31:
176:
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
281:
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
317:
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
442:
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.
289:
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
363:
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.
640:
359:, 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
410:
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
2264:
168: – 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
1236:
648:
2375:
1558:
947:
2077:
578:
Anderson, Stanley F.; Earle, John G.; Goldschmidt, Robert
Elliott; Powers, Don M. (January 1967). "The IBM System/360 Model 91: Floating-Point Execution Unit".
2234:
1800:
1617:
977:
1580:
2229:
2301:
580:
836:
2054:
427:
423:
763:
680:
2998:
2122:
1385:
1229:
3008:
2149:
1276:
502:
305:
The software that lists the necessary series of operations to emulate floating-point operations is often packaged in a floating-point
616:
558:
2316:
2144:
2117:
1496:
326:
1467:
3167:
3131:
2694:
1587:
1553:
1548:
1432:
379:, and typically sold as an optional add-on. It would only be purchased if needed to speed up or enable math-intensive programs.
185:
to run on systems with or without floating-point hardware. Emulation can be implemented on any of several levels: in the CPU as
3106:
3003:
2404:
2311:
2112:
1355:
1333:
1222:
212:
operations. This division varies significantly by architecture; some have dedicated floating-point registers, while some, like
1851:
1286:
1072:
844:
548:
283:
2306:
2154:
1988:
1602:
1563:
1420:
1127:
1104:
1081:
1055:
1025:
478:
series, but unlike IBM PC-compatible systems, sockets for adding the coprocessor were not as common in lower-end systems.
262:
3172:
2743:
2588:
2583:
2505:
1981:
1942:
1597:
1592:
1526:
1338:
1462:
955:
2370:
2067:
1765:
266:
1118:
1095:
3020:
2667:
2084:
1575:
1543:
1313:
1301:
1281:
112:
When a CPU is executing a program that calls for a floating-point operation, there are three ways to carry it out:
3111:
3074:
3064:
1452:
334:
157:
55:
3126:
2533:
2469:
2446:
2296:
2258:
2094:
2044:
2039:
1516:
1410:
1318:
553:
387:
383:
165:
1323:
705:
3182:
3079:
2862:
2756:
2720:
2637:
2621:
2463:
2252:
2211:
2199:
2062:
1976:
1897:
1662:
1266:
520:
376:
202:
135:
had floating-point arithmetic as a standard feature, one of its major improvements over its predecessor the
109:
do not have hardware support for floating-point operations (while they increasingly have them as standard).
102:
79:
1180:
902:
2885:
2857:
2767:
2732:
2481:
2475:
2457:
2191:
2185:
2089:
1993:
1884:
1823:
1685:
1328:
991:
494:
493:
capability. These add-on FPUs are host-processor-independent, possess their own programming requirements (
295:
255:
247:
90:
calculations, but the accuracy can be low, so some systems prefer to compute these functions in software.
71:
1016:
3177:
3059:
2968:
2714:
2426:
2244:
2003:
1971:
1929:
1841:
1642:
1457:
1447:
1437:
1427:
1397:
1380:
1245:
641:"FSIN Documentation Improvements in the "Intel® 64 and IA-32 Architectures Software Developer's Manual""
514:
486:
447:
299:
173:
94:
907:
1046:
3089:
3025:
2611:
2333:
2223:
2170:
1702:
1415:
1271:
1253:
386:, and most compatibles based on the 8088 or 8086 had a socket for the optional 8087 coprocessor. The
83:
3136:
2738:
526:
3121:
2941:
2792:
2774:
2380:
2327:
2132:
2127:
2104:
2020:
1902:
1757:
1652:
1511:
306:
903:"Implementing CORDIC algorithms – A single compact routine for computing transcendental functions"
669:
2993:
2985:
2837:
2812:
2616:
2491:
2015:
1956:
1836:
1568:
1296:
867:
Glass, L. Brent (January 1990). "Math
Coprocessors: A look at what they do, and how they do it".
824:
749:
431:
169:
106:
208:
In most modern computer architectures, there is some division of floating-point operations from
1207:
796:
Nave, Rafi (March 1983). "Implementation of
Transcendental Functions on a Numerics Processor".
2946:
2913:
2829:
2761:
2662:
2652:
2642:
2573:
2568:
2563:
2486:
2415:
2321:
2281:
1914:
1864:
1814:
1790:
1672:
1612:
1607:
1489:
1405:
971:
878:
840:
769:
759:
597:
153:
featured an "Auxiliary
Arithmetic Unit" for floating point and double-precision calculations.
3116:
3049:
2890:
2797:
2751:
2558:
2553:
2548:
2543:
2538:
2528:
2398:
2365:
2276:
2271:
2180:
2032:
2027:
2010:
1998:
1937:
1501:
1479:
1365:
1343:
1261:
805:
589:
439:
435:
190:
269:. Each physical integer core, two per module, is single-threaded, in contrast with Intel's
3030:
3015:
2963:
2867:
2842:
2679:
2672:
2523:
2518:
2513:
2452:
2360:
2350:
2072:
1907:
1859:
1622:
1506:
1474:
1375:
1370:
1291:
869:
498:
482:
419:
534:
routines are used in many FPUs to implement functions but not greatly increase gate count
258:, floating-point operations were sometimes pipelined separately from integer operations.
