558:
unified cache, hence any two working dual-core dies can be used, as opposed to producing four cores on a single die and requiring all four to work to produce a quad-core CPU. From an architectural point of view, ultimately, single CPU designs may make better use of the silicon surface area than multiprocessing cores, so a development commitment to this architecture may carry the risk of obsolescence. Finally, raw processing power is not the only constraint on system performance. Two processing cores sharing the same system bus and memory bandwidth limits the real-world performance advantage.
884:
5996:
58:
38:
1008:
46:
675:). Consequently, such code is much more difficult to debug than single-threaded code when it breaks. There has been a perceived lack of motivation for writing consumer-level threaded applications because of the relative rarity of consumer-level demand for maximum use of computer hardware. Also, serial tasks like decoding the
778:
a coordination language and program building blocks (programming libraries or higher-order functions). Each block can have a different native implementation for each processor type. Users simply program using these abstractions and an intelligent compiler chooses the best implementation based on the context.
777:
are written in these native languages and take advantage of multi-core processing. Balancing the application workload across processors can be problematic, especially if they have different performance characteristics. There are different conceptual models to deal with the problem, for example using
690:
The telecommunications market had been one of the first that needed a new design of parallel datapath packet processing because there was a very quick adoption of these multiple-core processors for the datapath and the control plane. These MPUs are going to replace the traditional
Network Processors
541:
Multi-core chips also allow higher performance at lower energy. This can be a big factor in mobile devices that operate on batteries. Since each core in a multi-core CPU is generally more energy-efficient, the chip becomes more efficient than having a single large monolithic core. This allows higher
493:
have turned to multi-core designs, sacrificing lower manufacturing-costs for higher performance in some applications and systems. Multi-core architectures are being developed, but so are the alternatives. An especially strong contender for established markets is the further integration of peripheral
391:
Several business motives drive the development of multi-core architectures. For decades, it was possible to improve performance of a CPU by shrinking the area of the integrated circuit (IC), which reduced the cost per device on the IC. Alternatively, for the same circuit area, more transistors could
557:
Integration of a multi-core chip can lower the chip production yields. They are also more difficult to manage thermally than lower-density single-core designs. Intel has partially countered this first problem by creating its quad-core designs by combining two dual-core ones on a single die with a
266:
In the consumer market, dual-core processors (that is, microprocessors with two units) started becoming commonplace on personal computers in the late 2000s. Quad-core processors were also being adopted in that era for higher-end systems before becoming standard. In the late 2010s, hexa-core (six
686:
Given the increasing emphasis on multi-core chip design, stemming from the grave thermal and power consumption problems posed by any further significant increase in processor clock speeds, the extent to which software can be multithreaded to take advantage of these new chips is likely to be the
403:
As the rate of clock speed improvements slowed, increased use of parallel computing in the form of multi-core processors has been pursued to improve overall processing performance. Multiple cores were used on the same CPU chip, which could then lead to better sales of CPU chips with two or more
898:
operates in an area of processor technology distinct from that of "mainstream" PCs. The same technological drives towards multi-core apply here too. Indeed, in many cases the application is a "natural" fit for multi-core technologies, if the task can easily be partitioned between the different
810:
In the third stage, development moves from the abstract toward the concrete. Developers revisit decisions made in the partitioning and communication phases with a view to obtaining an algorithm that will execute efficiently on some class of parallel computer. In particular, developers consider
801:
The tasks generated by a partition are intended to execute concurrently but cannot, in general, execute independently. The computation to be performed in one task will typically require data associated with another task. Data must then be transferred between tasks so as to allow computation to
538:(FSB). In terms of competing technologies for the available silicon die area, multi-core design can make use of proven CPU core library designs and produce a product with lower risk of design error than devising a new wider-core design. Also, adding more cache suffers from diminishing returns.
670:
thread waits for commands from the user (e.g. cancel the scan). In such cases, a multi-core architecture is of little benefit for the application itself due to the single thread doing all the heavy lifting and the inability to balance the work evenly across multiple cores. Programming truly
254:
problems may realize speedup factors near the number of cores, or even more if the problem is split up enough to fit within each core's cache(s), avoiding use of much slower main-system memory. Most applications, however, are not accelerated as much unless programmers invest effort in
533:
designs. Also, a dual-core processor uses slightly less power than two coupled single-core processors, principally because of the decreased power required to drive signals external to the chip. Furthermore, the cores share some circuitry, like the L2 cache and the interface to the
906:, legacy code or supporting independent developers less critical than is the case for PC or enterprise computing. As a result, it is easier for developers to adopt new technologies and as a result there is a greater variety of multi-core processing architectures and suppliers.
628:
How multiple cores are implemented and integrated significantly affects both the developer's programming skills and the consumer's expectations of apps and interactivity versus the device. A device advertised as being octa-core will only have independent cores if advertised as
262:
The parallelization of software is a significant ongoing topic of research. Cointegration of multiprocessor applications provides flexibility in network architecture design. Adaptability within parallel models is an additional feature of systems utilizing these protocols.
383:(TLP) methods, and multiple independent CPUs are commonly used to increase a system's overall TLP. A combination of increased available space (due to refined manufacturing processes) and the demand for increased TLP led to the development of multi-core CPUs.
934:
all manufacturing products with eight processors. For the system developer, a key challenge is how to exploit all the cores in these devices to achieve maximum networking performance at the system level, despite the performance limitations inherent in a
756:
Multi-core processing has also affected the ability of modern computational software development. Developers programming in newer languages might find that their modern languages do not support multi-core functionality. This then requires the use of
620:
The composition and balance of the cores in multi-core architecture show great variety. Some architectures use one core design repeated consistently ("homogeneous"), while others use a mixture of different cores, each optimized for a different,
687:
single greatest constraint on computer performance in the future. If developers are unable to design software to fully exploit the resources provided by multiple cores, then they will ultimately reach an insurmountable performance ceiling.
