%PDF-1.3 1 0 obj << /Type /Catalog /Outlines 2 0 R /Pages 3 0 R >> endobj 2 0 obj << /Type /Outlines /Count 0 >> endobj 3 0 obj << /Type /Pages /Kids [6 0 R 17 0 R ] /Count 2 /Resources << /ProcSet 4 0 R /Font << /F1 8 0 R /F2 9 0 R /F3 10 0 R /F4 19 0 R >> /XObject << /I1 11 0 R /I2 12 0 R >> >> /MediaBox [0.000 0.000 612.000 792.000] >> endobj 4 0 obj [/PDF /Text /ImageC ] endobj 5 0 obj << /Creator (DOMPDF) /CreationDate (D:20150923173533-04'00') /ModDate (D:20150923173533-04'00') /Title (The Green500 List - June 2014) >> endobj 6 0 obj << /Type /Page /Parent 3 0 R /Annots [ 13 0 R 15 0 R ] /Contents 7 0 R >> endobj 7 0 obj << /Length 5540 >> stream 1.000 1.000 1.000 rg 46.016 46.016 519.969 699.969 re f q 152.250 0 0 105.000 46.016 640.984 cm /I2 Do Q 0.000 0.000 0.000 rg BT 46.016 629.263 Td /F1 12.0 Tf [(Published on )] TJ ET BT 118.724 629.263 Td /F2 12.0 Tf [(The Green500)] TJ ET BT 196.100 629.263 Td /F1 12.0 Tf [( \()] TJ ET BT 203.432 629.263 Td /F1 12.0 Tf [(http://www.green500.org)] TJ ET 0.000 0.000 0.000 RG 0.6 w 0 J [ ] 0 d 203.432 626.863 m 334.160 626.863 l S BT 334.160 629.263 Td /F1 12.0 Tf [(\))] TJ ET BT 46.016 605.541 Td /F1 9.0 Tf [(Home)] TJ ET 0.45 w 0 J [ ] 0 d 46.016 603.741 m 70.019 603.741 l S BT 70.019 605.541 Td /F1 9.0 Tf [( > Printer-friendly PDF)] TJ ET 0.620 0.620 0.620 rg 46.016 596.593 519.969 0.750 re f 0.000 0.000 0.000 rg BT 46.016 557.070 Td /F3 24.0 Tf [(The Green500 List - June 2014)] TJ ET BT 46.016 523.408 Td /F1 12.0 Tf [(30 Jun 2014)] TJ ET BT 46.016 496.756 Td /F1 12.0 Tf [(The June 2014 release of the Green500 list shows an increasing trend of heterogeneous )] TJ ET BT 46.016 482.104 Td /F1 12.0 Tf [(supercomputing systems at the top of the list. Now the top 17 spots of the list are occupied by )] TJ ET BT 46.016 467.452 Td /F1 12.0 Tf [(heterogeneous computing systems, securing six more spots at the top over the previous list, a )] TJ ET BT 46.016 452.800 Td /F1 12.0 Tf [(55% increase. A heterogeneous supercomputing system is one that uses computational building )] TJ ET BT 46.016 438.148 Td /F1 12.0 Tf [(blocks that consist of two or more types of ?computing brains.? These types of computing brains )] TJ ET BT 46.016 423.496 Td /F1 12.0 Tf [(include traditional processors \(CPUs\), graphics processing units \(GPUs\), and co-processors. Of )] TJ ET BT 46.016 408.844 Td /F1 12.0 Tf [(the top 17 spots,)] TJ ET 69.816 385.613 m 69.816 