579:. To get better CPI values without pipelining, the number of execution units must be greater than the number of stages. For example, with six executions units, six new instructions are fetched in stage 1 only after the six previous instructions finish at stage 5, therefore on average the number of clock cycles it takes to execute an instruction is 5/6 (CPI = 5/6 < 1). To get better CPI values with pipelining, there must be at least two execution units. For example, with two executions units, two new instructions are fetched every clock cycle by exploiting instruction-level parallelism, therefore two different instructions would complete stage 5 in every clock cycle and on average the number of clock cycles it takes to execute an instruction is 1/2 (CPI = 1/2 < 1).
25:
366:
562:, therefore, since one could theoretically have five instructions in the five pipeline stages at once (one instruction per stage), a different instruction would complete stage 5 in every clock cycle and on average the number of clock cycles it takes to execute an instruction is 1 (CPI = 1). In this case, the processor is said to be
1269:
1109:
931:
183:
1351:
1194:
742:
1185:
554:, a new instruction is fetched in stage 1 only after the previous instruction finishes at stage 5, therefore the number of clock cycles it takes to execute an instruction is five (CPI = 5 > 1). In this case, the processor is said to be
1058:
1052:
512:
361:{\displaystyle \mathrm {CPI} :={\frac {\Sigma _{i}(\mathrm {IC} _{i})(\mathrm {CC} _{i})}{\mathrm {IC} }}={\frac {\Sigma _{i}(\mathrm {IC} _{i}\cdot \mathrm {CC} _{i})}{\Sigma _{i}(\mathrm {IC} _{i})}}}
982:
838:
1273:
1022:
1264:{\displaystyle {\text{Execution time}}(T)={\text{CPI}}\times {\text{Instruction count}}\times {\text{clock time}}={\frac {{\text{CPI}}\times {\text{Instruction Count}}}{\text{frequency}}}}
573:
processor, the best CPI attainable is 1. However, with a multiple-execution-unit processor, one may achieve even better CPI values (CPI < 1). In this case, the processor is said to be
453:
401:
650:
1113:
421:
2499:
1104:{\displaystyle {\text{Effective processor performance}}={\text{MIPS}}={\frac {\text{clock frequency}}{\text{CPI}}}\times {\frac {1}{\text{1 Million}}}}
760:
1471:
2610:
1793:
2312:
1590:
2469:
2035:
1852:
2823:
1815:
89:
2464:
1027:
926:{\displaystyle {\text{CPI}}={\frac {45000\times 1+32000\times 2+15000\times 2+8000\times 2}{100000}}={\frac {155000}{100000}}=1.55}
61:
458:
2536:
830:
42:
1346:{\displaystyle ={\frac {1.55\times 100000}{400\times 1000000}}={\frac {1.55}{4000}}=0.0003875\,{\text{sec}}=0.3875\,{\text{ms}}}
937:
2818:
2289:
68:
3233:
2357:
1620:
1464:
3243:
2384:
989:
75:
1511:
2551:
2379:
2352:
1731:
1428:
1397:
article provides a useful introduction to computer performance measurement for those readers interested in the topic.
108:
3366:
2929:
1822:
1788:
1783:
1702:
1667:
57:
3402:
3341:
3238:
2639:
2546:
2347:
1568:
1457:
1388:
514:
is the total instruction count. The summation sums over all instruction types for a given benchmarking process.
2367:
2086:
1521:
150:
46:
3417:
2541:
2389:
2362:
2223:
1837:
1798:
1655:
559:
618:
612:
606:
2978:
2740:
2216:
2177:
1832:
1827:
1761:
1573:
2605:
2302:
2000:
1697:
3255:
2902:
2319:
1810:
1778:
1548:
1536:
1516:
546:
Each stage requires one clock cycle and an instruction passes through the stages sequentially. Without
426:
374:
3346:
3309:
3299:
1687:
737:{\displaystyle {\text{CPI}}={\frac {5\times 50+4\times 25+4\times 15+3\times 8+3\times 2}{100}}=4.4}
82:
3361:
2768:
2704:
2681:
2531:
2493:
2329:
2279:
2274:
1751:
1645:
1553:
174:
3314:
3097:
2991:
2955:
2872:
2856:
2698:
2487:
2446:
2434:
2297:
2211:
2132:
1897:
1558:
1501:
1377:
1180:{\displaystyle ={\frac {400,000,000}{1.55\times 1000000}}={\frac {400}{1.55}}=258\,{\text{MIPS}}}
142:
35:
3120:
3092:
3002:
2967:
2716:
2710:
2692:
2420:
2324:
2228:
2119:
2058:
1920:
1563:
1371:
547:
158:
154:
3407:
3294:
3203:
2949:
2661:
2479:
2238:
2206:
2164:
2076:
1877:
1692:
1682:
1672:
1662:
1632:
1615:
1480:
1394:
551:
523:
122:
3324:
3260:
2846:
2568:
2458:
2405:
1937:
1650:
1506:
1488:
8:
3371:
3356:
3176:
3027:
3009:
2973:
2961:
2615:
2562:
2339:
2255:
2137:
1992:
1887:
1746:
3228:
3220:
3072:
3047:
2851:
2726:
2250:
2191:
2071:
1803:
1531:
406:
3181:
3148:
3064:
2996:
2897:
2887:
2877:
2808:
2803:
2798:
2721:
2650:
2556:
2516:
2149:
2099:
2049:
2025:
1907:
1847:
1842:
1724:
1640:
1424:
1417:
756:
593:
3351:
3284:
3270:
3125:
3032:
2986:
2793:
2788:
2783:
2778:
2773:
2763:
2633:
2600:
2511:
2506:
2415:
2267:
2262:
2245:
2233:
2172:
1736:
1714:
1600:
1578:
1496:
1361:
3265:
3250:
3198:
3102:
3077:
2914:
2907:
2758:
2753:
2748:
2687:
2595:
2585:
2307:
2142:
2094:
1857:
1741:
1709:
1610:
1605:
1526:
833:(Millions of instructions per second) rate, and execution time for this program.
