20:
165:
or a cabinet. Where floating-point calculation hardware has not been provided, floating-point calculations are done in software, which takes more processor time, but avoids the cost of the extra hardware. For a particular computer architecture, the floating-point unit instructions may be
270:
Some floating-point hardware only supports the simplest operations: addition, subtraction, and multiplication. But even the most complex floating-point hardware has a finite number of operations it can support – for example, no FPUs directly support
306:
In some cases, FPUs may be specialized, and divided between simpler floating-point operations (mainly addition and multiplication) and more complicated operations, like division. In some cases, only the simple operations may be implemented in hardware or
431:
range, introducing a gate array to interface the ARM2 processor with the WE32206 to support the additional ARM floating-point instructions. Acorn later offered the FPA10 coprocessor, developed by ARM, for various machines fitted with the ARM3 processor.
278:
When a CPU is executing a program that calls for a floating-point operation that is not directly supported by the hardware, the CPU uses a series of simpler floating-point operations. In systems without any floating-point hardware, the CPU
352:
models could execute floating-point instructions without an add-on FPU (the MicroPDP-11/23 required an add-on microcode option), and offered add-on accelerators to further speed the execution of those instructions.
629:
348:, such as the PDP-11/45, PDP-11/34a, PDP-11/44, and PDP-11/70, supported an add-on floating-point unit to support floating-point instructions. The PDP-11/60, MicroPDP-11/23 and several
399:
respectively, although early ones were socketed for the 80287, since the 80387 did not exist yet. Other companies manufactured co-processors for the Intel x86 series. These included
157: – that are coprocessors not always built into the CPU – have FPUs as a rule, while first generations of GPUs did not). This could be a single
637:
936:
567:
Anderson, Stanley F.; Earle, John G.; Goldschmidt, Robert
Elliott; Powers, Don M. (January 1967). "The IBM System/360 Model 91: Floating-Point Execution Unit".
2255:
967:
1227:
2366:
1549:
2068:
669:
1346:
2225:
1791:
1608:
2579:
294:
The software that lists the necessary series of operations to emulate floating-point operations is often packaged in a floating-point
1571:
605:
2220:
2292:
825:
569:
368:, and typically sold as an optional add-on. It would only be purchased if needed to speed up or enable math-intensive programs.
174:
to run on systems with or without floating-point hardware. Emulation can be implemented on any of several levels: in the CPU as
2574:
2045:
416:
412:
201:
operations. This division varies significantly by architecture; some have dedicated floating-point registers, while some, like
1063:
752:
2989:
2113:
1376:
1220:
467:
series, but unlike IBM PC-compatible systems, sockets for adding the coprocessor were not as common in lower-end systems.
2999:
2140:
1267:
944:
491:
2307:
2135:
2108:
1487:
547:
315:
1109:
1086:
3158:
3122:
2685:
1578:
1544:
1539:
1458:
1423:
101:
When a CPU is executing a program that calls for a floating-point operation, there are three ways to carry it out:
3097:
2994:
2395:
2302:
2103:
1324:
1213:
2123:
1842:
1277:
833:
537:
272:
694:
2297:
2145:
2118:
1979:
1593:
1554:
1411:
1118:
1095:
1072:
1046:
1016:
993:
251:
124:
had floating-point arithmetic as a standard feature, one of its major improvements over its predecessor the
98:
do not have hardware support for floating-point operations (while they increasingly have them as standard).
3163:
2734:
2496:
1972:
1933:
1588:
1583:
1517:
1329:
1171:
891:
482:
capability. These add-on FPUs are host-processor-independent, possess their own programming requirements (
79:
calculations, but the accuracy can be low, so some systems prefer to compute these functions in software.
2361:
2058:
1756:
1453:
1007:
255:
630:"FSIN Documentation Improvements in the "Intel® 64 and IA-32 Architectures Software Developer's Manual""
3011:
2658:
2075:
1566:
1534:
1304:
1292:
1272:
1037:
3102:
3065:
3055:
1443:
375:, and most compatibles based on the 8088 or 8086 had a socket for the optional 8087 coprocessor. The
323:
146:
44:
3117:
2524:
2460:
2437:
2287:
2249:
2085:
2035:
2030:
1507:
1401:
1309:
542:
376:
372:
154:
892:"Implementing CORDIC algorithms – A single compact routine for computing transcendental functions"
658:
3173:
3070:
2853:
2747:
2711:
2628:
2612:
2454:
2243:
2202:
2190:
2053:
1967:
1888:
1653:
1314:
1257:
856:
Glass, L. Brent (January 1990). "Math
Coprocessors: A look at what they do, and how they do it".
738:
509:
365:
191:
91:
68:
197:
In most modern computer architectures, there is some division of floating-point operations from
2876:
2848:
2758:
2723:
2472:
2466:
2448:
2182:
2176:
2080:
1984:
1875:
1814:
1676:
1319:
1198:
785:
Nave, Rafi (March 1983). "Implementation of
Transcendental Functions on a Numerics Processor".
483:
284:
244:
236:
60:
142:
featured an "Auxiliary
Arithmetic Unit" for floating point and double-precision calculations.