445:; the increasing gap between processor and memory speeds. This, in effect, pushes for cache sizes to be larger in order to mask the latency of memory. This helps only to the extent that memory bandwidth is not the bottleneck in performance.
811:
whether it is useful to combine, or agglomerate, tasks identified by the partitioning phase, so as to provide a smaller number of tasks, each of greater size. They also determine whether it is worthwhile to replicate data and computation.
819:
In the fourth and final stage of the design of parallel algorithms, the developers specify where each task is to execute. This mapping problem does not arise on uniprocessors or on shared-memory computers that provide automatic task
584:
core includes a high-performance core (called 'big') and a low-power core (called 'LITTLE'). There is also a trend towards improving energy-efficiency by focusing on performance-per-watt with advanced fine-grain or ultra fine-grain
792:
The partitioning stage of a design is intended to expose opportunities for parallel execution. Hence, the focus is on defining a large number of small tasks in order to yield what is termed a fine-grained decomposition of a
3196:
1936:
The research and development of multicore processors often compares many options, and benchmarks are developed to help such evaluations. Existing benchmarks include SPLASH-2, PARSEC, and COSMIC for heterogeneous systems.
2051:(SMP). In a partitioned architecture, each CPU boots into separate segments of physical memory and operate independently; in an SMP OS, processors work in a shared space, executing threads within the OS independently.
571:
The trend in processor development has been towards an ever-increasing number of cores, as processors with hundreds or even thousands of cores become theoretically possible. In addition, multi-core chips mixed with
463:; the trend of consuming exponentially increasing power (and thus also generating exponentially increasing heat) with each factorial increase of operating frequency. This increase can be mitigated by "
364:
have become a major design concern. These physical limitations can cause significant heat dissipation and data synchronization problems. Various other methods are used to improve CPU performance. Some
671:
multithreaded code often requires complex co-ordination of threads and can easily introduce subtle and difficult-to-find bugs due to the interweaving of processing on data shared between threads (see
2634:
2607:
506:
circuitry to operate at a much higher clock rate than what is possible if the signals have to travel off-chip. Combining equivalent CPUs on a single die significantly improves the performance of
554:(OS) support and to existing application software. Also, the ability of multi-core processors to increase application performance depends on the use of multiple threads within applications.
2480:
640:
Chuck Moore suggested computers should be like cellphones, using a variety of specialty cores to run modular software scheduled by a high-level applications programming interface.
2024:(DSPs) have used multi-core architectures for much longer than high-end general-purpose processors. A typical example of a DSP-specific implementation would be a combination of a
3186:
2346:
987:
Zynq UltraScale+ MPSoC has a quad-core ARM Cortex-A53 and dual-core ARM Cortex-R5. Software solutions such as OpenAMP are being used to help with inter-processor communication.
2032:. This allows for the design of products that require a general-purpose processor for user interfaces and a DSP for real-time data processing; this type of design is common in
844:
Vendors may license some software "per processor". This can give rise to ambiguity, because a "processor" may consist either of a single core or of a combination of cores.
943:
provide portable packet processing software designed so that the networking data plane runs in a fast path environment outside the operating system of the network device.
522:
less. These higher-quality signals allow more data to be sent in a given time period, since individual signals can be shorter and do not need to be repeated as often.
580:(or asymmetric) cores promise further performance and efficiency gains, especially in processing multimedia, recognition and networking applications. For example, a
2124:
109:(such as add, move data, and branch). However, the MCP can run instructions on separate cores at the same time, increasing overall speed for programs that support
434:
For general-purpose processors, much of the motivation for multi-core processors comes from greatly diminished gains in processor performance from increasing the
5322:
4349:
188:
have cores that do not share the same instruction set). Just as with single-processor systems, cores in multi-core systems may implement architectures such as
1891:
2069:
1885:
874:
counts an AMD X2 or an Intel dual-core CPU as a single processor but uses other metrics for other types, especially for processors with more than two cores.
400:
also increased by orders of magnitude in the decades of the late 20th century, from several megahertz in the 1980s to several gigahertz in the early 2000s.
2630:
963:
the four-core MSC8144 and six-core MSC8156 (and both have stated they are working on eight-core successors). Newer entries include the Storm-1 family from
3030:
3012:
2994:
2972:
2593:
3231:. 17th International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA-11). Las Vegas, Nevada, USA. pp. 551–557.
2404:
2173:
1063:
Epiphany, a many-core processor architecture which allows up to 4096 processors on-chip, although only a 16-core version has been commercially produced.
1066:
424:(SMP) designs using discrete CPUs, the issues regarding implementing multi-core processor architecture and supporting it with software are well known.
2393:
Patterson, David A. "Future of computer architecture." Berkeley EECS Annual
Research Symposium (BEARS), College of Engineering, UC Berkeley, US. 2006.
471:
poses manufacturing, system design and deployment problems that have not been justified in the face of the diminished gains in performance due to the
3070:
3321:
2439:
377:
are suitable for many applications, but are inefficient for others that contain difficult-to-predict code. Many applications are better suited to
2250:
657:, took the opposing view. He said multi-core chips need to be homogeneous collections of general-purpose cores to keep the software model simple.
2498:
2462:
320:
refers to multiple physically separate processing-units (which often contain special circuitry to facilitate communication between each other).