386.163 69.589 386.709 69.201 387.098 c 68.812 387.487 68.266 387.713 67.716 387.713 c 67.166 387.713 66.620 387.487 66.231 387.098 c 65.842 386.709 65.616 386.163 65.616 385.613 c 65.616 385.063 65.842 384.517 66.231 384.128 c 66.620 383.740 67.166 383.513 67.716 383.513 c 68.266 383.513 68.812 383.740 69.201 384.128 c 69.589 384.517 69.816 385.063 69.816 385.613 c f BT 76.016 382.192 Td /F1 12.0 Tf [(The first 15 systems are accelerated with NVIDIA Kepler K20 GPUs, coupled with Intel )] TJ ET BT 76.016 367.540 Td /F1 12.0 Tf [(Xeon CPUs, including TSUBAME-KFC at number 1.)] TJ ET 69.816 356.309 m 69.816 356.859 69.589 357.405 69.201 357.794 c 68.812 358.183 68.266 358.409 67.716 358.409 c 67.166 358.409 66.620 358.183 66.231 357.794 c 65.842 357.405 65.616 356.859 65.616 356.309 c 65.616 355.759 65.842 355.213 66.231 354.824 c 66.620 354.436 67.166 354.209 67.716 354.209 c 68.266 354.209 68.812 354.436 69.201 354.824 c 69.589 355.213 69.816 355.759 69.816 356.309 c f BT 76.016 352.888 Td /F1 12.0 Tf [(The number 16 system, Shadow, is accelerated with Intel Xeon Phi co-processors and )] TJ ET BT 76.016 338.236 Td /F1 12.0 Tf [(coupled with Intel Xeon CPUs.)] TJ ET 69.816 327.005 m 69.816 327.555 69.589 328.101 69.201 328.490 c 68.812 328.879 68.266 329.105 67.716 329.105 c 67.166 329.105 66.620 328.879 66.231 328.490 c 65.842 328.101 65.616 327.555 65.616 327.005 c 65.616 326.455 65.842 325.909 66.231 325.520 c 66.620 325.132 67.166 324.905 67.716 324.905 c 68.266 324.905 68.812 325.132 69.201 325.520 c 69.589 325.909 69.816 326.455 69.816 327.005 c f BT 76.016 323.584 Td /F1 12.0 Tf [(The number 17 system, SANAM, is accelerated with AMD FirePro S10000 GPUs, coupled )] TJ ET BT 76.016 308.932 Td /F1 12.0 Tf [(with Intel Xeon CPUs)] TJ ET BT 46.016 282.280 Td /F1 12.0 Tf [(While the above results might lead one to believe that heterogeneous systems permeate )] TJ ET BT 46.016 267.628 Td /F1 12.0 Tf [(throughout the Green500 list, they are only prevalent at the top end of the list. Of the 500 ranked )] TJ ET BT 46.016 252.976 Td /F1 12.0 Tf [(systems, only 64 of them are heterogeneous while the other 436 are homogeneous and contain )] TJ ET BT 46.016 238.324 Td /F1 12.0 Tf [(only one type of brain, that is, CPU brains. The total number of heterogeneous systems on the )] TJ ET BT 46.016 223.672 Td /F1 12.0 Tf [(list remains relatively unchanged from the November 2013 list, as does their