273:, where two virtual simultaneous threads share the resources of a single physical core.
3141:
2975:
2958:
2951:
2847:
2704:
2441:
2355:
2286:
1869:
1831:
1780:
1775:
1770:
1484:
1308:
537:
459:
455:
451:
372:
270:
239:
and 68882 for some kinds of floating-point instructions, mainly as a way to reduce the
236:
217:
98:
67:
3161:
2936:
2852:
1892:
1874:
1667:
1360:
809:
755:
407:
150:
87:
1165:
Fellows, Paul (March 1990). "Programming The ARM: The
Floating Point Co-processor".
3146:
3084:
2900:
2877:
2689:
2410:
1348:
17:
1141:
719:
329:
units to perform SIMD computation; an example of this is the augmentation of the
2931:
2895:
2606:
2578:
2436:
2291:
463:
399:
391:
350:
251:
232:
182:
161:
75:
63:
829:
2817:
2807:
2802:
2784:
2684:
2657:
1919:
1752:
1722:
1442:
490:
240:
228:
194:
882:
601:
2908:
2905:
2647:
1717:
1695:
471:
319:
186:
697:
2923:
1795:
1742:
1214:
542:
291:
178:
59:
51:
593:
164:
rather than as an integrated unit (but now in addition to the CPU, e.g.
1732:
1690:
209:
136:
132:
325:
In some current architectures, the FPU functionality is combined with
139:. This was carried forward to its successors the 709, 7090, and 7094.
3035:
1747:
1712:
1677:
531:
415:
368:
356:
338:
198:
116:
A floating-point unit emulator (a floating-point library in software)
470:
series. They were also commonly added to higher-end models of Apple
30:
2205:
1737:
1707:
773:
475:
467:
411:
143:
29:
322:, while the more complex operations are implemented as software.
3069:
2217:
2137:
1727:
1218:
1657:
1647:
403:
395:
360:
330:
224:
213:
181:
by a library of software functions; this may permit the same
34:
Collection of the x87 family of math coprocessors by Intel
932:
Yuen, A. K. (1988). "Intel's
Floating-Point Processors".
1067:
1065:
197:
code. When only integer functionality is available, the
1041:
1039:
1037:
1035:
617:"Intel Underestimates Error Bounds by 1.3 quintillion"
1183:(Press release). Acorn Computers Limited. 5 July 1993
341:
architecture used in newer Intel and AMD processors.
54:
system specially designed to carry out operations on
3099:
3048:
2984:
2922:
2876:
2828:
2783:
2703:
2630:
2599:
2504:
2425:
2389:
2343:
2243:
2169:
2103:
2053:
1964:
1955:
1928:
1883:
1850:
1822:
1813:
1633:
1536:
1525:
1396:
1252:
751:
Elementary
Functions: Algorithms and Implementation
828:
146:, which had floating point as a standard feature.
481:There are also add-on FPU coprocessor units for
394:-based systems were generally socketed for the
896:
894:
892:
501:, etc.) and are often provided with their own
489:(SBCs), which serve to provide floating-point
243:counts (and complexity) of the FPU subsystem.
1230:
1028:. 1973. Chapter 7 "Floating Point Processor".
743:
741:
739:
375:for the FPU to be entirely separate from the
265:uses a special FPU named FlexFPU, which uses
8:
927:
925:
2235:Computer performance by orders of magnitude
1181:"Acorn Releases Floating Point Accelerator"
791:
789:
2700:
2340:
1961:
1819:
1533:
1237:
1223:
1215:
402:-based machines – for the
235:microprocessor series, as well as in the
223:CORDIC routines have been implemented in
1120:VAX – Hardware Handbook Volume II – 1986
837:John Wiley & Sons Australia, Limited
97:, one or more FPUs may be integrated as
1097:VAX – Hardware Handbook Volume I – 1986
754:(2nd ed.). Boston, Massachusetts:
581:IBM Journal of Research and Development
570:
156:Historically, some systems implemented
976:: CS1 maint: archived copy as title (
969:
1142:"Western Electric 32206 co-processor"
78:. Some FPUs can also perform various
7:
2206:Floating-point operations per second
798:Microprocessing and Microprogramming
706:CPB-267_GE-235-SystemManual_1963.pdf
446:Coprocessors were available for the
246:Floating-point operations are often
503:integrated development environments
298:operations that run on the integer
201:methods are most commonly used for
25:
3132:Semiconductor device fabrication
686:from the original on 2022-10-09.
58:numbers. Typical operations are
3107:History of general-purpose CPUs
1334:Nondeterministic Turing machine
422:opted for the WE32206 to offer
367:In the 1980s, it was common in
142:In 1963, Digital announced the
1287:Deterministic finite automaton
1206:Filiatreault, Raymond (2003).
549:IBM hexadecimal floating point
284:arbitrary-precision arithmetic
254:architectures without general
1:
2078:Simultaneous and heterogenous
1128:Digital Equipment Corporation
1105:Digital Equipment Corporation
1082:Digital Equipment Corporation
1056:Digital Equipment Corporation
1026:Digital Equipment Corporation
559:Multiply–accumulate operation
294:it using a series of simpler
2762:Integrated memory controller
2744:Translation lookaside buffer
1943:Memory dependence prediction
1386:Random-access stored program
1339:Probabilistic Turing machine
1018:PDP-11/45 Processor Handbook
934:Electro/88 Conference Record
901:Jarvis, Pitts (1990-10-01).