4460:
3643:
3118:
647:, generally agreed. He suggested the cellphone's use of many specialty cores working in concert is a good model for future multi-core designs.
2484:
608:
Chips designed from the outset for a large number of cores (rather than having evolved from single core designs) are sometimes referred to as
5412:
4162:
3094:
140:
in a single physical package. Designers may couple cores in a multi-core device tightly or loosely. For example, cores may or may not share
5264:
3440:
2917:
2773:
2747:
4319:
3885:
3702:
2869:
964:
4673:
2723:
2699:
2342:
1608:
927:
1022:
683:
are impossible to parallelize because each result generated is used to help create the next result of the entropy decoding algorithm.
3665:
2096:
3273:
2563:
John
Darlinton; Moustafa Ghanem; Yike Guo; Hing Wing To (1996). "Guided Resource Organisation in Heterogeneous Parallel Computing".
4314:
1207:
345:
4386:
1878:
738:
5393:
4668:
4139:
2893:
2372:
201:
110:
3044:
2942:
2325:
1542:
242:
algorithms used and their implementation. In particular, possible gains are limited by the fraction of the software that can
5433:
2275:
2036:. In other applications, a growing number of companies have developed multi-core DSPs with very large numbers of processors.
5660:
5083:
4207:
3470:
3314:
3158:
1018:
5683:
5093:
4234:
3248:. The 20th IEEE International Conference on High Performance Computer Architecture (HPCA-14) workshop. Orlando, FL, USA.
828:, multi-core processors are ideal because they allow many users to connect to a site simultaneously and have independent
6026:
5572:
3361:
3143:
2545:
1571:(Polaris), a 3.16 GHz, 80-core processor prototype, which the company originally stated would be released by 2011.
1100:
485:
In order to continue delivering regular performance improvements for general-purpose processors, manufacturers such as
5428:
2845:
2507:
Galaxy smartphones run on either Octa-core (2.3GHz Quad + 1.6GHz Quad) or Quad-core (2.15GHz + 1.6GHz Dual) processors
5678:
5655:
4401:
4229:
4202:
3581:
2025:
785:
acquires a central role in developing parallel applications. The basic steps in designing parallel applications are:
393:
333:
are sometimes used to describe multi-core architectures with an especially high number of cores (tens to thousands).
3293:
2300:
1611:, 256-core processor, released 2012 (256 usable VLIW cores, Network-on-Chip (NoC), 32/64-bit IEEE 754 compliant FPU)
5257:
5216:
4779:
3672:
3638:
3633:
3552:
3517:
1993:
550:
Maximizing the usage of the computing resources provided by multi-core processors requires adjustments both to the
519:
5650:
5465:
5191:
5088:
4489:
4396:
4197:
3418:
3307:
2821:
2677:
2073:
1874:
1111:
542:
performance with less energy. A challenge in this, however, is the additional overhead of writing parallel code.
185:
2797:
6021:
5757:
5671:
5620:
4217:
3936:
3371:
3026:
3008:
2986:
2968:
2315:
1988:
1726:
core and eight specialized SPUs (Synergistic
Processing Unit) optimized for vector operations used in the Sony
983:, where a system uses more than one kind of processor or cores, multi-core solutions are becoming more common:
770:
2408:
1577:
dual-, quad-, 6-, 8-, 10-, 12-, 14-, 15-, 16-, 18-, 20-, 22-, 24-, 26-, 28-, 32-, 48-, and 56-core processors.
636:
The article "CPU designers debate multi-core future" by Rick
Merritt, EE Times 2008, includes these comments:
5981:
5815:
5666:
5353:
4391:
4239:
4212:
4073:
3687:
3648:
3505:
2177:
746:
742:
730:
453:
366:
341:
295:(SoC). The terms are generally used only to refer to multi-core microprocessors that are manufactured on the
6031:
4828:
4590:
4066:
4027:
3682:
3677:
3611:
3423:
181:
1172:, dual-, quad-, 6-, 8-, 12-, 16-, 24-, 32-, and 64-core desktop, mobile, and embedded platform processors.
6041:
6000:
5946:
5406:
5250:
4455:
4152:
3850:
3547:
1908:
1545:, a family of dual-, quad-, 6-, 8-, 10-, 12-, 14-, 16-, and 18-core processors, and the successor of the
1400:
758:
573:
267:
cores) started entering the mainstream and since the early 2020s has overtaken quad-core in many spaces.
41:
Diagram of a generic dual-core processor with CPU-local level-1 caches and a shared, on-die level-2 cache
17:
2431:
902:
In addition, embedded software is typically developed for a specific hardware release, making issues of
5925:
5720:
5605:
5567:
5417:
5307:
5105:
4752:
4169:
3660:
3628:
3398:
3386:
3366:
1450:
1426:
750:
706:
666:
An outdated version of an anti-virus application may create a new thread for a scan process, while its
189:
2125:"NVIDIA Announces the GeForce RTX 30 Series: Ampere For Gaming, Starting With RTX 3080 & RTX 3090"
633:, or similar styling, as opposed to being merely two sets of quad-cores each with fixed clock speeds.
431:
Using a proven processing-core design without architectural changes reduces design risk significantly.