overall share of )] TJ ET BT 46.016 209.020 Td /F1 12.0 Tf [(performance. The average energy efficiency of these systems, when measured in millions of )] TJ ET BT 46.016 194.368 Td /F1 12.0 Tf [(floating-point operations per second per watt, or \(MFLOPS/watt\) is 1,938 MFLOPS/watt, whereas )] TJ ET BT 46.016 179.716 Td /F1 12.0 Tf [(it is only 743.32 MFLOPS/watt for the measured homogeneous systems. As a result, )] TJ ET BT 46.016 165.064 Td /F1 12.0 Tf [(heterogeneous systems dominate the top end of the list. So, perhaps the old adage of ?two )] TJ ET BT 46.016 150.412 Td /F1 12.0 Tf [(heads being better than one,? or in this case, ?two types of computing brains being better than )] TJ ET BT 46.016 135.760 Td /F1 12.0 Tf [(one? appears to hold.)] TJ ET BT 46.016 109.108 Td /F1 12.0 Tf [(The energy efficiency at the top end of the Green500 list remains largely unchanged with )] TJ ET BT 46.016 94.456 Td /F1 12.0 Tf [(TSUBAME-KFC still at the top and remaining the only system above the 4,000 MFLOPS/watt )] TJ ET BT 46.016 79.804 Td /F1 12.0 Tf [(mark. The 10 greenest machines from the previous Green500 list all remain in the top 15 of this )] TJ ET BT 46.016 65.152 Td /F1 12.0 Tf [(Green500 list and were joined by similarly heterogeneous architectures. In addition, the ranking )] TJ ET endstream endobj 8 0 obj << /Type /Font /Subtype /Type1 /Name /F1 /BaseFont /Helvetica /Encoding /WinAnsiEncoding >> endobj 9 0 obj << /Type /Font /Subtype /Type1 /Name /F2 /BaseFont /Helvetica-Oblique /Encoding /WinAnsiEncoding >> endobj 10 0 obj << /Type /Font /Subtype /Type1 /Name /F3 /BaseFont /Helvetica-Bold /Encoding /WinAnsiEncoding >> endobj 11 0 obj << /Type /XObject /Subtype /Image /Width 203 /Height 140 /Filter /FlateDecode /DecodeParms << /Predictor 15 /Colors 1 /Columns 203 /BitsPerComponent 8>> /ColorSpace /DeviceGray /BitsPerComponent 8 /Length 5405>> stream x\]d%پJ)RJ)[)[ٶ-b-"""hѢEDhm-ZZk-FkE<18p]p\>\cy.w}Z_V}%?'{h7AZN;4cy)529Ƥ IVKIp\pRI24 PIT`6RT؇9-)V赨WݍϱUU4_tmTXb.E&AHZh", %ԢYt%?.M24DamEWY1ۛ_X̅hٝ^Ь3Y.,qK&\Pt)y$H1Ym,C^z^l\а:#಻`c}ma=sy>M^s3s.cfX }1AdMeK\<#/<[R]c{i,gۣ܎.z*dɜxWY1yಏXV|CdcP&A l0CIEWVGGC[̥aǣ=r/}3{jifpxxxt˖\6^tI.K3\ZDrX%\X8m>qr-ftɑDž5O`*rRIܢ%\޶Y沃b.>=9&ɗ.$}|^Sz'rj+6cH-I\VxBKxy.(A :!ZETbLuH%.k񼘋,[vu]ht-lwDTXotdK"mHLpΖn .>?ܻR&$mDBHź0|rf1,cF$sQsU4Ĉ/7znMD[It.T'm :)|'[B菫gqOc/TfnAa;2š6Qe+eN$q"xW9.:M,bwF$UvpA3 9* 4f9 UxzCEbe8˖>qa]Mp gDrBETZe~)YІU\ȺŘr8L"CN,2b\RE6)42֡B.x4 Y/#K'Bl5Hqj.kr1~߱M3ZpI1l?s1?