558:. With pipelining, a new instruction is fetched every clock cycle by exploiting
3412:
3376:
3210:
3193:
3186:
3082:
2939:
2676:
2590:
2521:
2104:
2066:
2015:
2010:
2005:
1719:
1543:
1383:
570:
3396:
3171:
3087:
2127:
2109:
1902:
1595:
2030:
1443:
Advanced
Computer Architecture by Kai Hwang, Chapter 1, Exercise Problem 1.1
3381:
3319:
3135:
3112:
2924:
2645:
1583:
3166:
3130:
2841:
2813:
2671:
2526:
1449:
764:
575:
146:
3052:
3042:
3037:
3019:
2919:
2892:
2154:
1987:
1957:
1677:
3143:
3140:
2882:
1952:
1930:
24:
3158:
1977:
1967:
1925:
1419:
Computer
Organization and Design: The Hardware/Software Interface
1982:
1947:
1912:
2440:
1972:
1942:
1365:
1047:{\displaystyle {\text{MIPS}}\propto {\text{clock frequency}}}
507:{\displaystyle \mathrm {IC} =\Sigma _{i}(\mathrm {IC} _{i})}
3304:
2452:
2372:
1962:
403:
is the number of instructions for a given instruction type
977:{\displaystyle 400\,{\text{MHz}}=400,000,000\,{\text{Hz}}}
169:
The average of Cycles Per
Instruction in a given process (
1892:
1882:
752:
1276:
1197:
1116:
1061:
1030:
992:
940:
841:
653:
461:
429:
409:
377:
186:
1017:{\displaystyle {\text{MIPS}}\propto 1/{\text{CPI}}}
49:. Unsourced material may be challenged and removed.
1416:
1345:
1263:
1179:
1103:
1046:
1016:
976:
925:
736:
506:
455:is the clock-cycles for that instruction type and
447:
415:
395:
360:
16:The average number of clock cycles per instruction
1414:
3394:
1415:Patterson, David A.; Hennessy, John L. (1994).
153:for a program or program fragment. It is the
1465:
533:Instruction decode/Register fetch cycle (ID).
2470:Computer performance by orders of magnitude
1479:
1472:
1458:
1337:
1325:
1171:
968:
944:
109:Learn how and when to remove this message
596:, there are five types of instructions:
536:Execution/Effective address cycle (EX).
3395:
145:'s performance: the average number of
1453:
2441:Floating-point operations per second
47:adding citations to reliable sources
18:
13:
526:, with the following five stages:
491:
488:
474:
466:
463:
435:
432:
383:
380:
342:
339:
325:
309:
306:
291:
288:
274:
261:
258:
243:
240:
222:
219:
205:
194:
191:
188:
14:
3429:
448:{\displaystyle \mathrm {CC} _{i}}
396:{\displaystyle \mathrm {IC} _{i}}
3367:Semiconductor device fabrication
755:processor was used to execute a
23:
3342:History of general-purpose CPUs
1569:Nondeterministic Turing machine
1064:Effective processor performance
34:needs additional citations for
1522:Deterministic finite automaton
1437:
1408:
1209:
1203:
517:
501:
483:
352:
334:
319:
283:
253:
235:
232:
214:
173:) is defined by the following
1:
2313:Simultaneous and heterogenous
1401:
829:Determine the effective CPI,
560:instruction-level parallelism
530:Instruction fetch cycle (IF).
164:
2997:Integrated memory controller
2979:Translation lookaside buffer
2178:Memory dependence prediction
1621:Random-access stored program
1574:Probabilistic Turing machine
746:
587:
131:clock cycles per instruction
7:
2453:Synaptic updates per second
1355:
582:
10:
3434:
2857:Heterogeneous architecture
1779:Orthogonal instruction set
1549:Alternating Turing machine
1537:Quantum cellular automaton
3347:Microprocessor chronology
3334:
3310:Dynamic frequency scaling
3283:
3219:
3157:
3111:
3063:
3018:
2938:
2865:
2834:
2739:
2660:
2624:
2578:
2478:
2465:Cache performance metrics
2404:
2338:
2288:
2199:
2190:
2163:
2118:
2085:
2057:
2048:
1868:
1771:
1760:
1631:
1487:
3362:Hardware security module
2705:Digital signal processor
2682:Graphics processing unit
2494:Graphics processing unit
58:"Cycles per instruction"
3315:Dynamic voltage scaling
3098:Memory address register
2992:Branch target predictor
2956:Address generation unit
2699:Physics processing unit
2488:Central processing unit
2447:Transactions per second
2435:Instructions per second
2358:Array processing (SIMT)
1502:Stored-program computer
1378:Instructions per second
635:15% R-type instructions
3403:Instruction processing
3121:Hardwired control unit
3003:Memory management unit
2968:Memory management unit
2717:Secure cryptoprocessor
2711:Tensor Processing Unit
2693:Vision processing unit
2427:Cycles per instruction
2421:Instructions per cycle
2368:Associative processing
2059:Instruction pipelining
1481:Processor technologies
1372:Instructions per cycle
1347:
1265:
1181:
1105:
1048:
1018:
978:
927:
738:
638:8% branch instructions
632:25% store instructions
542:Write-back cycle (WB).