810:10.1016/0165-6074(83)90151-5
748:Muller, Jean-Michel (2006).
647:. 2014-10-09. Archived from
615:Dawson, Bruce (2014-10-09).
261:The modular architecture of
122:Integrated FPU (in hardware)
2218:Synaptic updates per second
267:simultaneous multithreading
263:Bulldozer microarchitecture
216:, go as far as independent
3199:
2622:Heterogeneous architecture
1544:Orthogonal instruction set
1314:Alternating Turing machine
1302:Quantum cellular automaton
348:
231:, 80287, 80387) up to the
3112:Microprocessor chronology
3075:Dynamic frequency scaling
2230:Cache performance metrics
1048:PDP-11 Processor Handbook
621:randomascii.wordpress.com
27:Part of a computer system
3127:Hardware security module
2470:Digital signal processor
2447:Graphics processing unit
2259:Graphics processing unit
554:Graphics processing unit
80:transcendental functions
3168:Central processing unit
3080:Dynamic voltage scaling
2863:Memory address register
2757:Branch target predictor
2721:Address generation unit
2464:Physics processing unit
2253:Central processing unit
2212:Transactions per second
2200:Instructions per second
2123:Array processing (SIMT)
1267:Stored-program computer
545:floating-point standard
521:Address generation unit
454:. These were common in
337:instruction set in the
203:transcendental function
103:central processing unit
2886:Hardwired control unit
2768:Memory management unit
2733:Memory management unit
2482:Secure cryptoprocessor
2476:Tensor Processing Unit
2458:Vision processing unit
2192:Cycles per instruction
2186:Instructions per cycle
2133:Associative processing
1824:Instruction pipelining
1246:Processor technologies
355:Several models of the
333:instructions set with
296:fixed-point arithmetic
277:Floating-point library
256:out-of-order execution
95:computer architectures
35:
2969:Sum-addressed decoder
2715:Arithmetic logic unit
1842:Classic RISC pipeline
1796:Epiphany architecture
1643:Motorola 68000 series
1074:MICRO/PDP-11 Handbook
992:"AMD unveils Flex FP"
515:Arithmetic logic unit
487:single-board computer
448:Motorola 68000 family
300:arithmetic logic unit
33:
3090:Performance per watt
2668:replacement policies
2334:Package on a package
2224:Performance per watt
2128:Pipelined processing
1898:Tomasulo's algorithm
1703:Clipper architecture
1559:Application-specific
1272:Finite-state machine
720:"Intel 80287 family"
3173:Computer arithmetic
3122:Digital electronics
2775:Instruction decoder
2727:Floating-point unit
2381:Soft microprocessor
2328:System in a package
1903:Reservation station
1433:Transport-triggered
1084:. 1983. p. 33.
825:Morse, Stephen Paul
594:10.1147/rd.111.0034
119:Add-on FPU hardware
107:embedded processors
93:In general-purpose
40:floating-point unit
18:Floating point unit
2994:Integrated circuit
2838:Processor register
2492:Baseband processor
1837:Operand forwarding
1297:Cellular automaton
908:Dr. Dobb's Journal
698:"GE-2xx documents"
432:extended precision
170:integrated circuit
36:
3155:
3154:
3044:
3043:
2663:Instruction cache
2653:Scratchpad memory
2500:
2499:
2487:Network processor
2416:Network on a chip
2371:Ultra-low-voltage
2322:Multi-chip module
2165:
2164:
1951:
1950:
1938:Branch prediction
1915:Register renaming
1809:
1808:
1791:VISC architecture
1613:Quantum computing
1608:VISC architecture
1490:Secondary storage
1406:Microarchitecture
1366:Register machines
1169:. pp. 43–44.
1167:A&B Computing
1146:www.cpu-world.com
952:cdn3.wccftech.com
823:Palmer, John F.;
765:978-0-8176-4372-0
724:www.cpu-world.com
702:www.bitsavers.org
677:www.bitsavers.org
485:units (MCUs/ÎĽCs)/
50:) is a part of a
46:, colloquially a
16:(Redirected from
3190:
3117:Processor design
3009:Power management
2891:Instruction unit
2752:Branch predictor
2701:
2399:System on a chip
2341:
2181:Transistor count
2105:Flynn's taxonomy
1962:
1820:
1623:Addressing modes
1534:
1480:Memory hierarchy
1344:Hypercomputation
1262:Abstract machine
1239:
1232:
1225:
1216:
1211:
1193:
1192:
1190:
1188:
1177:
1171:
1170:
1162:
1156:
1155:
1153:
1152:
1138:
1132:
1131:
1125:
1115:
1109:
1108:
1102:
1092:
1086:
1085:
1079:
1069:
1060:
1059:
1053:
1043:
1030:
1029:
1023:
1013:
1007:
1006:
1004:
1002:
988:
982:
981:
975:
967:
965:
963:
954:. Archived from
944:
938:
937:
929:
920:
919:
917:
916:
898:
887:
886:
864:
858:
857:
855:
854:
835:(1st ed.).
834:
820:
814:
813:
804:(3–4): 221–225.
