403:
381:
I/O was not required. High priority microthreads provided (in decreasing priority) video, network, disk, a periodic timer, mouse, and keyboard. The microprogram did the complex logic of the I/O device, as well as the logic to integrate the device with the computer. For the actual hardware I/O, the microprogram read and wrote shift registers for most I/O, sometimes with resistor networks and transistors to shift output voltage levels (e.g. for video). To handle outside events, the microcontroller had microinterrupts to switch threads at the end of a thread's cycle, e.g. at the end of an instruction, or after a shift-register was accessed. The microprogram could be rewritten and reinstalled, which was very useful for a research computer.
321:
by constructing low leakage logic from large transistors that some processes provide for analog circuits. Some processes place the transistors above the surface of the silicon, in "fin fets", but these processes have more steps, so are more expensive. Special transistor doping materials (e.g. hafnium) can also reduce leakage, but this adds steps to the processing, making it more expensive. Some semiconductors have a larger band-gap than silicon. However, these materials and processes are currently (2020) more expensive than silicon.
139:
instructions, the pipeline sometimes must discard the data in process and restart. This is called a "stall." When two instructions could interfere, sometimes the control unit must stop processing a later instruction until an earlier instruction completes. This is called a "pipeline bubble" because a part of the pipeline is not processing instructions. Pipeline bubbles can occur when two instructions operate on the same register.
224:. If several instructions can be completed at the same time, the control unit will arrange it. So, the fastest computers can process instructions in a sequence that can vary somewhat, depending on when the operands or instruction destinations become available. Most supercomputers and many PC CPUs use this method. The exact organization of this type of control unit depends on the slowest part of the computer.
212:, the software also has to be designed to handle them. In general-purpose CPUs like PCs and smartphones, the threads are usually made to look very like normal time-sliced processes. At most, the operating system might need some awareness of them. In GPUs, the thread scheduling usually cannot be hidden from the application software, and is often controlled with a specialized subroutine library.
205:
usually have control units with a few threads, just enough to keep busy with affordable memory systems. Database computers often have about twice as many threads, to keep their much larger memories busy. Graphic processing units (GPUs) usually have hundreds or thousands of threads, because they have hundreds or thousands of execution units doing repetitive graphic calculations.
474:, that translates a microcode address into the control unit outputs. This truth table can be fed to a computer program that produces optimized electronic logic. The resulting control unit is almost as easy to design as microprogramming, but it has the fast speed and low number of logic elements of a hard wired control unit. The practical result resembles a
100:
very inexpensive, very simple, very reliable, or to get more work done, the control unit will finish the work in process before handling the interrupt. Finishing the work is inexpensive, because it needs no register to record the last finished instruction. It is simple and reliable because it has the fewest states. It also wastes the least amount of work.
313:
is the last to be turned off, and the first to be turned on. Also it then is the only CPU that requires special low-power features. A similar method is used in most PCs, which usually have an auxiliary embedded CPU that manages the power system. However, in PCs, the software is usually in the BIOS, not the operating system.
255:
an interrupt occurs to indicate that a memory access failed. This memory access must be associated with an exact instruction and an exact processor state, so that the processor's state can be saved and restored by the interrupt. A usual solution preserves copies of registers until a memory access completes.
194:. That is, they copy limited amounts of memory data into very fast memory. The CPU must be designed to process at the very fast speed of the cache memory. Therefore, the CPU might stall when it must access main memory directly. In modern PCs, main memory is as much as three hundred times slower than cache.
327:
Some CPUs make use of a special type of flip-flop (to store a bit) that couples a fast, high-leakage storage cell to a slow, large (expensive) low-leakage cell. These two cells have separated power supplies. When the CPU enters a power saving mode (e.g. because of a halt that waits for an interrupt),
320:
Some designs can use very low leakage transistors, but these usually add cost. The depletion barriers of the transistors can be made larger to have less leakage, but this makes the transistor larger and thus both slower and more expensive. Some vendors use this technique in selected portions of an IC
254:
Out of order controllers require special design features to handle interrupts. When there are several instructions in progress, it is not clear where in the instruction stream an interrupt occurs. For input and output interrupts, almost any solution works. However, when a computer has virtual memory,
235:
One kind of control unit for issuing uses an array of electronic logic, a "scoreboard" that detects when an instruction can be issued. The "height" of the array is the number of execution units, and the "length" and "width" are each the number of sources of operands. When all the items come together,
231:
It is common to have specialized execution units. For example, a modestly priced computer might have only one floating-point execution unit, because floating point units are expensive. The same computer might have several integer units, because these are relatively inexpensive, and can do the bulk of
204:
whose data has been fetched while the thread was idle. A thread has its own program counter, a stream of instructions and a separate set of registers. Designers vary the number of threads depending on current memory technologies and the type of computer. Typical computers such as PCs and smart phones
172:
can detect the most frequently-taken direction of a branch, the compiler can just produce instructions so that the most frequently taken branch is the preferred direction of branch. In a like way, a control unit might get hints from the compiler: Some computers have instructions that can encode hints
389:
Thus a program of instructions in memory will cause the CU to configure a CPU's data flows to manipulate the data correctly between instructions. This results in a computer that could run a complete program and require no human intervention to make hardware changes between instructions (as had to be
380:
had a multitasking microprogrammable control unit that performed almost all I/O. This design provided most of the features of a modern PC with only a tiny fraction of the electronic logic. The dual-thread computer was run by the two lowest-priority microthreads. These performed calculations whenever
339:
All modern CPUs have control logic to attach the CPU to the rest of the computer. In modern computers, this is usually a bus controller. When an instruction reads or writes memory, the control unit either controls the bus directly, or controls a bus controller. Many modern computers use the same bus
247:
If the execution is slower than writing the results, the memory write-back queue always has free entries. But what if the memory writes slowly? Or what if the destination register will be used by an "earlier" instruction that has not yet issued? Then the write-back step of the instruction might need
142:
Interrupts and unexpected exceptions also stall the pipeline. If a pipelined computer abandons work for an interrupt, more work is lost than in a multicycle computer. Predictable exceptions do not need to stall. For example, if an exception instruction is used to enter the operating system, it does
126:
In a pipelined computer, instructions flow through the computer. This design has several stages. For example, it might have one stage for each step of the Von
Neumann cycle. A pipelined computer usually has "pipeline registers" after each stage. These store the bits calculated by a stage so that the
99:
Control units can be designed to handle interrupts in one of two typical ways. If a quick response is most important, a control unit is designed to abandon work to handle the interrupt. In this case, the work in process will be restarted after the last completed instruction. If the computer is to be
312:
One common method is to spread the load to many CPUs, and turn off unused CPUs as the load reduces. The operating system's task switching logic saves the CPUs' data to memory. In some cases, one of the CPUs can be simpler and smaller, literally with fewer logic gates. So, it has low leakage, and it
296:
logic. CMOS wastes power in two common ways: By changing state, i.e. "active power", and by unintended leakage. The active power of a computer can be reduced by turning off control signals. Leakage current can be reduced by reducing the electrical pressure, the voltage, making the transistors with
227:
When the execution of calculations is the slowest, instructions flow from memory into pieces of electronics called "issue units." An issue unit holds an instruction until both its operands and an execution unit are available. Then, the instruction and its operands are "issued" to an execution unit.