303:; separate microprocessor dies in the same package are generally referred to by another name, such as
5941:
5920:
5865:
5752:
5742:
5715:
5577:
5196:
5159:
5149:
3537:
3282:
2502:
2048:
2004:
952:
936:
829:
734:
726:
594:
530:
421:
220:
3267:
722:
525:
Assuming that the die can physically fit into the package, multi-core CPU designs require much less
238:
The improvement in performance gained by the use of a multi-core processor depends very much on the
6036:
5895:
5521:
5460:
5373:
5211:
4618:
4554:
4531:
4381:
4343:
4179:
4129:
4124:
3601:
3495:
3403:
2656:"6WINDGATE Software: Network Optimization Software – SDN Software – Control Plane Software | 6WIND"
2577:
2021:
1898:
1833:
1802:
1444:
1438:
1432:
1314:
1292:
762:
733:
can be used on multi-core platforms. Intel introduced a new abstraction for C++ parallelism called
667:
379:
356:
While manufacturing technology improves, reducing the size of individual gates, physical limits of
300:
288:
224:
118:
2520:
5956:
5951:
5810:
5401:
5164:
4947:
4841:
4805:
4722:
4706:
4548:
4337:
4296:
4284:
4147:
4061:
3982:
3747:
3408:
3351:
1568:
1307:
1090:
1034:
980:
919:
622:
590:
577:
284:
251:
173:
86:
31:
1918:
5695:
5627:
5531:
5423:
5378:
4970:
4942:
4852:
4817:
4566:
4560:
4542:
4276:
4270:
4174:
4078:
3969:
3908:
3770:
3413:
2751:
2572:
1494:
1106:
374:
1030:
5787:
5747:
5700:
5690:
5485:
5348:
5287:
5144:
5053:
4799:
4511:
4329:
4088:
4056:
4014:
3926:
3727:
3542:
3532:
3522:
3512:
3482:
3465:
3330:
2199:
Duran, A (2011). "Ompss: a proposal for programming heterogeneous multi-core architectures".
1977:
1972:
1719:
1500:
774:
602:
526:
193:
5727:
5615:
5610:
5600:
5587:
5383:
5174:
5110:
4418:
4308:
4255:
3787:
3500:
3356:
3338:
1670:
HAL series tightly coupled 16-256 cores, L1 shared memory, hardware synchronized processor.
1301:
1215:
903:
861:
782:
702:
2103:
1677:
KC256, a 257-core microcontroller with a PowerPC core and 256 8-bit "processing elements".
1509:, the first dual-core (and, later, quad-core) processor for the budget/entry-level market.
8:
5890:
5845:
5645:
5511:
5221:
5206:
5026:
4877:
4859:
4823:
4811:
4465:
4412:
4189:
4105:
3987:
3842:
3737:
3596:
1667:
967:
with 40 and 80 general purpose ALUs per chip, all programmable in C as a SIMD engine and
865:
337:
832:
of execution. This allows for Web servers and application servers that have much better
309:. This article uses the terms "multi-core" and "dual-core" for CPUs manufactured on the
227:(GPU). Core count goes up to even dozens, and for specialized chips over 10,000, and in
5915:
5764:
5737:
5562:
5526:
5516:
5475:
5317:
5297:
5292:
5273:
5078:
5070:
4922:
4897:
4701:
4576:
4100:
4041:
3921:
3653:
3381:
3229:
Predicting CPU Availability of a Multi-core
Processor Executing Concurrent Java Threads
1912:
1651:
1218:
processor cores, intended for high-performance embedded and entertainment applications.
915:
895:
871:
325:
243:
114:
82:
5961:
5637:
5595:
5490:
5031:
4998:
4914:
4846:
4747:
4737:
4727:
4658:
4653:
4648:
4571:
4500:
4406:
4366:
3999:
3949:
3899:
3875:
3757:
3697:
3692:
3574:
3490:
2321:
1797:
1789:
1639:
1396:
1289:
1277:
CSX700, 192-core processor, released in 2008 (32/64-bit floating point; Integer ALU).
1263:
956:
435:
305:
216:
130:
2368:
2226:
5971:
5770:
5705:
5552:
5368:
5363:
5358:
5327:
5201:
5134:
5120:
4975:
4882:
4836:
4643:
4638:
4633:
4628:
4623:
4613:
4483:
4450:
4361:
4356:
4265:
4117:
4112:
4095:
4083:
4022:
3586:
3564:
3450:
3428:
3346:
3249:
3246:
A New
Composite CPU/Memory Model for Predicting Efficiency of Multi-core Processing
3232:
2208:
2044:
1733:
1595:
1204:
883:
802:
proceed. This information flow is specified in the communication phase of a design.
676:
586:
551:
408:
has produced a 48-core processor for research in cloud computing; each core has an
361:
292:
197:
153:
3243:
3226:
2946:
1077:
5835:
5775:
5710:
5557:
5547:
5480:
5470:
5312:
5302:
5115:
5100:
5048:
4952:
4927:
4764:
4757:
4608:
4603:
4598:
4537:
4445:
4435:
4157:
3992:
3944:
3707:
3591:
3559:
3460:
3455:
3376:
3187:"Meet KiloCore, a 1,000-core processor so efficient it could run on a AA battery"
2562:
2251:"Intel taking its six-core processors mainstream in 2018 with Coffee Lake family"
1962:
1685:
1655:
1285:
923:
515:
503:
247:
212:
207:
Multi-core processors are widely used across many application domains, including
169:
165:
145:
137:
106:
102:
1249:
BCM2836, BCM2837, BCM2710 and BCM2711 quad-core ARM SoC (designed for different
1142:, single-, dual-, quad-, 6-, 8-, 12-, and 16-core server/workstation processors.
1069:
5966:
5782:
5439:
5332:
5226:
5060:
5043:
5036:
4932:
4789:
4526:
4440:
4371:
3954:
3916:
3865:
3860:
3855:
3569:
3393:
2029:
1999:
1785:
1503:, single-core, dual-core, and quad-core processors for smartphones and tablets.