3168:
3050:
2959:
2705:
2417:
2235:
1994:
1962:
1920:
1832:
1633:
1448:
1438:
1428:
1418:
1388:
1371:
1236:
896:
503:
475:
436:
288:
162:
83:
3080:
3016:
2602:
2324:
2214:
2161:
1693:
1406:
1262:
1244:
258:. Each physical integer core, two per module, is single-threaded, in contrast with Intel's
72:
8:
3127:
3112:
2932:
2783:
2765:
2729:
2371:
2318:
2095:
2011:
1893:
1748:
1643:
1502:
523:
routines are used in many FPUs to implement functions but not greatly increase gate count
515:
295:
247:, floating-point operations were sometimes pipelined separately from integer operations.
262:, where two virtual simultaneous threads share the resources of a single physical core.
2984:
2976:
2828:
2803:
2607:
2482:
2006:
1947:
1827:
1559:
1287:
813:
420:
228:
and 68882 for some kinds of floating-point instructions, mainly as a way to reduce the
158:
95:
2937:
2904:
2820:
2752:
2653:
2643:
2633:
2564:
2559:
2554:
2477:
2406:
2312:
2272:
1905:
1855:
1805:
1781:
1663:
1603:
1598:
1480:
1396:
987:
867:
829:
798:
758:
748:
586:
1156:
Fellows, Paul (March 1990). "Programming The ARM: The
Floating Point Co-processor".
3107:
3040:
3026:
2881:
2788:
2742:
2549:
2544:
2539:
2534:
2529:
2519:
2389:
2356:
2267:
2262:
2171:
2023:
2018:
2001:
1989:
1928:
1492:
1470:
1356:
1334:
1252:
794:
578:
428:
424:
179:
1132:
3021:
3006:
2954:
2858:
2833:
2670:
2663:
2514:
2509:
2504:
2443:
2351:
2341:
2063:
1898:
1850:
1613:
1497:
1465:
1366:
1361:
1282:
858:
708:
487:
471:
408:
318:
units to perform SIMD computation; an example of this is the augmentation of the
3132:
2966:
2949:
2942:
2838:
2695:
2432:
2346:
2277:
1860:
1822:
1771:
1766:
1761:
1475:
1299:
818:
526:
448:
444:
440:
361:
259:
225:
206:
87:
56:
3152:
2927:
2843:
1883:
1865:
1658:
1351:
871:
590:
396:
139:
76:
1786:
744:
3137:
3075:
2891:
2868:
2680:
2401:
1339:
686:
2922:
2886:
2597:
2569:
2427:
2282:
1205:
452:
388:
380:
339:
240:
221:
171:
150:
64:
52:
582:
153:
rather than as an integrated unit (but now in addition to the CPU, e.g.
2808:
2798:
2793:
2775:
2675:
2648:
1910:
1743:
1713:
1433:
479:
229:
217:
183:
314:
In some current architectures, the FPU functionality is combined with
128:. This was carried forward to its successors the 709, 7090, and 7094.
2899:
2896:
2638:
1708:
1686:
460:
308:
175:
105:
A floating-point unit emulator (a floating-point library in software)
459:
series. They were also commonly added to higher-end models of Apple
2914:
1733:
531:
280:
167:
48:
40:
762:
1723:
1681:
311:, while the more complex operations are implemented as software.
198:
125:
121:
1738:
1703:
1668:
520:
404:
392:
384:
357:
345:
327:
187:
566:
2196:
1728:
1698:
464:
456:
400:
170:
by a library of software functions; this may permit the same
132:
3060:
2208:
2128:
1718:
19:
23:
Collection of the x87 family of math coprocessors by Intel
1648:
1638:
921:
Yuen, A. K. (1988). "Intel's
Floating-Point Processors".
349:
319:
213:
202:
1058:
1056:
186:
code. When only integer functionality is available, the
1032:
1030:
1028:
1026:
606:"Intel Underestimates Error Bounds by 1.3 quintillion"
1174:(Press release). Acorn Computers Limited. 5 July 1993
1149:
1053:
330:
architecture used in newer Intel and AMD processors.
43:
system specially designed to carry out operations on
1023:
740:
Elementary
Functions: Algorithms and Implementation
817:
135:, which had floating point as a standard feature.
3150:
470:There are also add-on FPU coprocessor units for
383:-based systems were generally socketed for the
885:
883:
881:
490:, etc.) and are often provided with their own
478:(SBCs), which serve to provide floating-point
232:counts (and complexity) of the FPU subsystem.
1221:
1019:. 1973. Chapter 7 "Floating Point Processor".
732:
730:
728:
364:for the FPU to be entirely separate from the
254:uses a special FPU named FlexFPU, which uses
1196:
974:. Archived from the original on Mar 22, 2017
916:
914:
2226:Computer performance by orders of magnitude
1172:"Acorn Releases Floating Point Accelerator"
1164:
878:
811:
780:
778:
1235:
1228:
1214:
849:
805:
725:
965:
911:
391:-based machines – for the
265:
224:microprocessor series, as well as in the
212:CORDIC routines have been implemented in
1111:VAX – Hardware Handbook Volume II – 1986
826:John Wiley & Sons Australia, Limited
775:
86:, one or more FPUs may be integrated as
18:
1155:
1088:VAX – Hardware Handbook Volume I – 1986
743:(2nd ed.). Boston, Massachusetts:
570:IBM Journal of Research and Development
145:Historically, some systems implemented
3151:
889:
736:
603:
1209:
1133:"Western Electric 32206 co-processor"
966:Halfacree, Gareth (28 October 2010).
855:
67:. Some FPUs can also perform various
2197:Floating-point operations per second
920:
787:Microprocessing and Microprogramming
784:
695:CPB-267_GE-235-SystemManual_1963.pdf
435:Coprocessors were available for the
235:Floating-point operations are often
492:integrated development environments
287:operations that run on the integer
190:methods are most commonly used for
13:
1190:
301:
14:
3185:
3123:Semiconductor device fabrication
675:from the original on 2022-10-09.