146:
For the same speed of electronic logic, a pipelined computer can execute more instructions per second than a multicycle computer. Also, even though the electronic logic has a fixed maximum speed, a pipelined computer can be made faster or slower by varying the number of stages in the pipeline. With
134:
In a pipelined computer, the control unit arranges for the flow to start, continue, and stop as a program commands. The instruction data is usually passed in pipeline registers from one stage to the next, with a somewhat separated piece of control logic for each stage. The control unit also assures
110:
It is common for multicycle computers to use more cycles. Sometimes it takes longer to take a conditional jump, because the program counter has to be reloaded. Sometimes they do multiplication or division instructions by a process, something like binary long multiplication and division. Very small
167:
Control units use many methods to keep a pipeline full and avoid stalls. For example, even simple control units can assume that a backwards branch, to a lower-numbered, earlier instruction, is a loop, and will be repeated. So, a control unit with this design will always fill the pipeline with the
316:
Theoretically, computers at lower clock speeds could also reduce leakage by reducing the voltage of the power supply. This affects the reliability of the computer in many ways, so the engineering is expensive, and it is uncommon except in relatively expensive computers such as PCs or cellphones.
158:
However, a pipelined computer is usually more complex and more costly than a comparable multicycle computer. It typically has more logic gates, registers and a more complex control unit. In a like way, it might use more total energy, while using less energy per instruction. Out-of-order CPUs can
83:
Multicycle control units typically use both the rising and falling edges of their square-wave timing clock. They operate a step of their operation on each edge of the timing clock, so that a four-step operation completes in two clock cycles. This doubles the speed of the computer, given the same
304:
Most computers also have a "halt" instruction. This was invented to stop non-interrupt code so that interrupt code has reliable timing. However, designers soon noticed that a halt instruction was also a good time to turn off a CPU's clock completely, reducing the CPU's active power to zero. The
361:
Many historic computers built some type of input and output directly into the control unit. For example, many historic computers had a front panel with switches and lights directly controlled by the control unit. These let a programmer directly enter a program and debug it. In later production
154:
A pipelined model of a computer often has less logic gates per instruction per second than multicycle and out-of-order computers. This is because the average stage is less complex than a multicycle computer. An out-of-order computer usually has large amounts of idle logic at any given instant.
76:
successively. This consists of fetching the instruction, fetching the operands, decoding the instruction, executing the instruction, and then writing the results back to memory. When the next instruction is placed in the control unit, it changes the behavior of the control unit to complete the
138:
When operating efficiently, a pipelined computer will have an instruction in each stage. It is then working on all of those instructions at the same time. It can finish about one instruction for each cycle of its clock. When a program makes a decision, and switches to a different sequence of
258:
Also, out of order CPUs have even more problems with stalls from branching, because they can complete several instructions per clock cycle, and usually have many instructions in various stages of progress. So, these control units might use all of the solutions used by pipelined processors.
358:. The cache controller and the associated cache memory is often the largest physical part of a modern, higher-performance CPU. When the memory, bus or cache is shared with other CPUs, the control logic must communicate with them to assure that no computer ever gets out-of-date old data.
308:
These methods are relatively easy to design, and became so common that others were invented for commercial advantage. Many modern low-power CMOS CPUs stop and start specialized execution units and bus interfaces depending on the needed instruction. Some computers even arrange the CPU's
300:
Active power is easier to reduce because data stored in the logic is not affected. The usual method reduces the CPU's clock rate. Most computer systems use this method. It is common for a CPU to idle during the transition to avoid side-effects from the changing clock.
288:
Many modern computers have controls that minimize power usage. In battery-powered computers, such as those in cell-phones, the advantage is longer battery life. In computers with utility power, the justification is to reduce the cost of power, cooling or noise.
243:
With some additional logic, a scoreboard can compactly combine execution reordering, register renaming and precise exceptions and interrupts. Further it can do this without the power-hungry, complex content-addressable memory used by the
Tomasulo algorithm.
236:
the signals from the operands and execution unit will cross. The logic at this intersection detects that the instruction can work, so the instruction is "issued" to the free execution unit. An alternative style of issuing control unit implements the
95:
is caused by the computer's operation. One crucial difference is that the timing of an interrupt cannot be predicted. Another is that some exceptions (e.g. a memory-not-available exception) can be caused by an instruction that needs to be restarted.
279:
In these, the "front" of the control unit manages the translation of instructions. Operands are not translated. The "back" of the CU is an out-of-order CPU that issues the micro-operations and operands to the execution units and data paths.
228:
The execution unit does the instruction. Then the resulting data is moved into a queue of data to be written back to memory or registers. If the computer has multiple execution units, it can usually do several instructions per clock cycle.
135:
that the instruction in each stage does not harm the operation of instructions in other stages. For example, if two stages must use the same piece of data, the control logic assures that the uses are done in the correct sequence.
267:
Some computers translate each single instruction into a sequence of simpler instructions. The advantage is that an out of order computer can be simpler in the bulk of its logic, while handling complex multi-step instructions.
461:
description. The main advantage of a microprogrammed control unit is the simplicity of its structure. Outputs from the controller are by microinstructions. The microprogram can be debugged and replaced similarly to software.
417:
units, featuring a finite number of gates that can generate specific results based on the instructions that were used to invoke those responses. Hardwired control units are generally faster than the microprogrammed designs.
248:
to be scheduled. This is sometimes called "retiring" an instruction. In this case, there must be scheduling logic on the back end of execution units. It schedules access to the registers or memory that will get the results.
240:, which reorders a hardware queue of instructions. In some sense, both styles utilize a queue. The scoreboard is an alternative way to encode and reorder a queue of instructions, and some designers call it a queue table.
130:
It is common for even numbered stages to operate on one edge of the square-wave clock, while odd-numbered stages operate on the other edge. This speeds the computer by a factor of two compared to single-edge designs.