\.?]^|𞧿q.9GVr1%6ؤi{b ۅRxTx\r˸z"3tnu\\ ʂ"|dؓ9g[h@)遞Z*Vri y^lIPH rihoJArkNEA .B|&˒n$?<2\[y.DMJL\\xlOKɰq+c@tVec&(1miGt#py!Qguv&Sb- 曉 E/HaAĚ%lkͰ:~ZV+꿼:@Jk|ŗǟ>^\^~D6UЧO9\`$ү>ѓ ^Ȕ*y+pu0r!cڎ7OOhWna~ 56}b6L!̣I]Lؤ;`wQ>J/i]b6<5mX0!yПKZJM-.JUɨ%njejP@A2u>UB+-I7WxKĴI-d4XVژ4zv4Eӧ6uCdRXH+ۭ$xDHEK G<+;Ujr`oU^\3]swзd/Q-؛Ak>vnn X\NO?{v+62@Ic!+LP^XwxUb.U%V:{.xXX+W?6٢K[%oI6,PRe3I u*KB!S%yS9ldcl5}9.8!ǖJ/>=QI *DpQ&(<䠑i NU4|`%ERmɔJ 4o'skd^nz֢<O'/1Y#}YX AJ>ʸ*BaWS%y0^Bz]{wX{mssS)%餻Bb>B\LNPWү;- G(peuAAI{-$mepE|"\VWVfZ[E%4 TjnbU9j\]ʠ%QyɑZ=z8<-Y?VŋԸݎn2br*3?LReu5L RS{X$F5D?r-)yH~,Uҭ=B =PEN1ܣBBBJU$%TMKhWr$,."&ee!<5ş5!sqK\ Ry\L:"VUL@]Z[ &"1,a.5|^yv.}E,|ko]m0/k1a9E͑&{ B] Fͳ .@W-0A-I,2dEί!(!nS4âg4VĒ1sUs u"yA\aecc%ݩ9dл WQ9J+mBDuWu71V!㪀hZD:B٬5_Q2:jݴae@nQ?>uwV);I¢#;TEF!$,1ojO (q^ iL!;$,K8T}t +vJrWe߂m&(G'e|n 0I?.+oP2إx LX !=4(\%:0AI*/ֹ*F ;zZUղ3~&@ aIі hx1-G|$O##גɒto)7?.(ɀO_yQ0ozc aHhѽlzg,8ɰ~ieh?[ΈK 3CZDB/P[^w2FVt7ٷC&,WFXE^Pg_$3h:&Gwƚ#g۲kDO ^tG.3wW&Sq~e-ohB6 ?. 2"/fU"n׬!)k1I-+Vn;ʳ%eR<&Fe%/]د3/8(c'oMh_15T5é5?7Vc{0Of6ɅS%Η4l'o⭚uP(d!3X˽<ĭVH&VA0a;`+ЪqVvjkKSs'O?mW,Hr_!j?!7G'&zsEBDqv{O>zKD+FvJoIZkbO^_֠)Y3ZhALHh\}R+E$$bC5̪M%R6 p:|T1u 샱Zƃ~L֑`na1FwH]bjKG_4S+Ix$/C+_Qy0˹ht ̰w/yQ6tu@K.zG RIPϐ5oL|7 np7?̤B endstream endobj 12 0 obj << /Type /XObject /Subtype /Image /Width 203 /Height 140 /SMask 11 0 R /Filter /FlateDecode /DecodeParms << /Predictor 15 /Colors 3 /Columns 203 /BitsPerComponent 8>> /ColorSpace /DeviceRGB /BitsPerComponent 8 /Length 17234>> stream x}w|յj{,Yro0%6!GI &䅖 š9B`dY-YzYj1z5;ڕVN޹s̙s=B.2 ?z9.Se,.._ii+\aS0MOOO5\ƿ"nx[ E% dlSd1z7cg QSI<?6_~6#qW[cj  CF|^Cvz*^-lݶg& H$hҥlI}}B%");3N9_5$?P1g=`ri=] .b,fA#1 mmv[ss35G@Crqz|#@eBT.42pE6HFFF{zNnOMK/=ҷo8$ɛo}GGXA5;m iR}S(cէO{_aMwMc"N8}L9oW/j。6' ?ѻ<Γ F㹆Ui>k#5-vHW[{gGg7I9M.? .WNW$jq~bb{ `q^|d݄aoewok9|qWzl]bP ¢S]$P_O$ EQZV!vYi|31V)j"UE`Őv^|vb~8-{!d A0iR Rs1^q=޻λP)G p\"phxd||"';K&.]Iz}4,z)w>JĊ eX J@vi=;^0ӈ'5FjF`%/{~3Л <@,AT$#T8gXaa0{='DҫZd-$ܿgϞ^xSXF((E]. fw}AAC|@愈Ɩӣ:@@ii~i:~{eQ6Nkit%}E2#(v3ij;6EO/wGE=] Ͽ;3gQ!>!!)@թ{wj0=S=o YHRcv 2 lV+a8 v'!Gm=F_kl}WVx駯* CK:gSzĆyşANfuk֬Zꫯ^t@r=]#""h4Y4A'&'o ABS8ASy|Ԓc~{ŦM .pwFX"1-. `0@(3 ÒnG萳wPe;;^<#չkU^jCjl*.}q: `/۴y@HMk3 J`xk_g# Hģ] o&رû|1$]/ C#=.@ a~9r/d잇R*rdWT׵x X7 w3 oh-qҳLId~gMOnzazۉ~5 pMQ/??_E$N8'ǰՆǏoظ1 QY ߼fӓ]bqVOv{;۱k_ʻ|6GC8l]e}eX "*Fh4JJRr\.J3O? /04Ll5u1کW#5gxsXoO h[-u^3QZ%e>쳅x`Y 7wuj@}yc D*bú۷-}{N<88*W2+;_"ðGu;.?2¡7bc 9 &%?`C<55A<B@QTfVwUJ%8&k{AѴ'0ؼsrsZmb٩0С<%}XE0 뮿G͗_;_J5dJg/-Yt1aW}㴑5 VZjd[&.