Using Heuristic Value Prediction and Dynamic Task Granularity Resizing to Improve Software Speculation

<table>The structures of the global tables (there are only one Conflict Variable Table. The number of Address Mapping Tables and Read Mapping Tables/Write Mapping Tables is equal to the <svg height="15.4" id="M97" style="vertical-align:-3.22281pt" version="1.1" viewbox="0 0 46.549999 15.4" width="46.549999" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g transform="matrix(.017,-0,0,-.017,.062,11.112)"><path d="M857 650l-6 -28q-44 -4 -61.5 -16.5t-29.5 -48.5q-11 -32 -37 -166l-78 -399h-29l-351 537h-4l-56 -276q-24 -120 -24 -164q0 -35 17.5 -46t75.5 -15l-6 -28h-245l7 28q41 2 62 14t31 44q10 30 41 171l53 245q8 44 6.5 60.5t-14.5 33.5q-10 15 -27 19.5t-64 6.5l6 28h153
l350 -516h5l48 257q25 131 25 171q0 34 -17.5 45t-77.5 15l7 28h240z" id="x1D441"></path></g>
<g transform="matrix(.012,-0,0,-.012,14.925,15.187)"><path d="M298 36l-68 -35q-25 -13 -42 -13q-98 0 -98 119v290h-69l-4 11l28 29h45v72l65 67l14 -2v-137h113q8 -8 6 -21t-12 -19h-107v-256q0 -96 59 -96q29 0 61 17z" id="x74"></path></g><g transform="matrix(.012,-0,0,-.012,18.614,15.187)"><path d="M513 0h-211v26q44 5 55 17t11 60v166q0 123 -98 123q-59 0 -104 -51v-238q0 -48 10 -60.5t55 -16.5v-26h-213l1 26q46 4 57 16t11 61v477q0 48 -9.5 58.5t-58.5 14.5v24q63 9 147 35v-336q73 73 143 73q64 0 101 -42t37 -129v-175q0 -49 11 -61t55 -16v-26z" id="x68"></path></g><g transform="matrix(.012,-0,0,-.012,24.968,15.187)"><path d="M181 342h2q63 107 121 107q24 0 41 -16t17 -34q0 -34 -32 -49q-17 -7 -26 2q-21 20 -43 20q-20 0 -42 -22t-38 -63v-183q0 -50 12.5 -62t69.5 -16v-26h-230v26q47 5 58 17t11 61v207q0 49 -9 60.5t-53 16.5v23q74 12 141 40v-109z" id="x72"></path></g><g transform="matrix(.012,-0,0,-.012,29.239,15.187)"><path d="M380 106l15 -22q-42 -52 -87 -74t-76 -22q-89 0 -142 62t-53 151q0 103 66 175q67 73 146 73h1q61 0 105.5 -46.5t44.5 -105.5q0 -16 -4.5 -22.5t-18.5 -9.5q-54 -10 -262 -17q1 -95 46 -143.5t104 -48.5q61 0 115 50zM225 412q-35 0 -64.5 -32.5t-40.5 -90.5q83 0 170 3
q14 0 19 4.5t5 15.5q1 41 -23.5 70.5t-65.5 29.5z" id="x65"></path></g><g transform="matrix(.012,-0,0,-.012,34.356,15.187)"><path d="M433 39l-85 -51q-27 0 -47 20q-16 16 -24 44q-5 -3 -24 -16.5t-27 -18.5t-23 -13.5t-25 -12t-19 -3.5q-52 0 -86 36.5t-34 85.5q0 71 81 99q128 43 155 65v17q0 54 -23 83.5t-62 29.5q-28 0 -45 -19t-29 -66q-7 -23 -29 -23q-15 0 -29 13t-14 30t31 40q25 19 68.5 40.5
t81.5 29.5q49 0 82 -27q45 -39 45 -123v-185q0 -60 35 -60q18 0 36 11zM275 84v156q-12 -6 -49 -23t-41 -19q-31 -14 -46 -31t-15 -43q0 -35 22.5 -55.5t48.5 -20.5q19 0 42 10.5t38 25.5z" id="x61"></path></g><g transform="matrix(.012,-0,0,-.012,39.579,15.187)"><path d="M517 51v-25q-101 -18 -155 -38v65l-61 -36q-50 -29 -78 -29q-68 0 -126.5 58t-58.5 155q0 103 77 175.5t180 72.5q36 0 67 -12v143q0 48 -11.5 58.5t-65.5 14.5v25q79 9 156 34v-597q0 -38 7.5 -48t36.5 -13zM362 85v275q-35 51 -103 51q-20 0 -40.5 -8.5t-42 -28
t-35 -57.5t-13.5 -89q0 -83 42 -130t96 -47q51 0 96 34z" id="x64"></path></g>
</svg>).</table>

The Scientific World Journal

fig10

Figure 10

Figure 10: Using Heuristic Value Prediction and Dynamic Task Granularity Resizing to Improve Software Speculation 