47:numbers. Typical operations are
3098:History of general-purpose CPUs
1325:Nondeterministic Turing machine
1125:
1102:
1079:
1000:
959:
929:
411:opted for the WE32206 to offer
356:In the 1980s, it was common in
131:In 1963, Digital announced the
1278:Deterministic finite automaton
1197:Filiatreault, Raymond (2003).
937:"AMD Steamroller vs Bulldozer"
701:
679:
651:
622:
597:
560:
538:IBM hexadecimal floating point
333:
273:arbitrary-precision arithmetic
243:architectures without general
1:
2069:Simultaneous and heterogenous
1119:Digital Equipment Corporation
1096:Digital Equipment Corporation
1073:Digital Equipment Corporation
1047:Digital Equipment Corporation
1017:Digital Equipment Corporation
553:
548:Multiply–accumulate operation
283:it using a series of simpler
2753:Integrated memory controller
2735:Translation lookaside buffer
1934:Memory dependence prediction
1377:Random-access stored program
1330:Probabilistic Turing machine
1009:PDP-11/45 Processor Handbook
923:Electro/88 Conference Record
890:Jarvis, Pitts (1990-10-01).
799:10.1016/0165-6074(83)90151-5
737:Muller, Jean-Michel (2006).
636:. 2014-10-09. Archived from
604:Dawson, Bruce (2014-10-09).
250:The modular architecture of
111:Integrated FPU (in hardware)
7:
2209:Synaptic updates per second
497:
256:simultaneous multithreading
252:Bulldozer microarchitecture
205:, go as far as independent
10:
3190:
2613:Heterogeneous architecture
1535:Orthogonal instruction set
1305:Alternating Turing machine
1293:Quantum cellular automaton
337:
220:, 80287, 80387) up to the
115:
3103:Microprocessor chronology
3090:
3066:Dynamic frequency scaling
3039:
2975:
2913:
2867:
2819:
2774:
2694:
2621:
2590:
2495:
2416:
2380:
2334:
2234:
2221:Cache performance metrics
2160:
2094:
2044:
1955:
1946:
1919:
1874:
1841:
1813:
1804:
1624:
1527:
1516:
1387:
1243:
1039:PDP-11 Processor Handbook
992:: CS1 maint: unfit URL (
610:randomascii.wordpress.com
16:Part of a computer system
3118:Hardware security module
2461:Digital signal processor
2438:Graphics processing unit
2250:Graphics processing unit
543:Graphics processing unit
69:transcendental functions
3159:Central processing unit
3071:Dynamic voltage scaling
2854:Memory address register
2748:Branch target predictor
2712:Address generation unit
2455:Physics processing unit
2244:Central processing unit
2203:Transactions per second
2191:Instructions per second
2114:Array processing (SIMT)
1258:Stored-program computer
534:floating-point standard
510:Address generation unit
443:. These were common in
326:instruction set in the
192:transcendental function
92:central processing unit
2877:Hardwired control unit
2759:Memory management unit
2724:Memory management unit
2473:Secure cryptoprocessor
2467:Tensor Processing Unit
2449:Vision processing unit
2183:Cycles per instruction
2177:Instructions per cycle
2124:Associative processing
1815:Instruction pipelining
1237:Processor technologies
344:Several models of the
322:instructions set with
285:fixed-point arithmetic
266:Floating-point library
245:out-of-order execution
84:computer architectures
24:
2960:Sum-addressed decoder
2706:Arithmetic logic unit
1833:Classic RISC pipeline
1787:Epiphany architecture
1634:Motorola 68000 series
1065:MICRO/PDP-11 Handbook
968:"AMD unveils Flex FP"
504:Arithmetic logic unit
476:single-board computer
437:Motorola 68000 family
289:arithmetic logic unit
22:
3081:Performance per watt
2659:replacement policies
2325:Package on a package
2215:Performance per watt
2119:Pipelined processing
1889:Tomasulo's algorithm
1694:Clipper architecture
1550:Application-specific
1263:Finite-state machine
709:"Intel 80287 family"
3164:Computer arithmetic
3113:Digital electronics
2766:Instruction decoder
2718:Floating-point unit
2372:Soft microprocessor
2319:System in a package
1894:Reservation station
1424:Transport-triggered
1075:. 1983. p. 33.
814:Morse, Stephen Paul
583:10.1147/rd.111.0034
108:Add-on FPU hardware
96:embedded processors
82:In general-purpose
29:floating-point unit
2985:Integrated circuit
2829:Processor register
2483:Baseband processor
1828:Operand forwarding
1288:Cellular automaton
897:Dr. Dobb's Journal
687:"GE-2xx documents"
421:extended precision
159:integrated circuit
25:
3146:
3145:
3035:
3034:
2654:Instruction cache
2644:Scratchpad memory
2491:
2490:
2478:Network processor
2407:Network on a chip
2362:Ultra-low-voltage
2313:Multi-chip module
2156:
2155:
1942:
1941:
1929:Branch prediction
1906:Register renaming
1800:
1799:
1782:VISC architecture
1604:Quantum computing
1599:VISC architecture
1481:Secondary storage
1397:Microarchitecture
1357:Register machines
1160:. pp. 43–44.