^u6-OzHGB/P˕][___ZZ L۫~CiHVuuvflv;/X'/Ҙ~*&Ë \QV)0Xs.edfoh|{# WA0,N@1H1 ڜӓ"^m;}Tu{.]z+/ `d u@%9֑@-HUb3E$К/s87Hb߈IFu1~8eX6urb{7ٲmz2LR@ʗ%ˤ Ngl98_ kߚ/ {t' 6h5O@Y w Dk Nzm<:9*Y'Nujk96Ĭv}H!aǗ8wWUqd1|m 4x  vSLdboabc7oڰDFj^{y\ђ%Q2~TZ* |#)UDca 0빪ͫ׬A"[aBd{ٙ 0l %۔Ǯa>vyMW4}js^~>HE;}p@ȼN lGyGNWUv79&\mVj  N- Ox))[ܘ * F hh4b$^~3B>#fsUM&SPqmIC`cIF `"yL9,L(.C(݌`-,L->^TF4M3 !qV>骪(Eib.EQV{힚mm 0 M'20ugz򬭩h4ll !;Pttvz뭞Qaa! ϧ㦛oocsszhї_~y-(yİ `)S;ctDz?a|z/ff~BQ/F]ճA}%M=,9\ja`Nve2X={׼{siHw2Ҙ7' .6 Lij4no@2e=: 75o wO:" A>NF˥Y?@ voSev㐚o Á:jo4&od@ ge O B\P>1(]ڻ;-p=lDv)(̃9I3mN}*dPpCx0>P:+{-r-#4$ߧƏw #- %`Se\9L(lV#@On7`DSi*,h(3A `PGm9=F$GaVub4}` ]B,z9SX )|A3Fs6!(COE0f9"tBRF0< an dK!mшP&pw填P#isLc v|rO'~ ԋ!Pdm Vp $I ᑑygժĄH @@:FgW709YMɉImOoA~bxQ" 2ueUF#+0yA=]^_@ynR` AtvVfcsKtTd\l,00t/a7mžֶֶ9qtuHyYi[kk)X VRc1M "k~!&tݹNW$'% +a"ԪƦfAdge8C(׏a&(rcwNzĩs D^,h1}1B @ t\-d!_ٕ[sã&AQTo_LL;;:^{USJT)nDQ4&&Z(-+IMIH$FRA]s`p[0tiqB. gPTmAkwNXUu6(*P 0 $I2ZQw4n\./!JbT*HĬOAPp8Nj F--).Na223|˗4VPX&𬒬w qZGQ5U 5 Abws}3B"h"JR"bXc8at86b sh<}fiիV#/PN#YbB|llJR).kV VR if4e0LOo74ҒSNݻy Ѵ_17MX +v{MJLHINezpLz6CQXPIlX޾&krOd,yU>3蟼HKMh"FFGޫ ϲ߾҄ܜl٭ xgg7"/x<03Ȱ [ qm;-FxbnĉSGD;;;[[[?3pPdɏ;=--++l6wtX@Y^,1!(as@QTeUu\TV*++_{cǎn>7s'')[Vr/O)y.##pQɋ+vNLgjL*-))NKM{yO{[{YjUVVFm9Фހ= 74,-) i T#/X(+3{L矯?jۖΦV,F g/D"FQ_E֝X,nݲ Ce˖zmA7NUV~w΢%Kjۖ< 4M֝:9pᄉRLtqђ%<˃YO=pņujֶɆI <f6sߨT"a O~C n\kƍbYr3grV,/Gdvl}bbbEEw?"]s|bjZj kG׬]&KBV;׀Ç˃h vYbc׬Y/>޾>9d%_Ε^3\-;f^TX0>>I4i4Nvof( A_|Xx[nv8##ziH7 @T"Rc~7=s>>kxQQѹ$Ӱoy ^AbeR%mr"HlvOPU2fLbn@^,T[0GEؘԔĄvf P))bdeB[Z##5N\~=g}H$bXҲ-A~iH$Y#P8g(``F.nYhv:2/Ja1ѹQQ#ÿV{/7//#-.r57^q}EE9#/HD"jzJ$< M3B"##%/BE 9޲0V5&&:8$~ ֚ۮyvUZZ:88vKxOL h߄ L.-KS "P[X KLT3LLan!<`iq/y²b7kK r}4?1 E t $UBh֭[<_099CDF8hbб@ͅRD,NKMޙD >د|衐Z䓍MM%KD(1ٲ9h]?Ώ{0`TfCNAZtQ醓 2: 7aXaA7|gͪ*2))qIQoϣnn* #7n"Z[\JX,VuAL>)!܏@`C$P`,2A71gx".a<썃ogd-?;I>a'M p8UXXLVS' Q$IzvvF "[\^8WJ#bzX(J$Kp[0i;\BBCӌbr!bfdvizڷ\.v9nJy9NލB);P ~mag8eLœbN#dLo,jwD^iRվuh6/p nu8S)>2SSSwC`4N%$$,q](htul`,3SAe2iuuu0‚mA DBHßfQYYPԧ^Hv19Dj"ӛ B_oIY`u&9L-;N$#Rt ) `0:|j"Eo?zBUu8<̜fdhx0G:Ym6N#|= x"yS_ B&SnLeM:kgEKNAF/ldTktC8WI_!>)|XN;5>6k4;~J.ލs_@QQ'|R$]*‚VAn ֖tG| ]\j[`odƯ# ^ZrЭ˽{~;\qvū1JKI!#vJ%m ݻk6'5_ N@JXj) Fq,==rzGMQ<l8K-cPhimegVn/Bi0y7QO8MVZ2 IHDDNH᯿9ٙ7p54ARRqohh^ s8> \>eX|"G1եK$|br6Wݹz^R*1*d|,CNСC4Y1ANt A~gd&&_sն&%%f9z47|!IZZ%EEi3/ꞘɄxjjJcc#o<ÑT*8 74fef~; 9K0_={w+~ dڦa0}̉D" X~T&MKMmgdt甇%911/:KLtTbBA*;~NllSjժ윜+2{7*:bKFwW$ @tT$bQ%ٜ6Ͳ ח.3isuMYYRlxB E͙3+V\S uEYcŎ.muMݱ\$Zx i׍;66f`S^ZZ*H y nTiy~|.6bl9Uuڟ5GB|\Zu7hϾDQGi6XN?ܫiTSSGStKsw=ߵиX|rs\mvGScM#+cx]l F";w_ f_[[ǻܛ-Sq8[W]?