180:: A control unit keeps an electronic list of the recent branches, encoded by the address of the branch instruction. This list has a few bits for each branch to remember the direction that was taken most recently.
373:
models had a data bus designed to let I/O devices borrow the control unit's memory read and write logic. This reduced the complexity and expense of high speed I/O controllers, e.g. for disk.
428:
A controller that uses this approach can operate at high speed; however, it has little flexibility. A complex instruction set can overwhelm a designer who uses ad hoc logic design.
362:
computers, the most common use of a front panel was to enter a small bootstrap program to read the operating system from disk. This was annoying. So, front panels were replaced by
340:
interface for memory, input and output. This is called "memory-mapped I/O". To a programmer, the registers of the I/O devices appear as numbers at specific memory addresses.
328:
data is transferred to the low-leakage cells, and the others are turned off. When the CPU leaves a low-leakage mode (e.g. because of an interrupt), the process is reversed.
850:
431:
The hardwired approach has become less popular as computers have evolved. Previously, control units for CPUs used ad hoc logic, and they were difficult to design.
816:
2035:
57:. In modern computer designs, the control unit is typically an internal part of the CPU with its overall role and operation unchanged since its introduction.
672:
331:
Older designs would copy the CPU state to memory, or even disk, sometimes with specialized software. Very simple embedded systems sometimes just restart.
187:, in which a computer might have two or more pipelines, calculate both directions of a branch, and then discard the calculations of the unused direction.
402:
1007:
457:
and stored in special control memory. The algorithm for the microprogram control unit, unlike the hardwired control unit, is usually specified by
324:
Managing leakage is more difficult, because before the logic can be turned-off, the data in it must be moved to some type of low-leakage storage.
470:
A popular variation on microcode is to debug the microcode using a software simulator. Then, the microcode is a table of bits. This is a logical
251:
Retiring logic can also be designed into an issuing scoreboard or a
Tomasulo queue, by including memory or register access in the issuing logic.
2146:
1329:
69:. These were the earliest designs. They are still popular in the very smallest computers, such as the embedded systems that operate machinery.
1848:
351:
controller. It handles interrupt signals from the system bus. The control unit is the part of the computer that responds to the interrupts.
200:
But what if all the calculations are complete, but the CPU is still stalled, waiting for main memory? Then, a control unit can switch to an
1126:
2005:
1571:
1388:
2359:
111:
computers might do arithmetic, one or a few bits at a time. Some other computers have very complex instructions that take many steps.
1351:
875:
527:
2000:
309:
microarchitecture to use transfer-triggered multiplexers so that each instruction only utilises the exact pieces of logic needed.
2072:
506:
2354:
1825:
77:
instruction correctly. So, the bits of the instruction directly control the control unit, which in turn controls the computer.
973:
948:
841:
706:
Alsup, Mitch; Leighton, Luke; Zaruba, Florian; Thornton, James; Kimmitt, Jonathon; Petrisko, Dan; Takano, S.; Falvo, Samuel.
197:
To help this, out-of-order CPUs and control units were developed to process data as it becomes available. (See next section)
2769:
1893:
1156:
1000:
707:
2779:
1920:
1047:
807:
755:
2087:
1915:
1888:
1267:
646:
42:
to convert coded instructions into timing and control signals that direct the operation of the other units (memory,
2938:
2902:
2465:
1358:
1324:
1319:
1238:
1203:
49:
Most computer resources are managed by the CU. It directs the flow of data between the CPU and the other devices.
2877:
2774:
2175:
2082:
1883:
1104:
993:
734:
1903:
1622:
1057:
896:
Williams, R. D.; Klenke, R. H.; Aylor, J. H. (May 2003). "Teaching computer design using virtual prototyping".
554:
2077:
1925:
1898:
1759:
1373:
1334:
1191:
20:
221:
120:
2943:
2514:
2276:
1752:
1713:
1368:
1363:
1297:
1109:
586:
66:
2141:
1838:
1536:
1233:
305:
interrupt controller might continue to need a clock, but that usually uses much less power than the CPU.
201:
2791:
2438:
1855:
1346:
1314:
1084:
1072:
1052:
91:
occurs because some type of input or output needs software attention in order to operate correctly. An
80:
The control unit may include a binary counter to tell the control unit's logic what step it should do.
2882:
2845:
2835:
1223:
2897:
2304:
2240:
2217:
2067:
2029:
1865:
1815:
1810:
1287:
1181:
1089:
54:
2850:
2633:
2527:
2491:
2408:
2392:
2234:
2023:
1982:
1970:
1833:
1747:
1668:
1433:
1094:
1037:
638:
35:
406:
Animation of the control matrix of a simple hardwired control unit performing an LDA-instruction
2656:
2628:
2538:
2503:
2252:
2246:
2228:
1962:
1956:
1860:
1764:
1655:
1594:
1456:
1099:
155:
Similar calculations usually show that a pipelined computer uses less energy per instruction.
2830:
2739:
2485:
2197:
2015:
1774:
1742:
1700:
1612:
1413:
1228:
1218:
1208:
1198:
1168:
1151:
1016:
605:
496:
184:
43:
630:
190:
Results from memory can become available at unpredictable times because very fast computers
2860:
2796:
2382:
2104:
1994:
1941:
1473:
1186:
1042:
1024:
905:
391:
867:
536:
159:
usually do more instructions per second because they can do several instructions at once.
8:
2907:
2892:
2712:
2563:
2545:
2509:
2497:
2151:
2098:
1875:
1791:
1673:
1528:
1423:
1282:
501:
479:
414:
909:
2764:
2756:
2608:
2583:
2387:
2262:
1786:
1727:
1607:
1339:
1067:
631:
237:
209:
107:
is required, then a memory-not-available exception must retry the failing instruction.
92:
2717:
2684:
2600:
2532:
2433:
2423:
2413:
2344:
2339:
2334:
2257:
2186:
2092:
2052:
1685:
1635:
1585:
1561:
1443:
1383:
1378:
1260:
1176:
969:
944:
941:
Logic synthesis for FSM based control units / Alexander
Barkalov and Larysa Titarenko
921:
642:
421:
This design uses a fixed architecture—it requires changes in the wiring if the
276:
translate complex CISC x86 instructions to more RISC-like internal micro-operations.