508:
449:
417:
397:
362:
159:instructions per cycle
155:multiplicative inverse
135:clocks per instruction
127:cycles per instruction
3204:Sum-addressed decoder
2950:Arithmetic logic unit
2077:Classic RISC pipeline
2031:Epiphany architecture
1878:Motorola 68000 series
1348:
1266:
1182:
1106:
1049:
1019:
979:
928:
739:
629:50% load instructions
552:multi-cycle processor
524:classic RISC pipeline
509:
450:
418:
398:
363:
141:) is one aspect of a
123:computer architecture
3418:Computer performance
3325:Performance per watt
2903:replacement policies
2569:Package on a package
2459:Performance per watt
2363:Pipelined processing
2133:Tomasulo's algorithm
1938:Clipper architecture
1794:Application-specific
1507:Finite-state machine
1274:
1195:
1114:
1059:
1028:
990:
938:
839:
651:
641:2% jump instructions
592:For the multi-cycle
539:Memory access (MEM).
459:
427:
407:
375:
184:
43:improve this article
3357:Digital electronics
3010:Instruction decoder
2962:Floating-point unit
2616:Soft microprocessor
2563:System in a package
2138:Reservation station
1668:Transport-triggered
784:Integer Arithmetic
759:with the following
3229:Integrated circuit
3073:Processor register
2727:Baseband processor
2072:Operand forwarding
1532:Cellular automaton
1343:
1261:
1177:
1101:
1044:
1014:
974:
923:
779:Clock cycle count
776:Instruction count
734:
645:then, the CPI is:
625:If a program has:
504:
445:
413:
393:
358:
3390:
3389:
3279:
3278:
2898:Instruction cache
2888:Scratchpad memory
2735:
2734:
2722:Network processor
2651:Network on a chip
2606:Ultra-low-voltage
2557:Multi-chip module
2400:
2399:
2186:
2185:
2173:Branch prediction
2150:Register renaming
2044:
2043:
2026:VISC architecture
1848:Quantum computing
1843:VISC architecture
1725:Secondary storage
1641:Microarchitecture
1601:Register machines
1341:
1329:
1317:
1304:
1259:
1258:
1253:
1252:Instruction Count
1245:
1234:
1226:
1225:Instruction count
1218:
1201:
1175:
1163:
1150:
1099:
1098:
1086:
1085:
1082:
1073:
1065:
1042:
1034:
1012:
996:
972:
948:
915:
902:
845:
827:
826:
817:Control transfer
773:Instruction TYPE
757:benchmark program
726:
657:
416:{\displaystyle i}
356:
265:
119:
118:
111:
93:
3425:
3352:Processor design
3244:Power management
3126:Instruction unit
2987:Branch predictor
2936:
2935:
2634:System on a chip
2576:
2575:
2416:Transistor count
2340:Flynn's taxonomy
2197:
2196:
2055:
2054:
1858:Addressing modes
1769:
1768:
1715:Memory hierarchy
1579:Hypercomputation
1497:Abstract machine
1474:
1467:
1460:
1451:
1450:
1444:
1441:
1435:
1434:
1422:
1412:
1362:Cycle per second
1352:
1350:
1349:
1344:
1342:
1339:
1330:
1327:
1318:
1310:
1305:
1303:
1292:
1281:
1270:
1268:
1267:
1262:
1260:
1256:
1255:
1254:
1251:
1246:
1243:
1240:
1235:
1232:
1227:
1224:
1219:
1216:
1202:
1199:
1186:
1184:
1183:
1178:
1176:
1173:
1164:
1156:
1151:
1149:
1138:
1121:
1110:
1108:
1107:
1102:
1100:
1096:
1092:
1087:
1083:
1080:
1079:
1074:
1071:
1066:
1063:
1053:
1051:
1050:
1045:
1043:
1040:
1035:
1032:
1023:
1021:
1020:
1015:
1013:
1010:
1008:
997:
994:
983:
981:
980:
975:
973:
970:
949:
946:
932:
930:
929:
924:
916:
908:
903:
898:
851:
846:
843:
770:
769:
743:
741:
740:
735:
727:
722:
663:
658:
655:
603:Store (4 cycles)
522:Let us assume a
513:
511:
510:
505:
500:
499:
494:
482:
481:
469:
454:
452:
451:
446:
444:
443:
438:
422:
420:
419:
414:
402:
400:
399:
394:
392:
391:
386:
367:
365:
364:
359:
357:
355:
351:
350:
345:
333:
332:
322:
318:
317:
312:
300:
299:
294:
282:
281:
271:
266:
264:
256:
252:
251:
246:
231:
230:
225:
213:
212:
202:
197:
175:weighted average
172:
114:
107:
103:
100:
94:
92:
51:
27:
19:
3433:
3432:
3428:
3427:
3426:
3424:
3423:
3422:
3393:
3392:
3391:
3386:
3372:Tick–tock model
3330:
3286:
3275:
3215:
3199:Address decoder
3153:
3107:
3103:Program counter
3078:Status register
3059:
3014:
2974:Load–store unit
2941:
2934:
2861:
2830:
2731:
2688:Image processor
2663:
2656:
2626:
2620:
2596:Microcontroller
2586:Embedded system
2574:
2474:
2407:
2396:
2334:
2284:
2182:
2159:
2143:Re-order buffer
2114:
2095:Data dependency
2081:
2040:
1870:
1864:
1763:
1762:Instruction set
1756:
1742:Multiprocessing
1710:Cache hierarchy
1703:Register/memory
1627:
1527:Queue automaton
1483:
1478:
1448:
1447:
1442:
1438:
1431:
1413:
1409:
1404:
1358:
1338:
1326:
1309:
1293:
1282:
1280:
1275:
1272:
1271:
1250:
1242:
1241:
1239:
1231:
1223:
1215:
1198:
1196:
1193:
1192:
1172:
1155:
1139:
1122:
1120:
1115:
1112:
1111:
1091:
1081:clock frequency
1078:
1070:
1062:
1060:
1057:
1056:
1041:clock frequency
1039:
1031:
1029:
1026:
1025:
1009:
1004:
993:
991:
988:
987:
969:
945:
939:
936:
935:
907:
852:
850:
842:
840:
837:
836:
806:Floating point
761:instruction mix
749:
664:
662:
654:
652:
649:
648:
600:Load (5 cycles)
590:
585:
520:
495:
487:
486:
477:
473:
462:
460:
457:
456:
439:
431:
430:
428:
425:
424:
408:
405:
404:
387:
379:
378:
376:
373:
372:
346:
338:
337:
328:
324:
323:
313:
305:
304:
295:
287:
286:
277:
273:
272:
270:
257:
247:
239:
238:
226:
218:
217:
208:
204:
203:
201:
187:
185:
182:
181:
170:
167:
115:
104:
98:
95:
52:
50:
40:
28:
17:
12:
11:
5:
3431:
3421:
3420:
3415:
3410:
3405:
3388:
3387:
3385:
3384:
3379:
3377:Pin grid array
3374:
3369:
3364:
3359:
3354:
3349:
3344:
3338:
3336:
3332:
3331:
3329:
3328:
3322:
3317:
3312:
3307:
3302:
3297:
3291:
3289:
3281:
3280:
3277:
3276:
3274:
3273:
3268:
3263:
3258:
3253:
3248:
3247:
3246:
3241:
3236:
3225:
3223:
3217:
3216:
3214:
3213:
3211:Barrel shifter
3208:
3207:
3206:
3201:
3194:Binary decoder
3191:
3190:
3189:
3179:
3174:
3169:
3163:
3161:
3155:
3154:
3152:
3151:
3146:
3138:
3133:
3128:
3123:
3117:
3115:
3109:
3108:
3106:
3105:
3100:
3095:
3090:
3085:
3083:Stack register
3080:
3075:
3069:
3067:
3061:
3060:
3058:
3057:
3056:
3055:
3050:
3040:
3035:
3030:
3024:
3022:
3016:
3015:
3013:
3012:
3007:
3006:
3005:
2994:
2989:
2984:
2983:
2982:
2976:
2965:
2959:
2953:
2946:
2944:
2933:
2932:
2927:
2922:
2917:
2912:
2911:
2910:
2905:
2900:
2895:
2890:
2885:
2875:
2869:
2867:
2863:
2862:
2860:
2859:
2854:
2849:
2844:
2838:
2836:
2832:
2831:
2829:
2828:
2827:
2826:
2816:
2811:
2806:
2801:
2796:
2791:
2786:
2781:
2776:
2771:
2766:
2761:
2756:
2751:
2745:
2743:
2737:
2736:
2733:
2732:
2730:
2729:
2724:
2719:
2714:
2708:
2702:
2696:
2690:
2685:
2679:
2677:AI accelerator
2674:
2668:
2666:
2658:
2657:
2655:
2654:
2648:
2643:
2640:Multiprocessor
2637:
2630:
2628:
2622:
2621:
2619:
2618:
2613:
2608:
2603:
2598:
2593:
2591:Microprocessor
2588:
2582:
2580:
2579:By application
2573:
2572:
2566:
2560:
2554:
2549:
2544:
2539:
2534:
2529:
2524:
2522:Tile processor
2519:
2514:
2509:
2504:
2503:
2502:
2491:
2484:
2482:
2476:
2475:
2473:
2472:
2467:
2462:
2456:
2450:
2444:
2438:
2432:
2431:
2430:
2418:
2412:
2410:
2402:
2401:
2398:
2397:
2395:
2394:
2393:
2392:
2382:
2377:
2376:
2375:
2370:
2365:
2360:
2350:
2344:
2342:
2336:
2335:
2333:
2332:
2327:
2322:
2317:
2316:
2315:
2310:
2308:Hyperthreading
2300:
2294:
2292:
2290:Multithreading
2286:
2285:
2283:
2282:
2277:
2272:
2271:
2270:
2260:
2259:
2258:
2253:
2243:
2242:
2241:
2236:
2226:
2221:
2220:
2219:
2214:
2203:
2201:
2194:
2188:
2187:
2184:
2183:
2181:
2180:
2175:
2169:
2167:
2161:
2160:
2158:
2157:
2152:
2147:
2146:
2145:
2140:
2130:
2124:
2122:
2116:
2115:
2113:
2112:
2107:
2102:
2097:
2091:
2089:
2083:
2082:
2080:
2079:
2074:
2069:
2067:Pipeline stall
2063:
2061:
2052:
2046:
2045:
2042:
2041:
2039:
2038:
2033:
2028:
2023:
2020:
2019:
2018:
2016:z/Architecture
2013:
2008:
2003:
1995:
1990:
1985:
1980:
1975:
1970:
1965:
1960:
1955:
1950:
1945:
1940:
1935:
1934:
1933:
1928:
1923:
1915:
1910:
1905:
1900:
1895:
1890:
1885:
1880:
1874:
1872:
1866:
1865:
1863:
1862:
1861:
1860:
1850:
1845:
1840:
1835:
1830:
1825:
1820:
1819:
1818:
1808:
1807:
1806:
1796:
1791:
1786:
1781:
1775:
1773:
1766:
1758:
1757:
1755:
1754:
1749:
1744:
1739:
1734:
1729:
1728:
1727:
1722:
1720:Virtual memory
1712:
1707:
1706:
1705:
1700:
1695:
1690:
1680:
1675:
1670:
1665:
1660:
1659:
1658:
1648:
1643:
1637:
1635:
1629:
1628:
1626:
1625:
1624:
1623:
1618:
1613:
1608:
1598:
1593:
1588:
1587:
1586:
1581:
1576:
1571:
1566:
1561:
1556:
1551:
1544:Turing machine
1541:
1540:
1539:
1534:
1529:
1524:
1519:
1514:
1504:
1499:
1493:
1491:
1485:
1484:
1477:
1476:
1469:
1462:
1454:
1446:
1445:
1436:
1429:
1406:
1405:
1403:
1400:
1399:
1398:
1391:
1386:
1384:Megahertz myth
1381:
1375:
1369:
1357:
1354:
1336:
1333:
1324:
1321:
1316:
1313:
1308:
1302:
1299:
1296:
1291:
1288:
1285:
1279:
1249:
1238:
1230:
1222:
1214:
1211:
1208:
1205:
1200:Execution time
1170:
1167:
1162:
1159:
1154:
1148:
1145:
1142:
1137:
1134:
1131:
1128:
1125:
1119:
1095:
1090:
1077:
1069:
1038:
1007:
1003:
1000:
967:
964:
961:
958:
955:
952:
943:
922:
919:
914:
911:
906:
901:
897:
894:
891:
888:
885:
882:
879:
876:
873:
870:
867:
864:
861:
858:
855:
849:
825:
824:
821:
818:
814:
813:
810:
807:
803:
802:
799:
796:
795:Data transfer
792:
791:
788:
785:
781:
780:
777:
774:
748:
745:
733:
730:
725:
721:
718:
715:
712:
709:
706:
703:
700:
697:
694:
691:
688:
685:
682:
679:
676:
673:
670:
667:
661:
643:
642:
639:
636:
633:
630:
623:
622:
616:
610:
604:
601:
589:
586:
584:
581:
571:execution-unit
569:With a single-
544:
543:
540:
537:
534:
531:
519:
516:
503:
498:
493:
490:
485:
480:
476:
472:
468:
465:
442:
437:
434:
412:
390:
385:
382:
369:
368:
354:
349:
344:
341:
336:
331:
327:
321:
316:
311:
308:
303:
298:
293:
290:
285:
280:
276:
269:
263:
260:
255:
250:
245:
242:
237:
234:
229:
224:
221:
216:
211:
207:
200:
196:
193:
190:
166:
163:
117:
116:
31:
29:
22:
15:
9:
6:
4:
3:
2:
3430:
3419:
3416:
3414:
3411:
3409:
3406:
3404:
3401:
3400:
3398:
3383:
3380:
3378:
3375:
3373:
3370:
3368:
3365:
3363:
3360:
3358:
3355:
3353:
3350:
3348:
3345:
3343:
3340:
3339:
3337:
3333:
3326:
3323:
3321:
3318:
3316:
3313:
3311:
3308:
3306:
3303:
3301:
3298:
3296:
3293:
3292:
3290:
3288:
3282:
3272:
3269:
3267:
3264:
3262:
3259:
3257:
3254:
3252:
3249:
3245:
3242:
3240:
3237:
3235:
3232:
3231:
3230:
3227:
3226:
3224:
3222:
3218:
3212:
3209:
3205:
3202:
3200:
3197:
3196:
3195:
3192:
3188:
3185:
3184:
3183:
3180:
3178:
3175:
3173:
3172:Demultiplexer
3170:
3168:
3165:
3164:
3162:
3160:
3156:
3150:
3147:
3145:
3142:
3139:
3137:
3134:
3132:
3129:
3127:
3124:
3122:
3119:
3118:
3116:
3114:
3110:
3104:
3101:
3099:
3096:
3094:
3093:Memory buffer
3091:
3089:
3088:Register file