1463:
1414:
1185:
991:
891:
on a plug-in card with processor, memory, power supply, and external interfaces
581:
535:
507:
177:
78:
3253:
2212:
2149:"Sunway TaihuLight - Sunway MPP, Sunway SW26010 260C 1.45GHz, Sunway | TOP500"
1580:
6015:
5855:
5732:
5021:
4937:
3977:
3959:
3752:
3445:
3287:
1947:
1762:
1756:
1750:
1744:
1727:
672:
650:
518:
between different CPUs travel shorter distances, and therefore those signals
357:
228:
149:
3880:
2603:
1178:, quad-, 8-, 12-, 16-, 24-, 32-, and 64-core server and embedded processors.
971:
with 300 processors on a single die, focused on communication applications.
235:
up to 20 million processing elements total in addition to host processors).
5455:
5231:
5169:
4985:
4962:
4774:
4495:
3433:
3210:
2655:
2598:
2047:
are able to use a dual-CPU multiprocessor: partitioned multiprocessing and
2033:
1927:, Kilocore, a 1000 core 1.78 GHz processor on a 32 nm IBM process
1469:
1390:
1250:
1227:
1211:
1121:
1117:
511:
161:
126:
62:
5976:
5016:
4980:
4691:
4663:
4521:
4376:
3299:
1565:, single, dual-core, and quad-core processors for the entry-level market.
1550:
1524:
1386:
825:
680:
371:
256:
157:
117:
techniques. Manufacturers typically integrate the cores onto a single IC
3244:
Khondker S. Hasan; John
Antonio; Sridhar Radhakrishnan (February 2014).
3236:
2549:
2148:
4902:
4892:
4887:
4869:
4769:
4742:
4004:
3837:
3807:
3527:
3045:"Intel shows off Xeon Platinum CPU with up to 56 cores and 112 threads"
1902:
1868:
1546:
1518:
1482:
1272:
1163:
1145:
853:
833:
464:
397:
50:
2669:
57:
5850:
5825:
5242:
4993:
4990:
4732:
3802:
3780:
1846:
1311:
1259:
1151:
1127:
1114:, dual-, triple-, and quad-core of Accelerated Processor Units (APU).
960:
849:
710:
692:
392:
be used in the design, which increased functionality, especially for
141:
3227:
Khondker S. Hasan; Nicolas G. Grounds; John K. Antonio (July 2011).
1485:, single, dual-core, quad-core, 8-, 12-, and 16-core processors for
37:
5900:
5880:
5805:
5008:
3827:
3191:
2594:"Windows Server 2016 moving to per core, not per socket, licensing"
2524:
1985:(Open Computing Language) – a framework for heterogeneous execution
1967:
1924:
1858:
1774:
1768:
1691:
1674:
1661:
1512:
1458:
1420:
1404:
1198:
1060:
968:
931:
769:, which perform math computations faster than newer languages like
718:
705:
techniques can benefit from multiple cores directly. Some existing
696:
609:
239:
208:
3296:, published on Feb 19, 2010 (more than one dead link in the slide)
2276:"Six-core CPUs are now more popular than quad-core chips on Steam"
231:(i.e. clusters of chips) the count can go over 10 million (and in
5905:
5885:
5860:
5495:
3817:
3775:
2343:"Intel Shows 48-core x86 Processor as Single-chip Cloud Computer"
1824:
1723:
1715:
1562:
1556:
1538:
1534:
1530:
1506:
1486:
1410:
1380:
1344:
1329:
1325:
1321:
1157:
1139:
857:
766:
644:
232:
1298:
467:" the processor by using smaller traces for the same logic. The
5875:
5870:
3832:
3797:
3762:
3119:"Intel® Xeon Phi™ 72x5 Processor Family Product Specifications"
1982:
1818:
1813:
1697:
1681:
1634:
1620:
1603:
1490:
1374:
1368:
1362:
1356:
1350:
1340:
1181:
1133:
1096:
984:
714:
654:
598:
502:
The proximity of multiple CPU cores on the same die allows the
172:. Homogeneous multi-core systems include only identical cores;
246:
simultaneously on multiple cores; this effect is described by
4290:
3822:
3792:
3095:"Intel® Xeon Phi™ x100 Product Family Product Specifications"
2990:
1957:
1664:
PC200 series 200–300 cores per device for DSP & wireless.
1643:
1477:
1267:
1221:
1189:
1169:
1086:
1083:
1073:
940:
486:
405:
2483:(Press release). MediaTek. November 20, 2013. Archived from
2463:"These 5 Myths About the Octa-Core Phones Are Actually True"
1282:
Cradle
Technologies CT3400 and CT3600, both multi-core DSPs.
5910:
5840:
5830:
5154:
4302:
4222:
3812:
2918:"Intel® Pentium® Processor D Series Product Specifications"
2774:"Intel® Celeron® Processor J Series Product Specifications"
2633:. OMT-CO Operations Management Technology Consulting GmbH.
1952:
1738:
1707:
1574:
1453:, an eight-core z/Architecture processor, released in 2021.
1447:, a twelve-core z/Architecture processor, released in 2019.
1435:, an eight-core z/Architecture processor, released in 2015.
1175:
176:
multi-core systems have cores that are not identical (e.g.
45:
1722:
processor, a nine-core processor with one general purpose
959:
has the three-core TMS320C6488 and four-core TMS320C5441,
5820:
5797:
3742:
3732:
2870:"Intel® Core™ X-series Processors Product Specifications"
2481:"MediaTek Launches MT6592 True Octa-Core Mobile Platform"
1864:
1711:
1423:, a quad-core z/Architecture processor, released in 2010.