73:
582:
535:, Moore School of Electrical Engineering, University of Pennsylvania, archived from
2887:
2820:
2806:
2661:
2568:
2522:
2329:
2324:
2319:
2314:
2309:
2299:
2169:
2136:
2047:
2042:
1951:
1803:
1798:
1781:
1769:
1708:
1272:
1250:
1136:
1114:
1032:
913:
491:
450:
177:
147:
more stages, each stage does less work, and so the stage has fewer delays from the
50:
2801:
2786:
2734:
2638:
2613:
2450:
2443:
2294:
2289:
2284:
2223:
2131:
2121:
1843:
1678:
1630:
1393:
1277:
1245:
1146:
1141:
1062:
422:
2912:
2746:
2729:
2722:
2618:
2475:
2212:
2126:
2057:
1640:
1602:
1551:
1546:
1541:
1255:
1079:
446:
104:
103:
Exceptions can be made to operate like interrupts in very simple computers. If
39:
966:
Synthesis of compositional microprogram control units for programmable devices
2932:
2707:
2623:
1663:
1645:
1438:
1131:
925:
475:
394:
for computations before stored programmed computers with CUs were invented).
19:
This article is about the component of a computer's CPU. For other uses, see
1566:
2917:
2855:
2671:
2460:
2181:
1119:
917:
355:
191:
344:
PCs use an older method, a separate I/O bus accessed by I/O instructions.
2702:
2666:
2377:
2349:
2207:
2062:
985:
471:
425:
is modified or changed. It can be convenient for simple, fast computers.
273:
38:(CPU) that directs the operation of the processor. A CU typically uses a
2588:
2578:
2573:
2555:
2455:
2428:
1690:
1523:
1493:
1213:
377:
148:
2679:
2676:
2418:
1488:
1466:
458:
440:
348:
88:
127:
logic gates of the next stage can use the bits to do the next step.
2694:
1513:
169:
1503:
1461:
87:
Many computers have two different types of unexpected events. An
1518:
1483:
1448:
297:
larger depletion regions or turning off the logic completely.
1976:
1508:
1478:
370:
453:
instructions. Microprograms were organized as a sequence of
2840:
1988:
1908:
1498:
363:
293:
123:. This design is popular because of its economy and speed.
283:
1428:
1418:
968:. Zielona GĂłra: University of Zielona GĂłra. p. 153.
341:
269:
72:
In a computer, the control unit often steps through the
705:
895:
708:"RISC-V HW Dev, 6600-style out-of-order scoreboard"
868:"MICRO-PROGRAMMED VERSUS HARDWIRED CONTROL UNITS;"
555:"Computer Organization - Control Unit and design"
173:from the compiler about the direction of branch.
2930:
465:
384:
334:
815:. Maynard Mass.: Digital Equipment Corp. 1970.
445:The idea of microprogramming was introduced by
1001:
215:
701:
699:
697:
695:
693:
594:(2.2 ed.). Berkeley: RISC-V Foundation.
449:in 1951 as an intermediate level to execute
434:
262:
2006:Computer performance by orders of magnitude
726:
552:
525:
60:
1015:
1008:
994:
963:
666:
664:
662:
660:
658:
637:. Atlanta: Scott, Foreman and Co. p.
114:
690:
397:
53:included the control unit as part of the
938:
628:
581:
401:
655:
577:
575:
284:Control units for low-powered computers
2931:
796:(v6, r0 ed.). Cambridge: ARM Ltd.
794:The ARM(tm) Technical Reference Manual
347:A modern CPU also tends to include an
989:
786:
735:"BOOM Docs, Rocketship SOC Generator"
354:There is often a cache controller to
46:and input and output devices, etc.).
1977:Floating-point operations per second
670:
572:
529:First Draft of a Report on the EDVAC
162:
13:
633:Design of a Computer: The CDC 6600
610:(3.0B ed.). Austin: IBM. 2017
220:A control unit can be designed to
14:
2955:
757:Introduction to MAXQ Architecture
732:
588:The RISC V Instruction Set Manual
553:Astha Singh (24 September 2018).
119:Many medium-complexity computers
34:) is a component of a computer's
2903:Semiconductor device fabrication
856:from the original on 2010-12-07.
2878:History of general-purpose CPUs
1105:Nondeterministic Turing machine
957:
932:
889:
878:from the original on 2017-04-30
860:
834:
822:from the original on 2015-04-22
800:
780:ARM Technical Reference, Cortex
772:
413:are implemented through use of
202:alternative thread of execution
1058:Deterministic finite automaton
964:Wiśniewski, Remigiusz (2009).
898:IEEE Transactions on Education
760:. Dallas: Maxim Integrated Inc
748:
622:
598:
546:
519:
1:
1849:Simultaneous and heterogenous
512:
466:Combination methods of design
385:Functions of the control unit
335:Integrating with the Computer
65:The simplest computers use a
21:Control unit (disambiguation)
16:Component of a computer's CPU
2533:Integrated memory controller
2515:Translation lookaside buffer
1714:Memory dependence prediction
1157:Random-access stored program
1110:Probabilistic Turing machine
939:Barkalov, Alexander (2009).
208:When a control unit permits
168:backwards branch path. If a
67:multicycle microarchitecture
7:
1989:Synaptic updates per second
485:
10:
2960:
2393:Heterogeneous architecture
1315:Orthogonal instruction set
1085:Alternating Turing machine
1073:Quantum cellular automaton
526:von Neumann, John (1945),
438:
292:Most modern computers use
216:Out of order control units
183:Some control units can do
18:
2883:Microprocessor chronology
2870:
2846:Dynamic frequency scaling
2819:
2755:
2693:
2647:
2599:
2554:
2474:
2401:
2370:
2275:
2196:
2160:
2114:
2014:
2001:Cache performance metrics
1940:
1874:
1824:
1735:
1726:
1699:
1654:
1621:
1593:
1584:
1404:
1307:
1296:
1167:
1023:
809:PDP-8L Maintenance Manual
435:Microprogram control unit
263:Translating control units
2898:Hardware security module
2241:Digital signal processor
2218:Graphics processing unit
2030:Graphics processing unit
61:Multicycle control units
55:von Neumann architecture
2939:Central processing unit
2851:Dynamic voltage scaling
2634:Memory address register
2528:Branch target predictor
2492:Address generation unit
2235:Physics processing unit
2024:Central processing unit
1983:Transactions per second
1971:Instructions per second
1894:Array processing (SIMT)
1038:Stored-program computer
782:(v8 ed.). ARM Ltd.
629:Thornton, J.E. (1970).