793:
784:
783:
781:
780:
745:
734:
733:
731:
730:
716:
710:
709:
694:
688:
687:
685:
674:
670:"PDP-6 Handbook"
666:
660:
659:
657:
656:
637:
631:
630:
628:
627:
612:
606:
605:
575:
499:instruction sets
193:function, or in
191:operating system
105:; however, many
48:math coprocessor
21:
3198:
3197:
3193:
3192:
3191:
3189:
3188:
3187:
3158:
3157:
3156:
3151:
3137:Tick–tock model
3095:
3051:
3040:
2980:
2964:Address decoder
2918:
2872:
2868:Program counter
2843:Status register
2824:
2779:
2739:Load–store unit
2706:
2699:
2626:
2595:
2496:
2453:Image processor
2428:
2421:
2391:
2385:
2361:Microcontroller
2351:Embedded system
2339:
2239:
2172:
2161:
2099:
2049:
1947:
1924:
1908:Re-order buffer
1879:
1860:Data dependency
1846:
1805:
1635:
1629:
1528:
1527:Instruction set
1521:
1507:Multiprocessing
1475:Cache hierarchy
1468:Register/memory
1392:
1292:Queue automaton
1248:
1243:
1205:
1202:
1200:Further reading
1197:
1196:
1186:
1184:
1179:
1178:
1174:
1164:
1163:
1159:
1150:
1148:
1140:
1139:
1135:
1123:
1117:
1116:
1112:
1100:
1094:
1093:
1089:
1077:
1071:
1070:
1063:
1051:
1045:
1044:
1033:
1021:
1015:
1014:
1010:
1000:
998:
990:
989:
985:
968:
961:
959:
948:"Archived copy"
946:
945:
941:
931:
930:
923:
914:
912:
900:
899:
890:
866:
865:
861:
852:
850:
849:. 9780471875697
847:
831:The 8087 Primer
822:
821:
817:
795:
794:
787:
778:
776:
766:
758:. p. 134.
747:
746:
737:
728:
726:
718:
717:
713:
696:
695:
691:
683:
672:
668:
667:
663:
654:
652:
639:
638:
634:
625:
623:
614:
613:
609:
577:
576:
572:
567:
527:Load–store unit
511:
483:microcontroller
452:68881 and 68882
420:Acorn Computers
353:
347:
315:
313:Integrated FPUs
279:
129:
99:execution units
28:
23:
22:
15:
12:
11:
5:
3196:
3194:
3186:
3185:
3183:Floating point
3180:
3175:
3170:
3160:
3159:
3153:
3152:
3150:
3149:
3144:
3142:Pin grid array
3139:
3134:
3129:
3124:
3119:
3114:
3109:
3103:
3101:
3097:
3096:
3094:
3093:
3087:
3082:
3077:
3072:
3067:
3062:
3056:
3054:
3046:
3045:
3042:
3041:
3039:
3038:
3033:
3028:
3023:
3018:
3013:
3012:
3011:
3006:
3001:
2990:
2988:
2982:
2981:
2979:
2978:
2976:Barrel shifter
2973:
2972:
2971:
2966:
2959:Binary decoder
2956:
2955:
2954:
2944:
2939:
2934:
2928:
2926:
2920:
2919:
2917:
2916:
2911:
2903:
2898:
2893:
2888:
2882:
2880:
2874:
2873:
2871:
2870:
2865:
2860:
2855:
2850:
2848:Stack register
2845:
2840:
2834:
2832:
2826:
2825:
2823:
2822:
2821:
2820:
2815:
2805:
2800:
2795:
2789:
2787:
2781:
2780:
2778:
2777:
2772:
2771:
2770:
2759:
2754:
2749:
2748:
2747:
2741:
2730:
2724:
2718:
2711:
2709:
2698:
2697:
2692:
2687:
2682:
2677:
2676:
2675:
2670:
2665:
2660:
2655:
2650:
2640:
2634:
2632:
2628:
2627:
2625:
2624:
2619:
2614:
2609:
2603:
2601:
2597:
2596:
2594:
2593:
2592:
2591:
2581:
2576:
2571:
2566:
2561:
2556:
2551:
2546:
2541:
2536:
2531:
2526:
2521:
2516:
2510:
2508:
2502:
2501:
2498:
2497:
2495:
2494:
2489:
2484:
2479:
2473:
2467:
2461:
2455:
2450:
2444:
2442:AI accelerator
2439:
2433:
2431:
2423:
2422:
2420:
2419:
2413:
2408:
2405:Multiprocessor
2402:
2395:
2393:
2387:
2386:
2384:
2383:
2378:
2373:
2368:
2363:
2358:
2356:Microprocessor
2353:
2347:
2345:
2344:By application
2338:
2337:
2331:
2325:
2319:
2314:
2309:
2304:
2299:
2294:
2289:
2287:Tile processor
2284:
2279:
2274:
2269:
2268:
2267:
2256:
2249:
2247:
2241:
2240:
2238:
2237:
2232:
2227:
2221:
2215:
2209:
2203:
2197:
2196:
2195:
2183:
2177:
2175:
2167:
2166:
2163:
2162:
2160:
2159:
2158:
2157:
2147:
2142:
2141:
2140:
2135:
2130:
2125:
2115:
2109:
2107:
2101:
2100:
2098:
2097:
2092:
2087:
2082:
2081:
2080:
2075:
2073:Hyperthreading
2065:
2059:
2057:
2055:Multithreading
2051:
2050:
2048:
2047:
2042:
2037:
2036:
2035:
2025:
2024:
2023:
2018:
2008:
2007:
2006:
2001:
1991:
1986:
1985:
1984:
1979:
1968:
1966:
1959:
1953:
1952:
1949:
1948:
1946:
1945:
1940:
1934:
1932:
1926:
1925:
1923:
1922:
1917:
1912:
1911:
1910:
1905:
1895:
1889:
1887:
1881:
1880:
1878:
1877:
1872:
1867:
1862:
1856:
1854:
1848:
1847:
1845:
1844:
1839:
1834:
1832:Pipeline stall
1828:
1826:
1817:
1811:
1810:
1807:
1806:
1804:
1803:
1798:
1793:
1788:
1785:
1784:
1783:
1781:z/Architecture
1778:
1773:
1768:
1760:
1755:
1750:
1745:
1740:
1735:
1730:
1725:
1720:
1715:
1710:
1705:
1700:
1699:
1698:
1693:
1688:
1680:
1675:
1670:
1665:
1660:
1655:
1650:
1645:
1639:
1637:
1631:
1630:
1628:
1627:
1626:
1625:
1615:
1610:
1605:
1600:
1595:
1590:
1585:
1584:
1583:
1573:
1572:
1571:
1561:
1556:
1551:
1546:
1540:
1538:
1531:
1523:
1522:
1520:
1519:
1514:
1509:
1504:
1499:
1494:
1493:
1492:
1487:
1485:Virtual memory
1477:
1472:
1471:
1470:
1465:
1460:
1455:
1445:
1440:
1435:
1430:
1425:
1424:
1423:
1413:
1408:
1402:
1400:
1394:
1393:
1391:
1390:
1389:
1388:
1383:
1378:
1373:
1363:
1358:
1353:
1352:
1351:
1346:
1341:
1336:
1331:
1326:
1321:
1316:
1309:Turing machine
1306:
1305:
1304:
1299:
1294:
1289:
1284:
1279:
1269:
1264:
1258:
1256:
1250:
1249:
1244:
1242:
1241:
1234:
1227:
1219:
1213:
1212:
1201:
1198:
1195:
1194:
1172:
1157:
1133:
1110:
1087:
1061:
1031:
1008:
983:
939:
921:
888:
877:(1): 337–348.
859:
845:
815:
785:
764:
735:
711:
689:
661:
632:
607:
569:
568:
566:
563:
562:
561:
556:
551:
546:
540:
538:Execution unit
535:
529:
524:
518:
510:
507:
474:and Commodore
456:Motorola 68020
373:microcomputers
349:Main article:
346:
343:
314:
311:
278:
275:
271:Hyperthreading
237:Motorola 68881
227:coprocessors (
158:floating point
128:
125:
124:
123:
120:
117:
68:multiplication
56:floating-point
26:
24:
14:
13:
10:
9:
6:
4:
3:
2:
3195:
3184:
3181:
3179:
3176:
3174:
3171:
3169:
3166:
3165:
3163:
3148:
3145:
3143:
3140:
3138:
3135:
3133:
3130:
3128:
3125:
3123:
3120:
3118:
3115:
3113:
3110:
3108:
3105:
3104:
3102:
3098:
3091:
3088:
3086:
3083:
3081:
3078:
3076:
3073:
3071:
3068:
3066:
3063:
3061:
3058:
3057:
3055:
3053:
3047:
3037:
3034:
3032:
3029:
3027:
3024:
3022:
3019:
3017:
3014:
3010:
3007:
3005:
3002:
3000:
2997:
2996:
2995:
2992:
2991:
2989:
2987:
2983:
2977:
2974:
2970:
2967:
2965:
2962:
2961:
2960:
2957:
2953:
2950:
2949:
2948:
2945:
2943:
2940:
2938:
2937:Demultiplexer
2935:
2933:
2930:
2929:
2927:
2925:
2921:
2915:
2912:
2910:
2907:
2904:
2902:
2899:
2897:
2894:
2892:
2889:
2887:
2884:
2883:
2881:
2879:
2875:
2869:
2866:
2864:
2861:
2859:
2858:Memory buffer
2856:
2854:
2853:Register file
2851:
2849:
2846:
2844:
2841:
2839:
2836:
2835:
2833:
2831:
2827:
2819:
2816:
2814:
2811:
2810:
2809:
2806:
2804:
2801:
2799:
2796:
2794:
2793:Combinational
2791:
2790:
2788:
2786:
2782:
2776:
2773:
2769:
2766:
2765:
2763:
2760:
2758:
2755:
2753:
2750:
2745:
2742:
2740:
2737:
2736:
2734:
2731:
2728:
2725:
2722:
2719:
2716:
2713:
2712:
2710:
2708:
2702:
2696:
2693:
2691:
2688:
2686:
2683:
2681:
2678:
2674:
2671:
2669:
2666:
2664:
2661:
2659:
2656:
2654:
2651:
2649:
2646:
2645:
2644:
2641:
2639:
2636:
2635:
2633:
2629:
2623:
2620:
2618:
2615:
2613:
2610:
2608:
2605:
2604:
2602:
2598:
2590:
2587:
2586:
2585:
2582:
2580:
2577:
2575:
2572:
2570:
2567:
2565:
2562:
2560:
2557:
2555:
2552:
2550:
2547:
2545:
2542:
2540:
2537:
2535:
2532:
2530:
2527:
2525:
2522:
2520:
2517:
2515:
2512:
2511:
2509:
2507:
2503:
2493:
2490:
2488:
2485:
2483:
2480:
2477:
2474:
2471:
2468:
2465:
2462:
2459:
2456:
2454:
2451:
2448:
2445:
2443:
2440:
2438:
2435:
2434:
2432:
2430:
2424:
2417:
2414:
2412:
2409:
2406:
2403:
2400:
2397:
2396:
2394:
2388:
2382:
2379:
2377:
2374:
2372:
2369:
2367:
2364:
2362:
2359:
2357:
2354:
2352:
2349:
2348:
2346:
2342:
2335:
2332:
2329:
2326:
2323:
2320:
2318:
2315:
2313:
2310:
2308:
2305:
2303:
2300:
2298:
2295:
2293:
2290:
2288:
2285:
2283:
2280:
2278:
2275:
2273:
2270:
2266:
2263:
2262:
2260:
2257:
2254:
2251:
2250:
2248:
2246:
2242:
2236:
2233:
2231:
2228:
2225:
2222:
2219:
2216:
2213:
2210:
2207:
2204:
2201:
2198:
2193:
2190:
2189:
2187:
2184:
2182:
2179:
2178:
2176:
2174:
2168:
2156:
2153:
2152:
2151:
2148:
2146:
2143:
2139:
2136:
2134:
2131:
2129:
2126:
2124:
2121:
2120:
2119:
2116:
2114:
2111:
2110:
2108:
2106:
2102:
2096:
2093:
2091:
2088:
2086:
2083:
2079:
2076:
2074:
2071:
2070:
2069:
2066:
2064:
2061:
2060:
2058:
2056:
2052:
2046:
2043:
2041:
2038:
2034:
2031:
2030:
2029:
2026:
2022:
2019:
2017:
2014:
2013:
2012:
2009:
2005:
2002:
2000:
1997:
1996:
1995:
1992:
1990:
1987:
1983:
1980:
1978:
1975:
1974:
1973:
1970:
1969:
1967:
1963:
1960:
1958:
1954:
1944:
1941:
1939:
1936:
1935:
1933:
1931:
1927:
1921:
1918:
1916:
1913:
1909:
1906:
1904:
1901:
1900:
1899:
1896:
1894:
1893:Scoreboarding
1891:
1890:
1888:
1886:
1882:
1876:
1875:False sharing
1873:
1871:
1868:
1866:
1863:
1861:
1858:
1857:
1855:
1853:
1849:
1843:
1840:
1838:
1835:
1833:
1830:
1829:
1827:
1825:
1821:
1818:
1816:
1812:
1802:
1799:
1797:
1794:
1792:
1789:
1786:
1782:
1779:
1777:
1774:
1772:
1769:
1767:
1764:
1763:
1761:
1759:
1756:
1754:
1751:
1749:
1746:
1744:
1741:
1739:
1736:
1734:
1731:
1729:
1726:
1724:
1721:
1719:
1716:
1714:
1711:
1709:
1706:
1704:
1701:
1697:
1694:
1692:
1689:
1687:
1684:
1683:
1681:
1679:
1676:
1674:
1671:
1669:
1668:Stanford MIPS
1666:
1664:
1661:
1659:
1656:
1654:
1651:
1649:
1646:
1644:
1641:
1640:
1638:
1632:
1624:
1621:
1620:
1619:
1616:
1614:
1611:
1609:
1606:
1604:
1601:
1599:
1596:
1594:
1591:
1589:
1586:
1582:
1579:
1578:
1577:
1574:
1570:
1567:
1566:
1565:
1562:
1560:
1557:
1555:
1552:
1550:
1547:
1545:
1542:
1541:
1539:
1535:
1532:
1530:
1529:architectures
1524:
1518:
1515:
1513:
1510:
1508:
1505:
1503:
1500:
1498:
1497:Heterogeneous
1495:
1491:
1488:
1486:
1483:
1482:
1481:
1478:
1476:
1473:
1469:
1466:
1464:
1461:
1459:
1456:
1454:
1451:
1450:
1449:
1448:Memory access
1446:
1444:
1441:
1439:
1436:
1434:
1431:
1429:
1426:
1422:
1419:
1418:
1417:
1414:
1412:
1409:
1407:
1404:
1403:
1401:
1399:
1395:
1387:
1384:
1382:
1381:Random-access
1379:
1377:
1374:
1372:
1369:
1368:
1367:
1364:
1362:
1361:Stack machine
1359:
1357:
1354:
1350:
1347:
1345:
1342:
1340:
1337:
1335:
1332:
1330:
1327:
1325:
1322:
1320:
1317:
1315:
1312:
1311:
1310:
1307:
1303:
1300:
1298:
1295:
1293:
1290:
1288:
1285:
1283:
1280:
1278:
1277:with datapath
1275:
1274:
1273:
1270:
1268:
1265:
1263:
1260:
1259:
1257:
1255:
1251:
1247:
1240:
1235:
1233:
1228:
1226:
1221:
1220:
1217:
1209:
1204:
1203:
1199:
1182:
1176:
1173:
1168:
1161:
1158:
1147:
1143:
1137:
1134:
1129:
1122:
1121:
1114:
1111:
1106:
1099:
1098:
1091:
1088:
1083:
1076:
1075:
1068:
1066:
1062:
1057:
1050:
1049:
1042:
1040:
1038:
1036:
1032:
1027:
1020:
1019:
1012:
1009:
997:
993:
987:
984:
979:
973:
958:on 9 May 2015
957:
953:
949:
943:
940:
935:
928:
926:
922:
910:
909:
904:
897:
895:
893:
889:
884:
880:
876:
872:
871:
863:
860:
848:
842:
838:
833:
832:
826:
819:
816:
811:
807:
803:
799:
792:
790:
786:
775:
771:
767:
761:
757:
753:
752:
744:
742:
740:
736:
725:
721:
715:
712:
707:
703:
699:
693:
690:
682:
678:
671:
665:
662:
651:on 2020-01-16
650:
646:
642:
636:
633:
622:
618:
611:
608:
603:
599:
595:
591:
587:
583:
582:
574:
571:
564:
560:
557:
555:
552:
550:
547:
544:
541:
539:
536:
533:
530:
528:
525:
522:
519:
516:
513:
512:
508:
506:
504:
500:
496:
492:
488:
484:
479:
477:
473:
469:
465:
461:
457:
453:
449:
444:
441:
437:
433:
429:
425:
421:
417:
413:
409:
405:
401:
400:80386/80386SX
397:
393:
389:
385:
380:
378:
374:
370:
365:
362:
358:
352:
344:
342:
340:
336:
332:
328:
323:
321:
312:
310:
308:
303:
301:
297:
293:
287:
285:
276:
274:
272:
268:
264:
259:
257:
253:
250:. In earlier
249:
244:
242:
238:
234:
230:
226:
221:
219:
215:
211:
206:
204:
200:
196:
192:
188:
184:
180:
175:
174:circuit board
171:
167:
163:
159:
154:
152:
149:In 1963, the
147:
145:
140:
138:
134:
131:In 1954, the
126:
121:
118:
115:
114:
113:
110:
108:
104:
100:
96:
91:
89:
88:trigonometric
85:
81:
77:
73:
69:
65:
61:
57:
53:
49:
45:
41:
32:
19:
3178:Coprocessors
3147:Chip carrier
3085:Clock gating
3004:Mixed-signal
2901:Write buffer
2878:Control unit
2726:
2690:Clock signal
2429:accelerators
2411:Cypress PSoC
2068:Simultaneous
1885:Out-of-order
1517:Neuromorphic
1398:Architecture
1356:Belt machine
1349:Zeno machine
1282:Hierarchical
1208:"SIMPLY FPU"
1185:. Retrieved
1175:
1166:
1160:
1149:. Retrieved
1145:
1136:
1119:
1113:
1096:
1090:
1073:
1047:
1017:
1011:
999:. Retrieved
996:bit-tech.net
995:
986:
960:. Retrieved
956:the original
951:
942:
933:
913:. Retrieved
906:
874:
868:
862:
851:. Retrieved
830:
818:
801:
797:
777:. Retrieved
750:
727:. Retrieved
723:
714:
701:
692:
676:
664:
653:. Retrieved
649:the original
644:
635:
624:. Retrieved
620:
610:
588:(1): 34–53.
585:
579:
573:
480:
464:workstations
445:
382:The IBM PC,
381:
371:/compatible
366:
354:
324:
316:
304:
288:
280:
260:
245:
222:
207:
205:evaluation.
172:, an entire
155:
148:
141:
130:
111:
92:
47:
43:
39:
37:
2932:Multiplexer
2896:Data buffer
2607:Single-core
2579:bit slicing
2437:Coprocessor
2292:Coprocessor
2173:performance
2095:Cooperative
2085:Speculative
2045:Distributed
2004:Superscalar
1989:Instruction
1957:Parallelism
1930:Speculative
1762:System/3x0
1634:Instruction
1411:Von Neumann
1324:Post–Turing
936:: 48/5/1–7.
466:, like the
351:Coprocessor
345:Add-on FPUs
252:superscalar
183:object code
162:coprocessor
101:within the
84:exponential
76:square root
64:subtraction
3162:Categories
3052:management
2947:Multiplier
2808:Logic gate
2798:Sequential
2705:Functional
2685:Clock rate
2658:Data cache
2631:Components
2612:Multi-core
2600:Core count
2090:Preemptive
1994:Pipelining
1977:Bit-serial
1920:Wide-issue
1865:Structural
1787:Tilera ISA
1753:MicroBlaze
1723:ETRAX CRIS
1618:Comparison
1463:Load–store
1443:Endianness
1151:2021-11-06
915:2016-01-02
853:2016-01-02
846:0471875694
779:2015-12-01
774:2005048094
756:Birkhäuser
729:2019-01-15
708:, p. IV-4.