1335:
490:
409:
129:. As of 2024, the microprocessors used in almost all new
2724:"Intel® Atom™ Processor Z Series Product Specifications"
2700:"Intel® Atom™ Processor C Series Product Specifications"
1441:, a ten-core z/Architecture processor, released in 2017.
1429:, a six-core z/Architecture processor, released in 2012.
1365:, a 4, 6 and 8-core PowerPC processor, released in 2010.
3144:"40-core processor with Forth-based IDE tools unveiled"
1403:-capable, PowerPC microprocessor used in the Microsoft
1160:, single-, dual-, and quad-core entry level processors.
1154:, dual-, triple-, quad-, and 6-core desktop processors.
1694:"SiCortex node" has six MIPS64 cores on a single chip.
1559:, single, dual-core, quad-core, and 8-core processors.
1383:, a 15 or 30-core PowerPC processor, released in 2021.
1377:, a 12 or 24-core PowerPC processor, released in 2017.
1262:
Xtensa LX6, available in a dual-core configuration in
456:
to keep a high-performance single-core processor busy.
1389:
MP, a dual-core PowerPC processor, used in the Apple
856:
licensing system. However, for some software such as
2369:"Intel unveils 48-core cloud computing silicon chip"
30:"Dual Core" redirects here. Not to be confused with
3211:"COSMIC:Statistical Multiprocessor Benchmark Suite"
1629:
XLS, an eight-core, quad-threaded MIPS64 processor.
1626:
XLR, an eight-core, quad-threaded MIPS64 processor.
1527:, 2 dual-core dies packaged in a multi-chip module.
1130:, dual-, triple-, and quad-core desktop processors.
420:Since computer manufacturers have long implemented
2894:"Intel® Itanium® Processor Product Specifications"
2405:"Q & A: Do multicores save energy? Not really"
2118:
2116:
1777:, a sixteen-core, 128-concurrent-thread processor.
1765:, a sixteen-core, 128-concurrent-thread processor.
1359:, a dual-core PowerPC processor, released in 2007.
1353:, a dual-core PowerPC processor, released in 2004.
1210:is a fully synthesizable multi-core container for
27:Microprocessor with more than one processing unit
6013:
3071:"2nd Gen Intel® Xeon® Scalable Processors Brief"
2670:"Sempron™ 3850 APU with Radeon™ R3 Series | AMD"
2301:"The 2,048-core PEZY-SC2 sets a Green500 record"
1771:, an eight-core, 64-concurrent-thread processor.
1759:, an eight-core, 64-concurrent-thread processor.
1583:57-, 60-, 61-, 64-, 68-, and 72-core processors.
1371:, a 12-core PowerPC processor, released in 2013.
848:Initially, for some of its enterprise software,
2585:
2113:
1224:ModemX, up to 128 cores, wireless applications.
918:have become mainstream, with companies such as
3294:Architecting solutions for the Manycore future
1238:Vega 3, a 54-core processor, released in 2008.
1235:Vega 2, a 48-core processor, released in 2006.
1232:Vega 1, a 24-core processor, released in 2005.
868:, Microsoft has shifted to per-core licensing.
612:designs, emphasising qualitative differences.
452:; the increasing difficulty of finding enough
93:to emphasize their multiplicity (for example,
5258:
3315:
2969:"Intel releases 15-core Xeon E7 v2 processor"
2631:"The Licensing Of Oracle Technology Products"
643:Atsushi Hasegawa, a senior chief engineer at
180:have heterogeneous cores that share the same
1594:SEAforth24, a 24-core processor designed by
946:
316:In contrast to multi-core systems, the term
313:integrated circuit, unless otherwise noted.
4320:Computer performance by orders of magnitude
2273:
1148:, dual-, triple-, and quad-core processors.
1136:, quad-, 6-, and 8-core desktop processors.
5265:
5251:
3329:
3322:
3308:
3283:"Multicore Is Bad News for Supercomputers"
2748:"Intel Preps Dual-Core Celeron Processors"
2432:"Intel: Why a 1,000-core chip is feasible"
1646:cores; plus other more specialized cores).
1166:, single- and dual-core laptop processors.
939:(SMP) operating system. Companies such as
514:) operations. Put simply, this means that
454:parallelism in a single instruction stream
2966:
2576:
1805:, a five-core multimedia video processor.
1747:and UltraSPARC IV+, dual-core processors.
1124:single- and dual-core desktop processors.
761:to access code written in languages like
653:, founder and chief executive of startup
287:(CPU), but are sometimes also applied to
152:inter-core communication methods. Common
125:(CMP), or onto multiple dies in a single
2521:"CPU designers debate multi-core future"
2460:
1310:QorIQ series processors, up to 8 cores,
1201:BF561, a symmetrical dual-core processor
1033:by adding descriptive text and removing
974:
882:
878:
438:. This is due to three primary factors:
386:
346:semiconductor intellectual property core
56:
44:
36:
3156:
2940:
2518:
2402:
2174:"What makes parallel programming hard?"
2171:
1879:Asynchronous array of simple processors
14:
6014:
5272:
3184:
2967:Kowaliski, Cyril (February 18, 2014).
2591:
2340:
2039:
576:, memory-on-chip, and special-purpose
5246:
3303:
2565:Journal of High Performance Computing
2375:from the original on December 6, 2012
2313:
2298:
2198:
2067:
1753:, an eight-core, 32-thread processor.