411:Hardwired control units
115:Pipelined control units
36:central processing unit
2657:Hardwired control unit
2539:Memory management unit
2504:Memory management unit
2253:Secure cryptoprocessor
2247:Tensor Processing Unit
2229:Vision processing unit
1963:Cycles per instruction
1957:Instructions per cycle
1904:Associative processing
1595:Instruction pipelining
1017:Processor technologies
918:10.1109/te.2002.808278
673:"Libre RISC-V M-Class"
507:Controller (computing)
407:
398:Hardwired control unit
176:Some control units do
2740:Sum-addressed decoder
2486:Arithmetic logic unit
1613:Classic RISC pipeline
1567:Epiphany architecture
1414:Motorola 68000 series
872:www.cs.binghamton.edu
497:Computer architecture
405:
390:done when using only
366:in read-only memory.
272:Intel CPUs since the
185:speculative execution
121:pipeline instructions
44:arithmetic logic unit
2861:Performance per watt
2439:replacement policies
2105:Package on a package
1995:Performance per watt
1899:Pipelined processing
1669:Tomasulo's algorithm
1474:Clipper architecture
1330:Application-specific
1043:Finite-state machine
943:. Berlin: Springer.
843:Alto Hardware Manual
2944:Digital electronics
2893:Digital electronics
2546:Instruction decoder
2498:Floating-point unit
2152:Soft microprocessor
2099:System in a package
1674:Reservation station
1204:Transport-triggered
910:2003ITEdu..46..296W
714:. RISC-V Foundation
502:Richards controller
480:Richards controller
415:combinational logic
143:not cause a stall.
2765:Integrated circuit
2609:Processor register
2263:Baseband processor
1608:Operand forwarding
1068:Cellular automaton
408:
364:bootstrap programs
238:Tomasulo algorithm
222:finish what it can
2926:
2925:
2815:
2814:
2434:Instruction cache
2424:Scratchpad memory
2271:
2270:
2258:Network processor
2187:Network on a chip
2142:Ultra-low-voltage
2093:Multi-chip module
1936:
1935:
1722:
1721:
1709:Branch prediction
1686:Register renaming
1580:
1579:
1562:VISC architecture
1384:Quantum computing
1379:VISC architecture
1261:Secondary storage
1177:Microarchitecture
1137:Register machines
975:978-83-7481-293-1
950:978-3-642-04308-6
542:on March 14, 2013
455:microinstructions
178:branch prediction
163:Preventing stalls
74:instruction cycle
2951:
2888:Processor design
2780:Power management
2662:Instruction unit
2523:Branch predictor
2472:
2471:
2170:System on a chip
2112:
2111:
1952:Transistor count
1876:Flynn's taxonomy
1733:
1732:
1591:
1590:
1394:Addressing modes
1305:
1304:
1251:Memory hierarchy
1115:Hypercomputation
1033:Abstract machine
1010:
1003:
996:
987:
986:
980:
979:
961:
955:
954:
936:
930:
929:
893:
887:
886:
884:
883:
864:
858:
857:
855:
848:
838:
832:
831:
829:
827:
821:
814:
804:
798:
797:
790:
784:
783:
776:
770:
769:
767:
765:
752:
746:
745:
743:
741:
730:
724:
723:
721:
719:
703:
688:
687:
685:
683:
671:Leighton, Luke.
668:
653:
652:
636:
626:
620:
619:
617:
615:
602:
596:
595:
593:
579:
570:
569:
567:
565:
550:
544:
543:
541:
534:
523:
492:Processor design
451:computer program
51:John von Neumann
2959:
2958:
2954:
2953:
2952:
2950:
2949:
2948:
2929:
2928:
2927:
2922:
2908:Tick–tock model
2866:
2822:
2811:
2751:
2735:Address decoder
2689:
2643:
2639:Program counter
2614:Status register
2595:
2550:
2510:Load–store unit
2477:
2470:
2397:
2366:
2267:
2224:Image processor
2199:
2192:
2162:
2156:
2132:Microcontroller
2122:Embedded system
2110:
2010:
1943:
1932:
1870:
1820:
1718:
1695:
1679:Re-order buffer
1650:
1631:Data dependency
1617:
1576:
1406:
1400:
1299:
1298:Instruction set
1292:
1278:Multiprocessing
1246:Cache hierarchy
1239:Register/memory
1163:
1063:Queue automaton
1019:
1014:
984:
983:
976:
962:
958:
951:
937:
933:
894:
890:
881:
879:
866:
865:
861:
853:
849:. Xerox. 1976.
846:
840:
839:
835:
825:
823:
819:
812:
806:
805:
801:
792:
791:
787:
778:
777:
773:
763:
761:
754:
753:
749:
739:
737:
731:
727:
717:
715:
704:
691:
681:
679:
669:
656:
649:
627:
623:
613:
611:
604:
603:
599:
591:
583:Asanovic, Krste
580:
573:
563:
561:
551:
547:
539:
532:
524:
520:
515:
488:
468:
443:
437:
423:instruction set
400:
387:
337:
286:
265:
218:
165:
117:
63:
24:
17:
12:
11:
5:
2957:
2947:
2946:
2941:
2924:
2923:
2921:
2920:
2915:
2913:Pin grid array
2910:
2905:
2900:
2895:
2890:
2885:
2880:
2874:
2872:
2868:
2867:
2865:
2864:
2858:
2853:
2848:
2843:
2838:
2833:
2827:
2825:
2817:
2816:
2813:
2812:
2810:
2809:
2804:
2799:
2794:
2789:
2784:
2783:
2782:
2777:
2772:
2761:
2759:
2753:
2752:
2750:
2749:
2747:Barrel shifter
2744:
2743:
2742:
2737:
2730:Binary decoder
2727:
2726:
2725:
2715:
2710:
2705:
2699:
2697:
2691:
2690:
2688:
2687:
2682:
2674:
2669:
2664:
2659:
2653:
2651:
2645:
2644:
2642:
2641:
2636:
2631:
2626:
2621:
2619:Stack register
2616:
2611:
2605:
2603:
2597:
2596:
2594:
2593:
2592:
2591:
2586:
2576:
2571:
2566:
2560:
2558:
2552:
2551:
2549:
2548:
2543:
2542:
2541:
2530:
2525:
2520:
2519:
2518:
2512:
2501:
2495:
2489:
2482:
2480:
2469:
2468:
2463:
2458:
2453:
2448:
2447:
2446:
2441:
2436:
2431:
2426:
2421:
2411:
2405:
2403:
2399:
2398:
2396:
2395:
2390:
2385:
2380:
2374:
2372:
2368:
2367:
2365:
2364:
2363:
2362:
2352:
2347:
2342:
2337:
2332:
2327:
2322:
2317:
2312:
2307:
2302:
2297:
2292:
2287:
2281:
2279:
2273:
2272:
2269:
2268:
2266:
2265:
2260:
2255:
2250:
2244:
2238:
2232:
2226:
2221:
2215:
2213:AI accelerator
2210:
2204:
2202:
2194:
2193:
2191:
2190:
2184:
2179:
2176:Multiprocessor
2173:
2166:
2164:
2158:
2157:
2155:
2154:
2149:
2144:
2139:
2134:
2129:
2127:Microprocessor
2124:
2118:
2116:
2115:By application
2109:
2108:
2102:
2096:
2090:
2085:
2080:
2075:
2070:
2065:
2060:
2058:Tile processor
2055:
2050:
2045:
2040:
2039:
2038:
2027:
2020:
2018:
2012:
2011:
2009:
2008:
2003:
1998:
1992:
1986:
1980:
1974:
1968:
1967:
1966:
1954:
1948:
1946:
1938:
1937:
1934:
1933:
1931:
1930:
1929:
1928:
1918:
1913:
1912:
1911:
1906:
1901:
1896:
1886:
1880:
1878:
1872:
1871:
1869:
1868:
1863:
1858:
1853:
1852:
1851:
1846:
1844:Hyperthreading
1836:
1830:
1828:
1826:Multithreading
1822:
1821:
1819:
1818:
1813:
1808:
1807:
1806:
1796:
1795:
1794:
1789:
1779:
1778:
1777:
1772:
1762:
1757:
1756:
1755:
1750:
1739:
1737:
1730:
1724:
1723:
1720:
1719:
1717:
1716:
1711:
1705:
1703:
1697:
1696:
1694:
1693:
1688:
1683:
1682:
1681:
1676:
1666:
1660:
1658:
1652:
1651:
1649:
1648:
1643:
1638:
1633:
1627:
1625:
1619:
1618:
1616:
1615:
1610:
1605:
1603:Pipeline stall
1599:
1597:
1588:
1582:
1581:
1578:
1577:
1575:
1574:
1569:
1564:
1559:
1556:
1555:
1554:
1552:z/Architecture
1549:
1544:
1539:
1531:
1526:
1521:
1516:
1511:
1506:
1501:
1496:
1491:
1486:
1481:
1476:
1471:
1470:
1469:
1464:
1459:
1451:
1446:
1441:
1436:
1431:
1426:
1421:
1416:
1410:
1408:
1402:
1401:
1399:
1398:
1397:
1396:
1386:
1381:
1376:
1371:
1366:
1361:
1356:
1355:
1354:
1344:
1343:
1342:
1332:
1327:
1322:
1317:
1311:
1309:
1302:
1294:
1293:
1291:
1290:
1285:
1280:
1275:
1270:
1265:
1264:
1263:
1258:
1256:Virtual memory
1248:
1243:
1242:
1241:
1236:
1231:
1226:
1216:
1211:
1206:
1201:
1196:
1195:
1194:
1184:
1179:
1173:
1171:
1165:
1164:
1162:
1161:
1160:
1159:
1154:
1149:
1144:
1134:
1129:
1124:
1123:
1122:
1117:
1112:
1107:
1102:
1097:
1092:
1087:
1080:Turing machine
1077:
1076:
1075:
1070:
1065:
1060:
1055:
1050:
1040:
1035:
1029:
1027:
1021:
1020:
1013:
1012:
1005:
998:
990:
982:
981:
974:
956:
949:
931:
904:(2): 296–301.
888:
859:
833:
799:
785:
771:
747:
733:Celio, Chris.
725:
689:
654:
647:
621:
597:
571:
545:
517:
516:
514:
511:
510:
509:
504:
499:
494:
487:
484:
467:
464:
447:Maurice Wilkes
439:Main article:
436:
433:
399:
396:
386:
383:
336:
333:
285:
282:
264:
261:
232:instructions.
217:
214:
164:
161:
116:
113:
105:virtual memory
84:logic family.
62:
59:
40:binary decoder
15:
9:
6:
4:
3:
2:
2956:
2945:
2942:
2940:
2937:
2936:
2934:
2919:
2916:
2914:
2911:
2909:
2906:
2904:
2901:
2899:
2896:
2894:
2891:
2889:
2886:
2884:
2881:
2879:
2876:
2875:
2873:
2869:
2862:
2859:
2857:
2854:
2852:
2849:
2847:
2844:
2842:
2839:
2837:
2834:
2832:
2829:
2828:
2826:
2824:
2818:
2808:
2805:
2803:
2800:
2798:
2795:
2793:
2790:
2788:
2785:
2781:
2778:
2776:
2773:
2771:
2768:
2767:
2766:
2763:
2762:
2760:
2758:
2754:
2748:
2745:
2741:
2738:
2736:
2733:
2732:
2731:
2728:
2724:
2721:
2720:
2719:
2716:
2714:
2711:
2709:
2708:Demultiplexer
2706:
2704:
2701:
2700:
2698:
2696:
2692:
2686:
2683:
2681:
2678:
2675:
2673:
2670:
2668:
2665:
2663:
2660:
2658:
2655:
2654:
2652:
2650:
2646:
2640:
2637:
2635:
2632:
2630:
2629:Memory buffer
2627:
2625:
2624:Register file
2622:
2620:
2617:
2615:
2612:
2610:
2607:
2606:
2604:
2602:
2598:
2590:
2587:
2585:
2582:
2581:
2580:
2577:
2575:
2572:
2570:
2567:
2565:
2564:Combinational
2562:
2561:
2559:
2557:
2553:
2547:
2544:
2540:
2537:
2536:
2534:
2531:
2529:
2526:
2524:
2521:
2516:
2513:
2511:
2508:
2507:
2505:
2502:
2499:
2496:
2493:
2490:
2487:
2484:
2483:
2481:
2479:
2473:
2467:
2464:
2462:
2459:
2457:
2454:
2452:
2449:
2445:
2442:
2440:
2437:
2435:
2432:
2430:
2427:
2425:
2422:
2420:
2417:
2416:
2415:
2412:
2410:
2407:
2406:
2404:
2400:
2394:
2391:
2389:
2386:
2384:
2381:
2379:
2376:
2375:
2373:
2369:
2361:
2358:
2357:
2356:
2353:
2351:
2348:
2346:
2343:
2341:
2338:
2336:
2333:
2331:
2328:
2326:
2323:
2321:
2318:
2316:
2313:
2311:
2308:
2306:
2303:
2301:
2298:
2296:
2293:
2291:
2288:
2286:
2283:
2282:
2280:
2278:
2274:
2264:
2261:
2259:
2256:
2254:
2251:
2248:
2245:
2242:
2239:
2236:
2233:
2230:
2227:
2225:
2222:
2219:
2216:
2214:
2211:
2209:
2206:
2205:
2203:
2201:
2195:
2188:
2185:
2183:
2180:
2177:
2174:
2171:
2168:
2167:
2165:
2159:
2153:
2150:
2148:
2145:
2143:
2140:
2138:
2135:
2133:
2130:
2128:
2125:
2123:
2120:
2119:
2117:
2113:
2106:
2103:
2100:
2097:
2094:
2091:
2089:
2086:
2084:
2081:
2079:
2076:
2074:
2071:
2069:
2066:
2064:
2061:
2059:
2056:
2054:
2051:
2049:
2046:
2044:
2041:
2037:
2034:
2033:
2031:
2028:
2025:
2022:
2021:
2019:
2017:
2013:
2007:
2004:
2002:
1999:
1996:
1993:
1990:
1987:
1984:
1981:
1978:
1975:
1972:
1969:
1964:
1961:
1960:
1958:
1955:
1953:
1950:
1949:
1947:
1945:
1939:
1927:
1924:
1923:
1922:
1919:
1917:
1914:
1910:
1907:
1905:
1902:
1900:
1897:
1895:
1892:
1891:
1890:
1887:
1885:
1882:
1881:
1879:
1877:
1873:
1867:
1864:
1862:
1859:
1857:
1854:
1850:
1847:
1845:
1842:
1841:
1840:
1837:
1835:
1832:
1831:
1829:
1827:
1823:
1817:
1814:
1812:
1809:
1805:
1802:
1801:
1800:
1797:
1793:
1790:
1788:
1785:
1784:
1783:
1780:
1776:
1773:
1771:
1768:
1767:
1766:
1763:
1761:
1758:
1754:
1751:
1749:
1746:
1745:
1744:
1741:
1740:
1738:
1734:
1731:
1729:
1725:
1715:
1712:
1710:
1707:
1706:
1704:
1702:
1698:
1692:
1689:
1687:
1684:
1680:
1677:
1675:
1672:
1671:
1670:
1667:
1665:
1664:Scoreboarding
1662:
1661:
1659:
1657:
1653:
1647:
1646:False sharing
1644:
1642:
1639:
1637:
1634:
1632:
1629:
1628:
1626:
1624:
1620:
1614:
1611:
1609:
1606:
1604:
1601:
1600:
1598:
1596:
1592:
1589:
1587:
1583:
1573:
1570:
1568:
1565:
1563:
1560:
1557:
1553:
1550:
1548:
1545:
1543:
1540:
1538:
1535:
1534:
1532:
1530:
1527:
1525:
1522:
1520:
1517:
1515:
1512:
1510:
1507:
1505:
1502:
1500:
1497:
1495:
1492:
1490:
1487:
1485:
1482:
1480:
1477:
1475:
1472:
1468:
1465:
1463:
1460:
1458:
1455:
1454:
1452:
1450:
1447:
1445:
1442:
1440:
1439:Stanford MIPS
1437:
1435:
1432:
1430:
1427:
1425:
1422:
1420:
1417:
1415:
1412:
1411:
1409:
1403:
1395:
1392:
1391:
1390:
1387:
1385:
1382:
1380:
1377:
1375:
1372:
1370:
1367:
1365:
1362:
1360:
1357:
1353:
1350:
1349:
1348:
1345:
1341:
1338:
1337:
1336:
1333:
1331:
1328:
1326:
1323:
1321:
1318:
1316:
1313:
1312:
1310:
1306:
1303:
1301:
1300:architectures
1295:
1289:
1286:
1284:
1281:
1279:
1276:
1274:
1271:
1269:
1268:Heterogeneous
1266:
1262:
1259:
1257:
1254:
1253:
1252:
1249:
1247:
1244:
1240:
1237:
1235:
1232:
1230:
1227:
1225:
1222:
1221:
1220:
1219:Memory access
1217:
1215:
1212:
1210:
1207:
1205:
1202:
1200:
1197:
1193:
1190:
1189:
1188:
1185:
1183:
1180:
1178:
1175:
1174:
1172:
1170:
1166:
1158:
1155:
1153:
1152:Random-access
1150:
1148:
1145:
1143:
1140:
1139:
1138:
1135:
1133:
1132:Stack machine
1130:
1128:
1125:
1121:
1118:
1116:
1113:
1111:
1108:
1106:
1103:
1101:
1098:
1096:
1093:
1091:
1088:
1086:
1083:
1082:
1081:
1078:
1074:
1071:
1069:
1066:
1064:
1061:
1059:
1056:
1054:
1051:
1049:
1048:with datapath
1046:
1045:
1044:
1041:
1039:
1036:
1034:
1031:
1030:
1028:
1026:
1022:
1018:
1011:
1006:
1004:
999:
997:
992:
991:
988:
977:
971:
967:
960:
952:
946:
942:
935:
927:
923:
919:
915:
911:
907:
903:
899:
892:
877:
873:
869:
863:
852:
845:
844:
837:
818:
811:
810:
803:
795:
789:
781:
775:
759:
758:
751:
736:
729:
713:
712:Google Groups
709:
702:
700:
698:
696:
694:
678:
674:
667:
665:
663:
661:
659:
650:
648:9780673059536
644:
640:
635:
634:
625:
609:
608:
607:Power ISA(tm)
601:
590:
589:
584:
578:
576:
560:
559:GeeksforGeeks
556:
549:
538:
531:
530:
522:
518:
508:
505:
503:
500:
498:
495:
493:
490:
489:
483:
481:
477:
476:Mealy machine
473:
463:
460:
456:
452:
448:
442:
432:
429:
426:
424:
419:
416:
412:
404:
395:
393:
382:
379:
374:
372:
367:
365:
359:
357:
352:
350:
345:
343:
332:
329:
325:
322:
318:
314:
310:
306:
302:
298:
295:
290:
281:
277:
275:
271:
260:
256:
252:
249:
245:
241:
239:
233:
229:
225:
223:
213:
211:
206:
203:
198:
195:
193:
188:
186:
181:
179:
174:
171:
160:
156:
152:
150:
144:
140:
136:
132:
128:
124:
122:
112:
108:
106:
101:
97:
94:
90:
85:
81:
78:
75:
70:
68:
58:
56:
52:
47:
45:
41:
37:
33:
29:
22:
2918:Chip carrier
2856:Clock gating
2775:Mixed-signal
2672:Write buffer
2649:Control unit
2648:
2461:Clock signal
2200:accelerators
2182:Cypress PSoC
1839:Simultaneous
1656:Out-of-order
1288:Neuromorphic
1169:Architecture
1127:Belt machine
1120:Zeno machine
1053:Hierarchical
965:
959:
940:
934:
901:
897:
891:
880:. Retrieved
871:
862:
842:
836:
824:. Retrieved
808:
802:
793:
788:
779:
774:
762:. Retrieved
756:
750:
738:. Retrieved
728:
716:. Retrieved
711:
680:. Retrieved
677:Crowd Supply
676:
632:
624:
612:. Retrieved
606:
600:
587:
562:. Retrieved
558:
548:
537:the original
528:
521:
469:
454:
444:
430:
427:
420:
410:
409:
388:
375:
368:
360:
356:cache memory
353:
346:
338:
330:
326:
323:
319:
315:
311:
307:
303:
299:
291:
287:
278:
266:
257:
253:
250:
246:
242:
234:
230:
226:
219:
207:
199:
196:
192:cache memory
189:
182:
175:
166:
157:
153:
145:
141:
137:
133:
129:
125:
118:
109:
102:
98:
86:
82:
79:
71:
64:
48:
31:
28:control unit
27:
25:
2703:Multiplexer
2667:Data buffer
2378:Single-core
2350:bit slicing
2208:Coprocessor
2063:Coprocessor
1944:performance
1866:Cooperative
1856:Speculative
1816:Distributed
1775:Superscalar
1760:Instruction
1728:Parallelism
1701:Speculative
1533:System/3x0
1405:Instruction
1182:Von Neumann
1095:Post–Turing
826:26 December
764:26 December
614:26 December
472:truth table
392:punch cards
274:Pentium Pro
149:logic gates
2933:Categories
2823:management
2718:Multiplier
2579:Logic gate
2569:Sequential
2476:Functional
2456:Clock rate
2429:Data cache
2402:Components
2383:Multi-core
2371:Core count
1861:Preemptive
1765:Pipelining
1748:Bit-serial
1691:Wide-issue
1636:Structural
1558:Tilera ISA
1524:MicroBlaze
1494:ETRAX CRIS
1389:Comparison
1234:Load–store
1214:Endianness
882:2017-02-17
740:16 January
718:16 January
682:16 January
513:References
378:Xerox Alto
2757:Circuitry
2677:Microcode
2601:Registers
2444:coherence
2419:CPU cache
2277:Word size
1942:Processor
1586:Execution
1489:DEC Alpha
1467:Power ISA
1283:Cognitive
1090:Universal
926:0018-9359
459:flowchart
441:Microcode
349:interrupt
93:exception
89:interrupt
2695:Datapath
2388:Manycore
2360:variable
2198:Hardware
1834:Temporal
1514:OpenRISC
1209:Cellular
1199:Dataflow
1192:modified
876:Archived
851:Archived
817:Archived
585:(2017).
486:See also
170:compiler
2871:Related
2802:Quantum
2792:Digital
2787:Boolean
2685:Counter
2584:Quantum
2345:512-bit
2340:256-bit
2335:128-bit
2178:(MPSoC)
2163:on chip
2161:Systems
1979:(FLOPS)
1792:Process
1641:Control
1623:Hazards
1509:Itanium
1504:Unicore
1462:PowerPC
1187:Harvard
1147:Pointer
1142:Counter
1100:Quantum
906:Bibcode
210:threads
2807:Switch
2797:Analog
2535:(IMC)
2506:(MMU)
2355:others
2330:64-bit
2325:48-bit
2320:32-bit
2315:24-bit
2310:16-bit
2305:15-bit
2300:12-bit
2137:Mobile
2053:Stream
2048:Barrel
2043:Vector
2032:(GPU)
1991:(SUPS)
1959:(IPC)
1811:Memory
1804:Vector
1787:Thread
1770:Scalar
1572:Others
1519:RISC-V
1484:SuperH
1453:Power
1449:MIPS-X
1424:PDP-11
1273:Fabric
1025:Models
972:
947:
924:
645:
564:25 May
2863:(PPW)
2821:Power
2713:Adder
2589:Array
2556:Logic
2517:(TLB)
2500:(FPU)
2494:(AGU)
2488:(ALU)
2478:units
2414:Cache
2295:8-bit
2290:4-bit
2285:1-bit
2249:(TPU)
2243:(DSP)
2237:(PPU)
2231:(VPU)
2220:(GPU)
2189:(NoC)
2172:(SoC)
2107:(PoP)
2101:(SiP)
2095:(MCM)
2036:GPGPU
2026:(CPU)
2016:Types
1997:(PPW)
1985:(TPS)
1973:(IPS)
1965:(CPI)
1736:Level
1547:S/390
1542:S/370
1537:S/360
1479:SPARC
1457:POWER
1340:TRIPS
1308:Types
854:(PDF)
847:(PDF)
820:(PDF)
813:(PDF)
592:(PDF)
540:(PDF)
533:(PDF)
371:PDP-8
369:Most
2841:ACPI
2574:Glue
2466:FIFO
2409:Core
2147:ASIP
2088:CPLD
2083:FPOA
2078:FPGA
2073:ASIC
1926:SPMD
1921:MIMD
1916:MISD
1909:SWAR
1889:SIMD
1884:SISD
1799:Data
1782:Task
1753:Word
1499:M32R
1444:MIPS
1407:sets
1374:ZISC
1369:NISC
1364:OISC
1359:MISC
1352:EPIC
1347:VLIW
1335:EDGE
1325:RISC
1320:CISC
1229:HUMA
1224:NUMA
970:ISBN
945:ISBN
922:ISSN
828:2019
766:2019
742:2020
720:2020
684:2020
643:ISBN
616:2019
566:2019
376:The
294:CMOS
26:The
2836:APM
2831:PMU
2723:CPU
2680:ROM
2451:Bus
2068:PAL
1743:Bit
1529:LMC
1434:ARM
1429:x86
1419:VAX
914:doi
639:125
478:or
342:x86
270:x86
2935::
2770:3D
920:.
912:.
902:46
900:.
874:.
870:.
710:.
692:^
675:.
657:^
641:.
574:^
557:.
482:.
151:.
32:CU
1009:e
1002:t
995:v
978:.
953:.
928:.
916::
908::
885:.
830:.
768:.
744:.
722:.
686:.
651:.
618:.
568:.
30:(
23:.
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.