1158:A&B Computing
1137:www.cpu-world.com
812:Palmer, John F.;
754:978-0-8176-4372-0
713:www.cpu-world.com
691:www.bitsavers.org
666:www.bitsavers.org
474:units (MCUs/ÎĽCs)/
39:) is a part of a
35:, colloquially a
3181:
3108:Processor design
3000:Power management
2882:Instruction unit
2743:Branch predictor
2692:
2691:
2390:System on a chip
2332:
2331:
2172:Transistor count
2096:Flynn's taxonomy
1953:
1952:
1811:
1810:
1614:Addressing modes
1525:
1524:
1471:Memory hierarchy
1335:Hypercomputation
1253:Abstract machine
1230:
1223:
1216:
1207:
1206:
1202:
1184:
1183:
1181:
1179:
1168:
1162:
1161:
1153:
1147:
1146:
1144:
1143:
1129:
1123:
1122:
1116:
1106:
1100:
1099:
1093:
1083:
1077:
1076:
1070:
1060:
1051:
1050:
1044:
1034:
1021:
1020:
1014:
1004:
998:
997:
991:
983:
981:
979:
963:
957:
956:
954:
952:
943:. Archived from
933:
927:
926:
918:
909:
908:
906:
905:
887:
876:
875:
853:
847:
846:
844:
843:
824:(1st ed.).
823:
809:
803:
802:
793:(3–4): 221–225.
782:
773:
772:
770:
769:
734:
723:
722:
720:
719:
705:
699:
698:
683:
677:
676:
674:
663:
659:"PDP-6 Handbook"
655:
649:
648:
646:
645:
626:
620:
619:
617:
616:
601:
595:
594:
564:
488:instruction sets
182:function, or in
180:operating system
94:; however, many
37:math coprocessor
3189:
3188:
3184:
3183:
3182:
3180:
3179:
3178:
3149:
3148:
3147:
3142:
3128:Tick–tock model
3086:
3042:
3031:
2971:
2955:Address decoder
2909:
2863:
2859:Program counter
2834:Status register
2815:
2770:
2730:Load–store unit
2697:
2690:
2617:
2586:
2487:
2444:Image processor
2419:
2412:
2382:
2376:
2352:Microcontroller
2342:Embedded system
2330:
2230:
2163:
2152:
2090:
2040:
1938:
1915:
1899:Re-order buffer
1870:
1851:Data dependency
1837:
1796:
1626:
1620:
1519:
1518:Instruction set
1512:
1498:Multiprocessing
1466:Cache hierarchy
1459:Register/memory
1383:
1283:Queue automaton
1239:
1234:
1193:
1191:Further reading
1188:
1187:
1177:
1175:
1170:
1169:
1165:
1154:
1150:
1141:
1139:
1131:
1130:
1126:
1114:
1108:
1107:
1103:
1091:
1085:
1084:
1080:
1068:
1062:
1061:
1054:
1042:
1036:
1035:
1024:
1012:
1006:
1005:
1001:
985:
984:
977:
975:
964:
960:
950:
948:
935:
934:
930:
919:
912:
903:
901:
888:
879:
854:
850:
841:
839:
838:. 9780471875697
836:
820:The 8087 Primer
810:
806:
783:
776:
767:
765:
755:
747:. p. 134.
735:
726:
717:
715:
707:
706:
702:
685:
684:
680:
672:
661:
657:
656:
652:
643:
641:
628:
627:
623:
614:
612:
602:
598:
565:
561:
556:
516:Load–store unit
500:
472:microcontroller
441:68881 and 68882
409:Acorn Computers
342:
336:
304:
302:Integrated FPUs
268:
118:
88:execution units
17:
12:
11:
5:
3187:
3177:
3176:
3174:Floating point
3171:
3166:
3161:
3144:
3143:
3141:
3140:
3135:
3133:Pin grid array
3130:
3125:
3120:
3115:
3110:
3105:
3100:
3094:
3092:
3088:
3087:
3085:
3084:
3078:
3073:
3068:
3063:
3058:
3053:
3047:
3045:
3037:
3036:
3033:
3032:
3030:
3029:
3024:
3019:
3014:
3009:
3004:
3003:
3002:
2997:
2992:
2981:
2979:
2973:
2972:
2970:
2969:
2967:Barrel shifter
2964:
2963:
2962:
2957:
2950:Binary decoder
2947:
2946:
2945:
2935:
2930:
2925:
2919:
2917:
2911:
2910:
2908:
2907:
2902:
2894:
2889:
2884:
2879:
2873:
2871:
2865:
2864:
2862:
2861:
2856:
2851:
2846:
2841:
2839:Stack register
2836:
2831:
2825:
2823:
2817:
2816:
2814:
2813:
2812:
2811:
2806:
2796:
2791:
2786:
2780:
2778:
2772:
2771:
2769:
2768:
2763:
2762:
2761:
2750:
2745:
2740:
2739:
2738:
2732:
2721:
2715:
2709:
2702:
2700:
2689:
2688:
2683:
2678:
2673:
2668:
2667:
2666:
2661:
2656:
2651:
2646:
2641:
2631:
2625:
2623:
2619:
2618:
2616:
2615:
2610:
2605:
2600:
2594:
2592:
2588:
2587:
2585:
2584:
2583:
2582:
2572:
2567:
2562:
2557:
2552:
2547:
2542:
2537:
2532:
2527:
2522:
2517:
2512:
2507:
2501:
2499:
2493:
2492:
2489:
2488:
2486:
2485:
2480:
2475:
2470:
2464:
2458:
2452:
2446:
2441:
2435:
2433:AI accelerator
2430:
2424:
2422:
2414:
2413:
2411:
2410:
2404:
2399:
2396:Multiprocessor
2393:
2386:
2384:
2378:
2377:
2375:
2374:
2369:
2364:
2359:
2354:
2349:
2347:Microprocessor
2344:
2338:
2336:
2335:By application
2329:
2328:
2322:
2316:
2310:
2305:
2300:
2295:
2290:
2285:
2280:
2278:Tile processor
2275:
2270:
2265:
2260:
2259:
2258:
2247:
2240:
2238:
2232:
2231:
2229:
2228:
2223:
2218:
2212:
2206:
2200:
2194:
2188:
2187:
2186:
2174:
2168:
2166:
2158:
2157:
2154:
2153:
2151:
2150:
2149:
2148:
2138:
2133:
2132:
2131:
2126:
2121:
2116:
2106:
2100:
2098:
2092:
2091:
2089:
2088:
2083:
2078:
2073:
2072:
2071:
2066:
2064:Hyperthreading
2056:
2050:
2048:
2046:Multithreading
2042:
2041:
2039:
2038:
2033:
2028:
2027:
2026:
2016:
2015:
2014:
2009:
1999:
1998:
1997:
1992:
1982:
1977:
1976:
1975:
1970:
1959:
1957:
1950:
1944:
1943:
1940:
1939:
1937:
1936:
1931:
1925:
1923:
1917:
1916:
1914:
1913:
1908:
1903:
1902:
1901:
1896:
1886:
1880:
1878:
1872:
1871:
1869:
1868:
1863:
1858:
1853:
1847:
1845:
1839:
1838:
1836:
1835:
1830:
1825:
1823:Pipeline stall
1819:
1817:
1808:
1802:
1801:
1798:
1797:
1795:
1794:
1789:
1784:
1779:
1776:
1775:
1774:
1772:z/Architecture
1769:
1764:
1759:
1751:
1746:
1741:
1736:
1731:
1726:
1721:
1716:
1711:
1706:
1701:
1696:
1691:
1690:
1689:
1684:
1679:
1671:
1666:
1661:
1656:
1651:
1646:
1641:
1636:
1630:
1628:
1622:
1621:
1619:
1618:
1617:
1616:
1606:
1601:
1596:
1591:
1586:
1581:
1576:
1575:
1574:
1564:
1563:
1562:
1552:
1547:
1542:
1537:
1531:
1529:
1522:
1514:
1513:
1511:
1510:
1505:
1500:
1495:
1490:
1485:
1484:
1483:
1478:
1476:Virtual memory
1468:
1463:
1462:
1461:
1456:
1451:
1446:
1436:
1431:
1426:
1421:
1416:
1415:
1414:
1404:
1399:
1393:
1391:
1385:
1384:
1382:
1381:
1380:
1379:
1374:
1369:
1364:
1354:
1349:
1344:
1343:
1342:
1337:
1332:
1327:
1322:
1317:
1312:
1307:
1300:Turing machine
1297:
1296:
1295:
1290:
1285:
1280:
1275:
1270:
1260:
1255:
1249:
1247:
1241:
1240:
1233:
1232:
1225:
1218:
1210:
1204:
1203:
1192:
1189:
1186:
1185:
1163:
1148:
1124:
1101:
1078:
1052:
1022:
999:
958:
928:
910:
877:
866:(1): 337–348.
848:
834:
804:
774:
753:
724:
700:
678:
650:
621:
596:
558:
557:
555:
552:
551:
550:
545:
540:
535:
529:
527:Execution unit
524:
518:
513:
507:
499:
496:
463:and Commodore
445:Motorola 68020
362:microcomputers
338:Main article:
335:
332:
303:
300:
267:
264:
260:Hyperthreading
226:Motorola 68881
216:coprocessors (
147:floating point
117:
114:
113:
112:
109:
106:
57:multiplication
45:floating-point
15:
9:
6:
4:
3:
2:
3186:
3175:
3172:
3170:
3167:
3165:
3162:
3160:
3157:
3156:
3154:
3139:
3136:
3134:
3131:
3129:
3126:
3124:
3121:
3119:
3116:
3114:
3111:
3109:
3106:
3104:
3101:
3099:
3096:
3095:
3093:
3089:
3082:
3079:
3077:
3074:
3072:
3069:
3067:
3064:
3062:
3059:
3057:
3054:
3052:
3049:
3048:
3046:
3044:
3038:
3028:
3025:
3023:
3020:
3018:
3015:
3013:
3010:
3008:
3005:
3001:
2998:
2996:
2993:
2991:
2988:
2987:
2986:
2983:
2982:
2980:
2978:
2974:
2968:
2965:
2961:
2958:
2956:
2953:
2952:
2951:
2948:
2944:
2941:
2940:
2939:
2936:
2934:
2931:
2929:
2928:Demultiplexer
2926:
2924:
2921:
2920:
2918:
2916:
2912:
2906:
2903:
2901:
2898:
2895:
2893:
2890:
2888:
2885:
2883:
2880:
2878:
2875:
2874:
2872:
2870:
2866:
2860:
2857:
2855:
2852:
2850:
2849:Memory buffer
2847:
2845:
2844:Register file
2842:
2840:
2837:
2835:
2832:
2830:
2827:
2826:
2824:
2822:
2818:
2810:
2807:
2805:
2802:
2801:
2800:
2797:
2795:
2792:
2790:
2787:
2785:
2784:Combinational
2782:
2781:
2779:
2777:
2773:
2767:
2764:
2760:
2757:
2756:
2754:
2751:
2749:
2746:
2744:
2741:
2736:
2733:
2731:
2728:
2727:
2725:
2722:
2719:
2716:
2713:
2710:
2707:
2704:
2703:
2701:
2699:
2693:
2687:
2684:
2682:
2679:
2677:
2674:
2672:
2669:
2665:
2662:
2660:
2657:
2655:
2652:
2650:
2647:
2645:
2642:
2640:
2637:
2636:
2635:
2632:
2630:
2627:
2626:
2624:
2620:
2614:
2611:
2609:
2606:
2604:
2601:
2599:
2596:
2595:
2593:
2589:
2581:
2578:
2577:
2576:
2573:
2571:
2568:
2566:
2563:
2561:
2558:
2556:
2553:
2551:
2548:
2546:
2543:
2541:
2538:
2536:
2533:
2531:
2528:
2526:
2523:
2521:
2518:
2516:
2513:
2511:
2508:
2506:
2503:
2502:
2500:
2498:
2494:
2484:
2481:
2479:
2476:
2474:
2471:
2468:
2465:
2462:
2459:
2456:
2453:
2450:
2447:
2445:
2442:
2439:
2436:
2434:
2431:
2429:
2426:
2425:
2423:
2421:
2415:
2408:
2405:
2403:
2400:
2397:
2394:
2391:
2388:
2387:
2385:
2379:
2373:
2370:
2368:
2365:
2363:
2360:
2358:
2355:
2353:
2350:
2348:
2345:
2343:
2340:
2339:
2337:
2333:
2326:
2323:
2320:
2317:
2314:
2311:
2309:
2306:
2304:
2301:
2299:
2296:
2294:
2291:
2289:
2286:
2284:
2281:
2279:
2276:
2274:
2271:
2269:
2266:
2264:
2261:
2257:
2254:
2253:
2251:
2248:
2245:
2242:
2241:
2239:
2237:
2233:
2227:
2224:
2222:
2219:
2216:
2213:
2210:
2207:
2204:
2201:
2198:
2195:
2192:
2189:
2184:
2181:
2180:
2178:
2175:
2173:
2170:
2169:
2167:
2165:
2159:
2147:
2144:
2143:
2142:
2139:
2137:
2134:
2130:
2127:
2125:
2122:
2120:
2117:
2115:
2112:
2111:
2110:
2107:
2105:
2102:
2101:
2099:
2097:
2093:
2087:
2084:
2082:
2079:
2077:
2074:
2070:
2067:
2065:
2062:
2061:
2060:
2057:
2055:
2052:
2051:
2049:
2047:
2043:
2037:
2034:
2032:
2029:
2025:
2022:
2021:
2020:
2017:
2013:
2010:
2008:
2005:
2004:
2003:
2000:
1996:
1993:
1991:
1988:
1987:
1986:
1983:
1981:
1978:
1974:
1971:
1969:
1966:
1965:
1964:
1961:
1960:
1958:
1954:
1951:
1949:
1945:
1935:
1932:
1930:
1927:
1926:
1924:
1922:
1918:
1912:
1909:
1907:
1904:
1900:
1897:
1895:
1892:
1891:
1890:
1887:
1885:
1884:Scoreboarding
1882:
1881:
1879:
1877:
1873:
1867:
1866:False sharing
1864:
1862:
1859:
1857:
1854:
1852:
1849:
1848:
1846:
1844:
1840:
1834:
1831:
1829:
1826:
1824:
1821:
1820:
1818:
1816:
1812:
1809:
1807:
1803:
1793:
1790:
1788:
1785:
1783:
1780:
1777:
1773:
1770:
1768:
1765:
1763:
1760:
1758:
1755:
1754:
1752:
1750:
1747:
1745:
1742:
1740:
1737:
1735:
1732:
1730:
1727:
1725:
1722:
1720:
1717:
1715:
1712:
1710:
1707:
1705:
1702:
1700:
1697:
1695:
1692:
1688:
1685:
1683:
1680:
1678:
1675:
1674:
1672:
1670:
1667:
1665:
1662:
1660:
1659:Stanford MIPS
1657:
1655:
1652:
1650:
1647:
1645:
1642:
1640:
1637:
1635:
1632:
1631:
1629:
1623:
1615:
1612:
1611:
1610:
1607:
1605:
1602:
1600:
1597:
1595:
1592:
1590:
1587:
1585:
1582:
1580:
1577:
1573:
1570:
1569:
1568:
1565:
1561:
1558:
1557:
1556:
1553:
1551:
1548:
1546:
1543:
1541:
1538:
1536:
1533:
1532:
1530:
1526:
1523:
1521:
1520:architectures
1515:
1509:
1506:
1504:
1501:
1499:
1496:
1494:
1491:
1489:
1488:Heterogeneous
1486:
1482:
1479:
1477:
1474:
1473:
1472:
1469:
1467:
1464:
1460:
1457:
1455:
1452:
1450:
1447:
1445:
1442:
1441:
1440:
1439:Memory access
1437:
1435:
1432:
1430:
1427:
1425:
1422:
1420:
1417:
1413:
1410:
1409:
1408:
1405:
1403:
1400:
1398:
1395:
1394:
1392:
1390:
1386:
1378:
1375:
1373:
1372:Random-access
1370:
1368:
1365:
1363:
1360:
1359:
1358:
1355:
1353:
1352:Stack machine
1350:
1348:
1345:
1341:
1338:
1336:
1333:
1331:
1328:
1326:
1323:
1321:
1318:
1316:
1313:
1311:
1308:
1306:
1303:
1302:
1301:
1298:
1294:
1291:
1289:
1286:
1284:
1281:
1279:
1276:
1274:
1271:
1269:
1268:with datapath
1266:
1265:
1264:
1261:
1259:
1256:
1254:
1251:
1250:
1248:
1246:
1242:
1238:
1231:
1226:
1224:
1219:
1217:
1212:
1211:
1208:
1200:
1195:
1194:
1173:
1167:
1159:
1152:
1138:
1134:
1128:
1120:
1113:
1112:
1105:
1097:
1090:
1089:
1082:
1074:
1067:
1066:
1059:
1057:
1048:
1041:
1040:
1033:
1031:
1029:
1027:
1018:
1011:
1010:
1003:
995:
989:
973:
969:
962:
947:on 9 May 2015
946:
942:
938:
932:
924:
917:
915:
899:
898:
893:
886:
884:
882:
873:
869:
865:
861:
860:
852:
837:
831:
827:
822:
821:
815:
808:
800:
796:
792:
788:
781:
779:
764:
760:
756:
750:
746:
742:
741:
733:
731:
729:
714:
710:
704:
696:
692:
688:
682:
671:
667:
660:
654:
640:on 2020-01-16
639:
635:
631:
625:
611:
607:
600:
592:
588:
584:
580:
576:
572:
571:
563:
559:
549:
546:
544:
541:
539:
536:
533:
530:
528:
525:
522:
519:
517:
514:
511:
508:
505:
502:
501:
495:
493:
489:
485:
481:
477:
473:
468:
466:
462:
458:
454:
450:
446:
442:
438:
433:
430:
426:
422:
418:
414:
410:
406:
402:
398:
394:
390:
389:80386/80386SX
386:
382:
378:
374:
369:
367:
363:
359:
354:
351:
347:
341:
331:
329:
325:
321:
317:
312:
310:
299:
297:
292:
290:
286:
282:
276:
274:
263:
261:
257:
253:
248:
246:
242:
239:. In earlier
238:
233:
231:
227:
223:
219:
215:
210:
208:
204:
200:
195:
193:
189:
185:
181:
177:
173:
169:
164:
163:circuit board
160:
156:
152:
148:
143:
141:
138:In 1963, the
136:
134:
129:
127:
123:
120:In 1954, the
110:
107:
104:
103:
102:
99:
97:
93:
89:
85:
80:
78:
77:trigonometric
74:
70:
66:
62:
58:
54:
50:
46:
42:
38:
34:
30:
21:
3169:Coprocessors
3138:Chip carrier
3076:Clock gating
2995:Mixed-signal
2892:Write buffer
2869:Control unit
2717:
2681:Clock signal
2420:accelerators
2402:Cypress PSoC
2059:Simultaneous
1876:Out-of-order
1508:Neuromorphic
1389:Architecture
1347:Belt machine
1340:Zeno machine
1273:Hierarchical
1199:"SIMPLY FPU"
1176:. Retrieved
1166:
1157:
1151:
1140:. Retrieved
1136:
1127:
1110:
1104:
1087:
1081:
1064:
1038:
1008:
1002:
976:. Retrieved
972:bit-tech.net
971:
961:
949:. Retrieved
945:the original
940:
931:
922:
902:. Retrieved
895:
863:
857:
851:
840:. Retrieved
819:
807:
790:
786:
766:. Retrieved
739:
716:. Retrieved
712:
703:
690:
681:
665:
653:
642:. Retrieved
638:the original
633:
624:
613:. Retrieved
609:
599:
577:(1): 34–53.
574:
568:
562:
469:
453:workstations
434:
371:The IBM PC,
370:
360:/compatible
355:
343:
313:
305:
293:
277:
269:
249:
234:
211:
196:
194:evaluation.
161:, an entire
144:
137:
130:
119:
100:
81:
36:
32:
28:
26:
2923:Multiplexer
2887:Data buffer
2598:Single-core
2570:bit slicing
2428:Coprocessor
2283:Coprocessor
2164:performance
2086:Cooperative
2076:Speculative
2036:Distributed
1995:Superscalar
1980:Instruction
1948:Parallelism
1921:Speculative
1753:System/3x0
1625:Instruction
1402:Von Neumann
1315:Post–Turing
925:: 48/5/1–7.
455:, like the
340:Coprocessor
334:Add-on FPUs
241:superscalar
172:object code
151:coprocessor
90:within the
73:exponential
65:square root
53:subtraction
3153:Categories
3043:management
2938:Multiplier
2799:Logic gate
2789:Sequential
2696:Functional
2676:Clock rate
2649:Data cache
2622:Components
2603:Multi-core
2591:Core count
2081:Preemptive
1985:Pipelining
1968:Bit-serial
1911:Wide-issue
1856:Structural
1778:Tilera ISA
1744:MicroBlaze
1714:ETRAX CRIS
1609:Comparison
1454:Load–store
1434:Endianness
1142:2021-11-06
904:2016-01-02
842:2016-01-02
835:0471875694
768:2015-12-01
763:2005048094
745:Birkhäuser
718:2019-01-15
697:, p. IV-4.