909:
737:. Other research efforts include the
4291:Floating-point operations per second
3157:Hammond, Lance; et al. (1999).
3142:Cole, Bernard (September 24, 2008).
3141:
2941:Zazaian, Mike (September 26, 2006).
2610:from the original on 4 December 2015
2349:from the original on January 5, 2016
2097:"Multicore Processors – A Necessity"
1808:TMS320TMS320C66, 2-, 4-, 8-core DSP.
1591:SEAforth 40C18, a 40-core processor.
1001:
997:
415:
283:most commonly refer to some sort of
3199:from the original on June 23, 2016.
3027:"Intel Xeon Processor E3 v2 Family"
3009:"Intel Xeon Processor E7 v2 Family"
2987:"Intel Xeon Processor E7 v3 Family"
2546:"Multicore Packet Processing Forum"
2094:
1788:, a 260-core processor used in the
661:
344:. Each "core" can be considered a "
156:used to interconnect cores include
24:
3220:
2519:Merritt, Rick (February 6, 2008).
2442:from the original on 6 August 2015
2070:"Definition: multi-core processor"
2068:Rouse, Margaret (March 27, 2007).
1101:massively parallel processor array
136:A multi-core processor implements
25:
6053:
3261:
2750:. 11 October 2007. Archived from
2592:Bright, Peter (4 December 2015).
2461:Kudikala, Chakri (Aug 27, 2016).
2429:
2341:Shrout, Ryan (December 2, 2009).
2122:
1468:Danube, a dual-core, MIPS-based,
394:complex instruction set computing
5995:
5994:
5217:Semiconductor device fabrication
3033:from the original on 2015-07-07.
3015:from the original on 2015-07-07.
2997:from the original on 2015-07-07.
2975:from the original on 2014-10-11.
2499:"What is an Octa-core processor"
2017:
1994:Partitioned global address space
1006:
545:
186:AMD Accelerated Processing Units
101:). Each core reads and executes
5466:Analysis of parallel algorithms
5192:History of general-purpose CPUs
3419:Nondeterministic Turing machine
3203:
3178:
3150:
3135:
3111:
3087:
3068:
3062:
3037:
3019:
3001:
2979:
2960:
2934:
2910:
2886:
2862:
2838:
2814:
2790:
2766:
2740:
2716:
2692:
2680:from the original on 4 May 2019
2662:
2648:
2637:from the original on 2014-03-21
2622:
2556:
2538:
2512:
2491:
2473:
2454:
2423:
2403:Suleman, Aater (May 19, 2011).
2396:
2387:
2361:
2334:
2307:
2292:
2172:Suleman, Aater (May 20, 2011).
1875:University of California, Davis
1304:, a 100-core DSP/GPP processor.
691:that were based on proprietary
615:
529:(PCB) space than do multi-chip
85:(IC) with two or more separate
3372:Deterministic finite automaton
3185:Chacos, Brad (June 20, 2016).
2846:"Products formerly Kentsfield"
2299:Schor, David (November 2017).
2267:
2243:
2219:
2192:
2165:
2141:
2088:
2061:
1989:Parallel random access machine
1741:5200, two-core VLIW processor.
1619:XLP, a 32-core, quad-threaded
351:
270:
250:. In the best case, so-called
13:
1:
5413:Simultaneous and heterogenous
4163:Simultaneous and heterogenous
3274:"Embedded moves to multicore"
2176:. FutureChips. Archived from
2055:
1931:
1827:, a 72-core 64-bit processor.
1821:, a 64-core 32-bit processor.
1493:, embedded applications, and
1246:SiByte SB1250, SB1255, SB1455
1054:
497:
367:instruction-level parallelism
6001:Category: Parallel computing
4847:Integrated memory controller
4829:Translation lookaside buffer
4028:Memory dependence prediction
3471:Random-access stored program
3424:Probabilistic Turing machine
2314:Vajda, András (2011-06-10).
2072:. TechTarget. Archived from
1417:processor, released in 2008.
1347:processor, released in 2001.
839:
7:
4303:Synaptic updates per second
3268:"What Is a Processor Core?"
2317:Programming Many-Core Chips
2201:Parallel Processing Letters
1940:
1909:University of Texas, Austin
1852:
1501:Atom SoC (system on a chip)
990:Mobile devices may use the
707:parallel programming models
574:simultaneous multithreading
561:
10:
6058:
5308:High-performance computing
4707:Heterogeneous architecture
3629:Orthogonal instruction set
3399:Alternating Turing machine
3387:Quantum cellular automaton
2822:"Products formerly Conroe"
2274:Alan Dexter (2022-04-05).
824:On the other hand, on the
29:
6027:Digital signal processing
5990:
5942:Automatic parallelization
5934:
5796:
5636:
5586:
5578:Application checkpointing
5540:
5504:
5448:
5392:
5341:
5280:
5197:Microprocessor chronology
5184:
5160:Dynamic frequency scaling
5133:
5069:
5007:
4961:
4913:
4868:
4788:
4715:
4684:
4589:
4510:
4474:
4428:
4328:
4315:Cache performance metrics
4254:
4188:
4138:
4049:
4040:
4013:
3968:
3935:
3907:
3898:
3718:
3621:
3610:
3481:
3337:
3278:Embedded Computing Design
3254:10.13140/RG.2.1.3051.9207
2943:"Intel: 80 Cores by 2011"
2798:"Products formerly Yonah"
2501:. Samsung. Archived from
2371:. BBC. December 3, 2009.