@C~^ԁu0{5ollm108T}ljF!=;KdrRL&&䦛offr:rbBQ[{ n{ddtdd0XL#,yX9,B.Tfjf{l51eŞeBF(J`0<3sp =5BQt0xI!K^|_5W^dDoټ1~;~ I7m\_20yv zkvNNtT䶭N4mZƩqQ!*C.+NjPk"TQQXBn> Ga;zHaA^B|ygL@R, O7pO3BB^~~C};;:}@`bb|9zNa]-Nȡ'!*EP8T(@  a'y˖\PmeK"]CL&+,۶u3r/_YPRZ)IW_5''K`2?䤍f~G6nP3*()-$%&\}Lǐ$$L-l۶7|w>[QysT*Ս D\\lttT\\,r}>>80p2-ûrRSR""""ht:F}=?0أbc䔔DRT*^ ݅@,_~Y&BWDtp=V 0]RR&3o*bQm/[ZP( Bi|Qo)h4B!+*a$uAG +K ?ھSp<6"; "Aad%eNY%pl1)> endobj 14 0 obj << /Type /Action /S /URI /URI (http://www.green500.org) >> endobj 15 0 obj << /Type /Annot /Subtype /Link /A 16 0 R /Border [0 0 0] /H /I /Rect [ 46.0157 604.7086 70.0187 613.8661 ] >> endobj 16 0 obj << /Type /Action /S /URI /URI (http://www.green500.org/) >> endobj 17 0 obj << /Type /Page /Parent 3 0 R /Annots [ 20 0 R ] /Contents 18 0 R >> endobj 18 0 obj << /Length 3213 >> stream 0.000 0.000 0.000 rg 0.000 0.000 0.000 RG 0.45 w 0 J [ ] 0 d 1.000 1.000 1.000 rg 46.016 46.016 519.969 699.969 re f 0.000 0.000 0.000 rg BT 46.016 734.263 Td /F1 12.0 Tf [(of the most energy-efficient heterogeneous supercomputing systems is based more on the GPU )] TJ ET BT 46.016 719.611 Td /F1 12.0 Tf [(than on the CPU. For instance, the 15 greenest systems are all accelerated by NVIDIA Kepler )] TJ ET BT 46.016 704.959 Td /F1 12.0 Tf [(K20 GPUs, but different Intel CPUs are sprinkled throughout the top 15 with Piz Daint at number )] TJ ET BT 46.016 690.307 Td /F1 12.0 Tf [(5 and CSIRO at number 7 having Intel Sandy Bridge CPUs and TSUBAME at number 8 having )] TJ ET BT 46.016 675.655 Td /F1 12.0 Tf [(Intel Westmere CPUs, both older CPU generations. The rest of the top 15 systems all use Intel )] TJ ET BT 46.016 661.003 Td /F1 12.0 Tf [(Ivy Bridge, the latest generation of CPU.)] TJ ET BT 46.016 634.351 Td /F1 12.0 Tf [(Homogeneous supercomputing systems from IBM, dubbed BlueGene/Q, dominate 24 of the next )] TJ ET BT 46.016 619.699 Td /F1 12.0 Tf [(25 spots of the Green500 with energy efficiency ranging between 2,178 MFLOPS/watt and 2,182 )] TJ ET BT 46.016 605.047 Td /F1 12.0 Tf [(MFLOPS/watt. In the meantime, Tianhe-2, the fastest supercomputer in the world according to )] TJ ET BT 46.016 590.395 Td /F1 12.0 Tf [(the Top500 list, still ranks as one of the 50 most energy-efficient supercomputers in the world at )] TJ ET BT 46.016 575.743 Td /F1 12.0 Tf [(number 49. Its heterogeneous architecture, accelerated by Intel Xeon Phi, achieves an efficiency )] TJ ET BT 46.016 561.091 Td /F1 12.0 