3086:
3084:
3081:
3079:
3076:
3074:
3071:
3070:
3068:
3066:
3062:
3054:
3051:
3049:
3046:
3045:
3044:
3041:
3039:
3036:
3034:
3031:
3029:
3028:Combinational
3026:
3025:
3023:
3021:
3017:
3011:
3008:
3004:
3001:
3000:
2998:
2995:
2993:
2990:
2988:
2985:
2980:
2977:
2975:
2972:
2971:
2969:
2966:
2963:
2960:
2957:
2954:
2951:
2948:
2947:
2945:
2943:
2937:
2931:
2928:
2926:
2923:
2921:
2918:
2916:
2913:
2909:
2906:
2904:
2901:
2899:
2896:
2894:
2891:
2889:
2886:
2884:
2881:
2880:
2879:
2876:
2874:
2871:
2870:
2868:
2864:
2858:
2855:
2853:
2850:
2848:
2845:
2843:
2840:
2839:
2837:
2833:
2825:
2822:
2821:
2820:
2817:
2815:
2812:
2810:
2807:
2805:
2802:
2800:
2797:
2795:
2792:
2790:
2787:
2785:
2782:
2780:
2777:
2775:
2772:
2770:
2767:
2765:
2762:
2760:
2757:
2755:
2752:
2750:
2747:
2746:
2744:
2742:
2738:
2728:
2725:
2723:
2720:
2718:
2715:
2712:
2709:
2706:
2703:
2700:
2697:
2694:
2691:
2689:
2686:
2683:
2680:
2678:
2675:
2673:
2670:
2669:
2667:
2665:
2659:
2652:
2649:
2647:
2644:
2641:
2638:
2635:
2632:
2631:
2629:
2623:
2617:
2614:
2612:
2609:
2607:
2604:
2602:
2599:
2597:
2594:
2592:
2589:
2587:
2584:
2583:
2581:
2577:
2570:
2567:
2564:
2561:
2558:
2555:
2553:
2550:
2548:
2545:
2543:
2540:
2538:
2535:
2533:
2530:
2528:
2525:
2523:
2520:
2518:
2515:
2513:
2510:
2508:
2505:
2501:
2498:
2497:
2495:
2492:
2489:
2486:
2485:
2483:
2481:
2477:
2471:
2468:
2466:
2463:
2460:
2457:
2454:
2451:
2448:
2445:
2442:
2439:
2436:
2433:
2428:
2425:
2424:
2422:
2419:
2417:
2414:
2413:
2411:
2409:
2403:
2391:
2388:
2387:
2386:
2383:
2381:
2378:
2374:
2371:
2369:
2366:
2364:
2361:
2359:
2356:
2355:
2354:
2351:
2349:
2346:
2345:
2343:
2341:
2337:
2331:
2328:
2326:
2323:
2321:
2318:
2314:
2311:
2309:
2306:
2305:
2304:
2301:
2299:
2296:
2295:
2293:
2291:
2287:
2281:
2278:
2276:
2273:
2269:
2266:
2265:
2264:
2261:
2257:
2254:
2252:
2249:
2248:
2247:
2244:
2240:
2237:
2235:
2232:
2231:
2230:
2227:
2225:
2222:
2218:
2215:
2213:
2210:
2209:
2208:
2205:
2204:
2202:
2198:
2195:
2193:
2189:
2179:
2176:
2174:
2171:
2170:
2168:
2166:
2162:
2156:
2153:
2151:
2148:
2144:
2141:
2139:
2136:
2135:
2134:
2131:
2129:
2128:Scoreboarding
2126:
2125:
2123:
2121:
2117:
2111:
2110:False sharing
2108:
2106:
2103:
2101:
2098:
2096:
2093:
2092:
2090:
2088:
2084:
2078:
2075:
2073:
2070:
2068:
2065:
2064:
2062:
2060:
2056:
2053:
2051:
2047:
2037:
2034:
2032:
2029:
2027:
2024:
2021:
2017:
2014:
2012:
2009:
2007:
2004:
2002:
1999:
1998:
1996:
1994:
1991:
1989:
1986:
1984:
1981:
1979:
1976:
1974:
1971:
1969:
1966:
1964:
1961:
1959:
1956:
1954:
1951:
1949:
1946:
1944:
1941:
1939:
1936:
1932:
1929:
1927:
1924:
1922:
1919:
1918:
1916:
1914:
1911:
1909:
1906:
1904:
1903:Stanford MIPS
1901:
1899:
1896:
1894:
1891:
1889:
1886:
1884:
1881:
1879:
1876:
1875:
1873:
1867:
1859:
1856:
1855:
1854:
1851:
1849:
1846:
1844:
1841:
1839:
1836:
1834:
1831:
1829:
1826:
1824:
1821:
1817:
1814:
1813:
1812:
1809:
1805:
1802:
1801:
1800:
1797:
1795:
1792:
1790:
1787:
1785:
1782:
1780:
1777:
1776:
1774:
1770:
1767:
1765:
1764:architectures
1759:
1753:
1750:
1748:
1745:
1743:
1740:
1738:
1735:
1733:
1732:Heterogeneous
1730:
1726:
1723:
1721:
1718:
1717:
1716:
1713:
1711:
1708:
1704:
1701:
1699:
1696:
1694:
1691:
1689:
1686:
1685:
1684:
1683:Memory access
1681:
1679:
1676:
1674:
1671:
1669:
1666:
1664:
1661:
1657:
1654:
1653:
1652:
1649:
1647:
1644:
1642:
1639:
1638:
1636:
1634:
1630:
1622:
1619:
1617:
1616:Random-access
1614:
1612:
1609:
1607:
1604:
1603:
1602:
1599:
1597:
1596:Stack machine
1594:
1592:
1589:
1585:
1582:
1580:
1577:
1575:
1572:
1570:
1567:
1565:
1562:
1560:
1557:
1555:
1552:
1550:
1547:
1546:
1545:
1542:
1538:
1535:
1533:
1530:
1528:
1525:
1523:
1520:
1518:
1515:
1513:
1512:with datapath
1510:
1509:
1508:
1505:
1503:
1500:
1498:
1495:
1494:
1492:
1490:
1486:
1482:
1475:
1470:
1468:
1463:
1461:
1456:
1455:
1452:
1440:
1432:
1430:9781558602816
1426:
1421:
1420:
1411:
1407:
1396:
1392:
1390:
1387:
1385:
1382:
1379:
1376:
1373:
1370:
1367:
1363:
1360:
1359:
1353:
1334:
1331:
1322:
1319:
1314:
1311:
1306:
1300:
1297:
1294:
1289:
1286:
1283:
1277:
1247:
1236:
1228:
1220:
1212:
1206:
1190:
1187:
1168:
1165:
1160:
1157:
1152:
1146:
1143:
1140:
1135:
1132:
1129:
1126:
1123:
1117:
1093:
1088:
1075:
1067:
1054:
1036:
1005:
1001:
998:
984:
965:
962:
959:
956:
953:
950:
941:
933:
920:
917:
912:
909:
904:
899:
895:
892:
889:
886:
883:
880:
877:
874:
871:
868:
865:
862:
859:
856:
853:
847:
834:
832:
822:
819:
816:
815:
811:
808:
805:
804:
800:
797:
794:
793:
789:
786:
783:
782:
778:
775:
772:
771:
768:
766:
762:
758:
754:
744:
731:
728:
723:
719:
716:
713:
710:
707:
704:
701:
698:
695:
692:
689:
686:
683:
680:
677:
674:
671:
668:
665:
659:
646:
640:
637:
634:
631:
628:
627:
626:
620:
617:
614:
611:
608:
605:
602:
599:
598:
597:
595:
580:
578:
577:
572:
567:
565:
561:
557:
553:
549:
541:
538:
535:
532:
529:
528:
527:
525:
515:
496:
478:
470:
440:
410:
388:
347:
329:
314:
301:
296:
278:
267:
248:
227:
209:
198:
180:
179:
178:
176:
162:
160:
156:
152:
148:
144:
140:
136:
132:
128:
124:
113:
110:
102:
99:December 2009
91:
88:
84:
81:
77:
74:
70:
67:
63:
60: –
59:
55:
54:Find sources:
48:
44:
38:
37:
32:This article
30:
26:
21:
20:
3408:Clock signal
3382:Chip carrier
3320:Clock gating
3239:Mixed-signal
3136:Write buffer
3113:Control unit
2925:Clock signal
2664:accelerators
2646:Cypress PSoC
2426:
2303:Simultaneous
2120:Out-of-order
1752:Neuromorphic
1633:Architecture
1591:Belt machine
1584:Zeno machine
1517:Hierarchical
1439:
1418:
1410:
1191:
1188:
1055:
985:
934:
835:
828:
750:
647:
644:
624:
591:
574:
568:
563:
555:
545:
521:
370:
168:
147:clock cycles
138:
134:
130:
126:
120:
105:
96:
86:
79:
72:
65:
53:
41:Please help
36:verification
33:
3167:Multiplexer
3131:Data buffer
2842:Single-core
2814:bit slicing
2672:Coprocessor
2527:Coprocessor
2408:performance
2330:Cooperative
2320:Speculative
2280:Distributed
2239:Superscalar
2224:Instruction
2192:Parallelism
2165:Speculative
1997:System/3x0
1869:Instruction
1646:Von Neumann
1559:Post–Turing
1189:Therefore:
765:clock cycle
576:superscalar
518:Explanation
151:instruction
3397:Categories
3287:management
3182:Multiplier
3043:Logic gate
3033:Sequential
2940:Functional
2920:Clock rate
2893:Data cache
2866:Components
2847:Multi-core
2835:Core count
2325:Preemptive
2229:Pipelining
2212:Bit-serial
2155:Wide-issue
2100:Structural
2022:Tilera ISA
1988:MicroBlaze
1958:ETRAX CRIS
1853:Comparison
1698:Load–store
1678:Endianness
1402:References
1233:clock time
621:(3 cycles)
615:(3 cycles)
609:(4 cycles)
548:pipelining
165:Definition
69:newspapers
3221:Circuitry
3141:Microcode
3065:Registers
2908:coherence
2883:CPU cache
2741:Word size
2406:Processor
2050:Execution
1953:DEC Alpha
1931:Power ISA
1747:Cognitive
1554:Universal
1395:benchmark
1323:0.0003875
1298:×
1287:×
1257:frequency
1248:×
1229:×
1221:×
1144:×
1097:1 Million
1089:×
1037:∝
999:∝