655:2020-01-16
626:2020-01-16
565:References
495:operations
491:arithmetic
440:Archimedes
195:user-space
2986:Circuitry
2906:Microcode
2830:Registers
2673:coherence
2648:CPU cache
2506:Word size
2171:Processor
1815:Execution
1718:DEC Alpha
1696:Power ISA
1512:Cognitive
1319:Universal
911:: 152–156
883:0360-5280
645:intel.com
602:0018-8646
472:Macintosh
320:microcode
248:pipelined
225:Intel x87
220:schemes.
214:Intel x86
187:microcode
2924:Datapath
2617:Manycore
2589:variable
2427:Hardware
2063:Temporal
1743:OpenRISC
1438:Cellular
1428:Dataflow
1421:modified
1001:29 March
972:cite web
962:14 March
827:(1984).
681:Archived
543:IEEE 754
509:See also
505:(IDEs).
438:powered
292:emulates
218:clocking
189:, as an
179:emulated
82:such as
72:division
60:addition
52:computer
3100:Related
3031:Quantum
3021:Digital
3016:Boolean
2914:Counter
2813:Quantum
2574:512-bit
2569:256-bit
2564:128-bit
2407:(MPSoC)
2392:on chip
2390:Systems
2208:(FLOPS)
2021:Process
1870:Control
1852:Hazards
1738:Itanium
1733:Unicore
1691:PowerPC
1416:Harvard
1376:Pointer
1371:Counter
1329:Quantum
1187:7 April
1130:. 1986.
1107:. 1985.
1058:. 1979.
462:-based
434:to its
408:80387SX
307:library
210:integer
160:with a
137:IBM 701
133:IBM 704
127:History
3036:Switch
3026:Analog
2764:(IMC)
2735:(MMU)
2584:others
2559:64-bit
2554:48-bit
2549:32-bit
2544:24-bit
2539:16-bit
2534:15-bit
2529:12-bit
2366:Mobile
2282:Stream
2277:Barrel
2272:Vector
2261:(GPU)
2220:(SUPS)
2188:(IPC)
2040:Memory
2033:Vector
2016:Thread
1999:Scalar
1801:Others
1748:RISC-V
1713:SuperH
1682:Power
1678:MIPS-X
1653:PDP-11
1502:Fabric
1254:Models
881:
843:
772:
762:
600:
532:CORDIC
450:, the
428:double
424:single
416:Weitek
398:, and
369:IBM PC
357:PDP-11
339:x86-64
199:CORDIC
151:GE-235
74:, and
3092:(PPW)
3050:Power
2942:Adder
2818:Array
2785:Logic
2746:(TLB)
2729:(FPU)
2723:(AGU)
2717:(ALU)
2707:units
2643:Cache
2524:8-bit
2519:4-bit
2514:1-bit
2478:(TPU)
2472:(DSP)
2466:(PPU)
2460:(VPU)
2449:(GPU)
2418:(NoC)
2401:(SoC)
2336:(PoP)
2330:(SiP)
2324:(MCM)
2265:GPGPU
2255:(CPU)
2245:Types
2226:(PPW)
2214:(TPS)
2202:(IPS)
2194:(CPI)
1965:Level
1776:S/390
1771:S/370
1766:S/360
1708:SPARC
1686:POWER
1569:TRIPS
1537:Types
1124:(PDF)
1101:(PDF)
1078:(PDF)
1052:(PDF)
1022:(PDF)
684:(PDF)
673:(PDF)
523:(AGU)
517:(ALU)
476:Amiga
468:Sun-3
460:68030
412:Cyrix
404:80387
396:80287
392:80286
233:80486
144:PDP-6
3070:ACPI
2803:Glue
2695:FIFO
2638:Core
2376:ASIP
2317:CPLD
2312:FPOA
2307:FPGA
2302:ASIC
2155:SPMD
2150:MIMD
2145:MISD
2138:SWAR
2118:SIMD
2113:SISD
2028:Data
2011:Task
1982:Word
1728:M32R
1673:MIPS
1636:sets
1603:ZISC
1598:NISC
1593:OISC
1588:MISC
1581:EPIC
1576:VLIW
1564:EDGE
1554:RISC
1549:CISC
1458:HUMA
1453:NUMA
1189:2021
1003:2018
978:link
964:2022
879:ISSN
870:Byte
841:ISBN
770:LCCN
760:ISBN
598:ISSN
430:and
414:and
406:and
390:and
327:SIMD
241:gate
229:8087
166:GPUs
3065:APM
3060:PMU
2952:CPU
2909:ROM
2680:Bus
2297:PAL
1972:Bit
1758:LMC
1663:ARM
1658:x86
1648:VAX
806:doi
590:doi
436:ARM
377:CPU
361:VAX
335:SSE
331:x87
86:or
44:FPU
3164::
2999:3D
1144:.
1126:.
1103:.
1080:.
1064:^
1054:.
1034:^
1024:.
994:.
974:}}
970:{{
950:.
924:^
905:.
891:^
875:15
873:.
839:.
802:11
800:.
788:^
768:.
738:^
722:.
704:.
700:.
679:.
675:.
643:.
619:.
596:.
586:11
584:.
497:,
426:,
418:.
388:AT
384:XT
309:.
302:.
286:.
70:,
66:,
62:,
38:A
1238:e
1231:t
1224:v
1210:.
1191:.
1154:.
1005:.
980:)
966:.
918:.
885:.
856:.
812:.
808::
782:.
732:.
658:.
629:.
604:.
592::
458:/
42:(
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