Tf [(of 1,902 MFLOPS/watt even in spite of its tremendous size.)] TJ ET BT 46.016 534.439 Td /F1 12.0 Tf [(Assuming that TSUBAME-KFC?s energy efficiency could be scaled linearly to an exaflop )] TJ ET BT 46.016 519.787 Td /F1 12.0 Tf [(supercomputing system, one that can perform one trillion floating-point operations per second, )] TJ ET BT 46.016 505.135 Td /F1 12.0 Tf [(such a system would consume on the order of 225 megawatts \(MW\). ?Although this 225-)] TJ ET BT 46.016 490.483 Td /F1 12.0 Tf [(megawatt power envelope is still quite far from DARPA?s optimistic target of a 67-megawatt )] TJ ET BT 46.016 475.831 Td /F1 12.0 Tf [(power envelope, it is an order of magnitude better than the initial projection of a nearly 3000-)] TJ ET BT 46.016 461.179 Td /F1 12.0 Tf [(megawatt power envelope from 2007 when the first official Green500 list was launched,? says )] TJ ET BT 46.016 446.527 Td /F1 12.0 Tf [(Wu Feng of the Green500.)] TJ ET BT 65.460 419.875 Td /F1 12.0 Tf [(2007-2013 CompuGreen, LLC. Green500 and the Green500 logo are trademarks of )] TJ ET BT 202.632 405.223 Td /F1 12.0 Tf [(CompuGreen, LLC. All rights reserved.)] TJ ET 0.620 0.620 0.620 rg 46.016 395.542 519.969 0.750 re f 0.000 0.000 0.000 rg BT 46.016 379.122 Td /F3 10.7 Tf [(Source URL \(retrieved on )] TJ ET BT 177.616 379.122 Td /F4 10.7 Tf [(09/23/2015 - 17:35)] TJ ET BT 267.761 379.122 Td /F3 10.7 Tf [(\):)] TJ ET BT 277.831 379.122 Td /F1 10.7 Tf [(http://www.green500.org/news/green500-list-june-2014)] TJ ET 0.5334 w 0 J [ ] 0 d 277.831 376.988 m 537.533 376.988 l S endstream endobj 19 0 obj << /Type /Font /Subtype /Type1 /Name /F4 /BaseFont /Helvetica-BoldOblique /Encoding /WinAnsiEncoding >> endobj 20 0 obj << /Type /Annot /Subtype /Link /A 21 0 R /Border [0 0 0] /H /I /Rect [ 277.8314 378.1350 537.5332 388.9896 ] >> endobj 21 0 obj << /Type /Action /S /URI /URI (http://www.green500.org/news/green500-list-june-2014) >> endobj xref 0 22 0000000000 65535 f 0000000008 00000 n 0000000073 00000 n 0000000119 00000 n 0000000350 00000 n 0000000387 00000 n 0000000540 00000 n 0000000629 00000 n 0000006221 00000 n 0000006328 00000 n 0000006443 00000 n 0000006556 00000 n 0000012207 00000 n 0000029701 00000 n 0000029829 00000 n 0000029904 00000 n 0000030030 00000 n 0000030106 00000 n 0000030190 00000 n 0000033456 00000 n 0000033576 00000 n 0000033704 00000 n trailer << /Size 22 /Root 1 0 R /Info 5 0 R >> startxref 33808 %%EOF