644:2020-01-16
615:2020-01-16
554:References
484:operations
480:arithmetic
429:Archimedes
184:user-space
2977:Circuitry
2897:Microcode
2821:Registers
2664:coherence
2639:CPU cache
2497:Word size
2162:Processor
1806:Execution
1709:DEC Alpha
1687:Power ISA
1503:Cognitive
1310:Universal
900:: 152–156
872:0360-5280
634:intel.com
591:0018-8646
461:Macintosh
309:microcode
237:pipelined
214:Intel x87
209:schemes.
203:Intel x86
176:microcode
2915:Datapath
2608:Manycore
2580:variable
2418:Hardware
2054:Temporal
1734:OpenRISC
1429:Cellular
1419:Dataflow
1412:modified
988:cite web
978:29 March
951:14 March
941:WCCFtech
816:(1984).
670:Archived
532:IEEE 754
498:See also
494:(IDEs).
427:powered
281:emulates
207:clocking
178:, as an
168:emulated
71:such as
61:division
49:addition
41:computer
3091:Related
3022:Quantum
3012:Digital
3007:Boolean
2905:Counter
2804:Quantum
2565:512-bit
2560:256-bit
2555:128-bit
2398:(MPSoC)
2383:on chip
2381:Systems
2199:(FLOPS)
2012:Process
1861:Control
1843:Hazards
1729:Itanium
1724:Unicore
1682:PowerPC
1407:Harvard
1367:Pointer
1362:Counter
1320:Quantum
1178:7 April
1121:. 1986.
1098:. 1985.
1049:. 1979.
451:-based
423:to its
397:80387SX
296:library
199:integer
149:with a
126:IBM 701
122:IBM 704
116:History
3027:Switch
3017:Analog
2755:(IMC)
2726:(MMU)
2575:others
2550:64-bit
2545:48-bit
2540:32-bit
2535:24-bit
2530:16-bit
2525:15-bit
2520:12-bit
2357:Mobile
2273:Stream
2268:Barrel
2263:Vector
2252:(GPU)
2211:(SUPS)
2179:(IPC)
2031:Memory
2024:Vector
2007:Thread
1990:Scalar
1792:Others
1739:RISC-V
1704:SuperH
1673:Power
1669:MIPS-X
1644:PDP-11
1493:Fabric
1245:Models
870:
832:
761:
751:
589:
521:CORDIC
439:, the
417:double
413:single
405:Weitek
387:, and
358:IBM PC
346:PDP-11
328:x86-64
188:CORDIC
140:GE-235
63:, and
3083:(PPW)
3041:Power
2933:Adder
2809:Array
2776:Logic
2737:(TLB)
2720:(FPU)
2714:(AGU)
2708:(ALU)
2698:units
2634:Cache
2515:8-bit
2510:4-bit
2505:1-bit
2469:(TPU)
2463:(DSP)
2457:(PPU)
2451:(VPU)
2440:(GPU)
2409:(NoC)
2392:(SoC)
2327:(PoP)
2321:(SiP)
2315:(MCM)
2256:GPGPU
2246:(CPU)
2236:Types
2217:(PPW)
2205:(TPS)
2193:(IPS)
2185:(CPI)
1956:Level
1767:S/390
1762:S/370
1757:S/360
1699:SPARC
1677:POWER
1560:TRIPS
1528:Types
1115:(PDF)
1092:(PDF)
1069:(PDF)
1043:(PDF)
1013:(PDF)
673:(PDF)
662:(PDF)
512:(AGU)
506:(ALU)
465:Amiga
457:Sun-3
449:68030
401:Cyrix
393:80387
385:80287
381:80286
222:80486
133:PDP-6
3061:ACPI
2794:Glue
2686:FIFO
2629:Core
2367:ASIP
2308:CPLD
2303:FPOA
2298:FPGA
2293:ASIC
2146:SPMD
2141:MIMD
2136:MISD
2129:SWAR
2109:SIMD
2104:SISD
2019:Data
2002:Task
1973:Word
1719:M32R
1664:MIPS
1627:sets
1594:ZISC
1589:NISC
1584:OISC
1579:MISC
1572:EPIC
1567:VLIW
1555:EDGE
1545:RISC
1540:CISC
1449:HUMA
1444:NUMA
1180:2021
994:link
980:2018
953:2022
868:ISSN
859:Byte
830:ISBN
759:LCCN
749:ISBN
587:ISSN
419:and
403:and
395:and
379:and
316:SIMD
230:gate
218:8087
155:GPUs
3056:APM
3051:PMU
2943:CPU
2900:ROM
2671:Bus
2288:PAL
1963:Bit
1749:LMC
1654:ARM
1649:x86
1639:VAX
795:doi
579:doi
425:ARM
366:CPU
350:VAX
324:SSE
320:x87
75:or
33:FPU
3155::
2990:3D
1135:.
1117:.
1094:.
1071:.
1055:^
1045:.
1025:^
1015:.
990:}}
986:{{
970:.
939:.
913:^
894:.
880:^
864:15
862:.
828:.
791:11
789:.
777:^
757:.
727:^
711:.
693:.
689:.
668:.
664:.
632:.
608:.
585:.
575:11
573:.
486:,
415:,
407:.
377:AT
373:XT
298:.
291:.
275:.
59:,
55:,
51:,
27:A
1229:e
1222:t
1215:v
1201:.
1182:.
1145:.
996:)
982:.
955:.
907:.
874:.
845:.
801:.
797::
771:.
721:.
647:.
618:.
593:.
581::
447:/
31:(
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