893:×
881:×
869:×
857:×
747:Example 2
717:×
705:×
693:×
681:×
669:×
588:Example 1
556:subscalar
475:Σ
326:Σ
302:⋅
275:Σ
206:Σ
143:processor
3159:Datapath
2852:Manycore
2824:variable
2662:Hardware
2298:Temporal
1978:OpenRISC
1673:Cellular
1663:Dataflow
1656:modified
1356:See also
583:Examples
3335:Related
3266:Quantum
3256:Digital
3251:Boolean
3149:Counter
3048:Quantum
2809:512-bit
2804:256-bit
2799:128-bit
2642:(MPSoC)
2627:on chip
2625:Systems
2443:(FLOPS)
2256:Process
2105:Control
2087:Hazards
1973:Itanium
1968:Unicore
1926:PowerPC
1651:Harvard
1611:Pointer
1606:Counter
1564:Quantum
1301:1000000
1147:1000000
986:since:
767:count:
550:, in a
83:scholar
3271:Switch
3261:Analog
2999:(IMC)
2970:(MMU)
2819:others
2794:64-bit
2789:48-bit
2784:32-bit
2779:24-bit
2774:16-bit
2769:15-bit
2764:12-bit
2601:Mobile
2517:Stream
2512:Barrel
2507:Vector
2496:(GPU)
2455:(SUPS)
2423:(IPC)
2275:Memory
2268:Vector
2251:Thread
2234:Scalar
2036:Others
1983:RISC-V
1948:SuperH
1917:Power
1913:MIPS-X
1888:PDP-11
1737:Fabric
1489:Models
1427:
1335:0.3875
1290:100000
913:100000
910:155000
900:100000
809:15000
798:32000
787:45000
613:Branch
607:R-type
564:scalar
371:Where
85:
78:
71:
64:
56:
3413:Rates
3327:(PPW)
3285:Power
3177:Adder
3053:Array
3020:Logic
2981:(TLB)
2964:(FPU)
2958:(AGU)
2952:(ALU)
2942:units
2878:Cache
2759:8-bit
2754:4-bit
2749:1-bit
2713:(TPU)
2707:(DSP)
2701:(PPU)
2695:(VPU)
2684:(GPU)
2653:(NoC)
2636:(SoC)
2571:(PoP)
2565:(SiP)
2559:(MCM)
2500:GPGPU
2490:(CPU)
2480:Types
2461:(PPW)
2449:(TPS)
2437:(IPS)
2429:(CPI)
2200:Level
2011:S/390
2006:S/370
2001:S/360
1943:SPARC
1921:POWER
1804:TRIPS
1772:Types
1380:(IPS)
1374:(IPC)
878:15000
866:32000
854:45000
820:8000
751:A 400
137:, or
129:(aka
90:JSTOR
76:books
3305:ACPI
3038:Glue
2930:FIFO
2873:Core
2611:ASIP
2552:CPLD
2547:FPOA
2542:FPGA
2537:ASIC
2390:SPMD
2385:MIMD
2380:MISD
2373:SWAR
2353:SIMD
2348:SISD
2263:Data
2246:Task
2217:Word
1963:M32R
1908:MIPS
1871:sets
1838:ZISC
1833:NISC
1828:OISC
1823:MISC
1816:EPIC
1811:VLIW
1799:EDGE
1789:RISC
1784:CISC
1693:HUMA
1688:NUMA
1425:ISBN
1393:The
1389:MIPS
1315:4000
1312:1.55
1284:1.55
1174:MIPS
1161:1.55
1141:1.55
1072:MIPS
1033:MIPS
1024:and
995:MIPS
921:1.55
890:8000
831:MIPS
763:and
619:Jump
594:MIPS
149:per
62:news
3300:APM
3295:PMU
3187:CPU
3144:ROM
2915:Bus
2532:PAL
2207:Bit
1993:LMC
1898:ARM
1893:x86
1883:VAX
1328:sec
1295:400
1244:CPI
1217:CPI
1169:258
1158:400
1136:000
1130:000
1124:400
1084:CPI
1011:CPI
966:000
960:000
954:400
947:MHz
942:400
844:CPI
753:MHz
732:4.4
724:100
656:CPI
171:CPI
157:of
139:CPI
121:In
45:by
3399::
3234:3D
1423:.
1366:Hz
1340:ms
971:Hz
823:2
812:2
801:2
790:1
696:15
684:25
672:50
566:.
423:,
199::=
177::
161:.
133:,
125:,
1473:e
1466:t
1459:v
1433:.
1368:)
1364:(
1332:=
1320:=
1307:=
1278:=
1237:=
1213:=
1210:)
1207:T
1204:(
1166:=
1153:=
1133:,
1127:,
1118:=
1094:1
1076:=
1068:=
1006:/
1002:1
963:,
957:,
951:=
918:=
905:=
896:2
887:+
884:2
875:+
872:2
863:+
860:1
848:=
729:=
720:2
714:3
711:+
708:8
702:3
699:+
690:4
687:+
678:4
675:+
666:5
660:=
502:)
497:i
492:C
489:I
484:(
479:i
471:=
467:C
464:I
441:i
436:C
433:C
411:i
389:i
384:C
381:I
353:)
348:i
343:C
340:I
335:(
330:i
320:)
315:i
310:C
307:C
297:i
292:C
289:I
284:(
279:i
268:=
262:C
259:I
254:)
249:i
244:C
241:C
236:(
233:)
228:i
223:C
220:I
215:(
210:i
195:I
192:P
189:C
112:)
106:(
101:)
97:(
87:·
80:·
73:·
66:·
39:.
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
