+͜{|,!M:-xmGl}.|vaW^fG6LuD00٥{rλd>g싹EG*5EfTU.qwe% . ;Mv $$\Q p[ ^ p*6'i{#q[XDž{8!ZF|Q`*Zb0 gl,nzi5F#g~ٴK(ݼ 5x>odҩ'D%Jpj(s QkQnXb,#{30Z6pDNZdFrC"ְW"6,sg%X>7 ݵ٥E>pahݶ^;?pM}ܐ -tx&4
WqRFF6{=By֍Aġ{s2fMfk+9ou->NKγnZ
Z588!&PޠJ V%8 |Q+L:Q4hIENTX oQ5.GڽM??eg-gG U!%$LlnL;'
+o'Ec
)EJ] oo$L-E
h*|tn0:DHu*^3y$|b^mwD!pc $7y|d
fӏ55zINJ"\-q~F4HO=lbnUJpEyX%[hxZ; a>i64N'60ssh2q{'$1w2K"7s}Ǿ-ɧ8։Ջ>puR6JQɠH,<`#bNu^Êk>^yDh[N}W*_c_ʠ;Ђa
wqR)R=ui7at}3fferR,CmwlYi/>(\b^ٴi\Zœ^k䵩21GIzYj6aF1ڻ.p_)hQr^mJPơAWt
boѤ']wtݢTު{SǯihrnޒS4~pM',&o,Z{vVd7!TXGk,::{:dM'{CNZ6(K&