2227:"Definition of dual core"
2213:10.1142/S0129626411000151
2049:symmetric multiprocessing
2022:Digital signal processors
1921:, Sweden, ePUMA processor
953:digital signal processing
947:Digital signal processing
937:symmetric multiprocessing
566:
494:functions into the chip.
422:symmetric multiprocessing
348:" as well as a CPU core.
340:cores placed on a single
289:digital signal processors
221:digital signal processing
144:, and they may implement
53:E6750 dual-core processor
5212:Hardware security module
4555:Digital signal processor
4532:Graphics processing unit
4344:Graphics processing unit
2011:
1899:University of Washington
1861:, 4-core Hydra processor
1834:Software Defined Silicon
1652:Parallax Propeller P8X32
1521:, a dual-core processor.
1515:, a dual-core processor.
955:the same trend applies:
773:. Intel's MKL and AMD's
380:thread-level parallelism
105:, specifically ordinary
87:central processing units
5957:Embarrassingly parallel
5952:Deterministic algorithm
5165:Dynamic voltage scaling
4948:Memory address register
4842:Branch target predictor
4806:Address generation unit
4549:Physics processing unit
4338:Central processing unit
4297:Transactions per second
4285:Instructions per second
4208:Array processing (SIMT)
3352:Stored-program computer
2320:. Springer. p. 3.
1840:
1569:Teraflops Research Chip
1495:mobile internet devices
1308:Freescale Semiconductor
1258:Cadence Design Systems
1091:physics processing unit
1035:less pertinent examples
981:heterogeneous computing
920:Freescale Semiconductor
914:As of 2010, multi-core
852:continued to use a per-
285:central processing unit
252:embarrassingly parallel
32:Dual Core (hip hop duo)
5672:Associative processing
5628:Non-blocking algorithm
5434:Clustered multi-thread
4971:Hardwired control unit
4853:Memory management unit
4818:Memory management unit
4567:Secure cryptoprocessor
4561:Tensor Processing Unit
4543:Vision processing unit
4277:Cycles per instruction
4271:Instructions per cycle
4218:Associative processing
3909:Instruction pipelining
3331:Processor technologies
3160:The Stanford Hydra CMP
1890:167-core 1.2 GHz
1616:NetLogic Microsystems
1078:fault-tolerant version
1076:that also exists in a
965:Stream Processors, Inc
892:
659:
603:portable media players
396:(CISC) architectures.
370:(ILP) methods such as
336:Some systems use many
66:
54:
42:
6022:Computer architecture
5788:Hardware acceleration
5701:Superscalar processor
5691:Dataflow architecture
5288:Distributed computing
5054:Sum-addressed decoder
4800:Arithmetic logic unit
3927:Classic RISC pipeline
3881:Epiphany architecture
3728:Motorola 68000 series
2505:on January 17, 2022.
1973:Multicore Association
1884:36-core 610 MHz
975:Heterogeneous systems
886:
879:Embedded applications
739:Codeplay Sieve System
638:
527:printed circuit board
387:Commercial incentives
60:
48:
40:
5667:Pipelined processing
5616:Explicit parallelism
5611:Implicit parallelism
5601:Dataflow programming
5175:Performance per watt
4753:replacement policies
4419:Package on a package
4309:Performance per watt
4213:Pipelined processing
3983:Tomasulo's algorithm
3788:Clipper architecture
3644:Application-specific
3357:Finite-state machine
2487:on October 29, 2020.
2411:on December 16, 2012
1919:Linköping University
1702:U74 includes 4 cores
1216:ARM Cortex-A9 MPCore
904:software portability
703:Parallel programming
404:cores. For example,
331:massively multi-core
103:program instructions
71:multi-core processor
5891:Parallel Extensions
5696:Pipelined processor
5207:Digital electronics
4860:Instruction decoder
4812:Floating-point unit
4466:Soft microprocessor
4413:System in a package
3988:Reservation station
3518:Transport-triggered
1642:(10496 CUDA cores,
1099:Am2045, a 336-core
1031:improve the article
866:Windows Server 2016
858:BizTalk Server 2013
759:numerical libraries
679:algorithms used in
436:operating frequency
338:soft microprocessor
299:integrated circuit
123:chip multiprocessor
65:dual-core processor
6042:Parallel computing
5765:Massively parallel
5743:distributed shared
5563:Cache invalidation
5527:Instruction window
5318:Manycore processor
5298:Massively parallel
5293:Parallel computing
5274:Parallel computing
5079:Integrated circuit
4923:Processor register
4577:Baseband processor
3922:Operand forwarding
3382:Cellular automaton
2754:on 4 November 2007
1680:Raspberry Pi Ltd.
1288:Octeon, a 32-core
916:network processors
910:Network processors
896:Embedded computing
893:
872:Oracle Corporation
164:, two-dimensional
154:network topologies
131:personal computers
115:parallel computing
83:integrated circuit
67:
55:
43:
6009:
6008:
5962:Parallel slowdown
5596:Stream processing
5486:Karp–Flatt metric
5240:
5239:
5129:
5128:
4748:Instruction cache
4738:Scratchpad memory
4585:
4584:
4572:Network processor
4501:Network on a chip
4456:Ultra-low-voltage
4407:Multi-chip module
4250:
4249:
4036:
4035:
4023:Branch prediction
4000:Register renaming
3894:
3893:
3876:VISC architecture
3698:Quantum computing
3693:VISC architecture
3575:Secondary storage
3491:Microarchitecture
3451:Register machines
2327:978-1-4419-9739-5
2129:www.anandtech.com
2076:on August 5, 2010
2045:operating systems
1836:quad-core XS1-G4.
1798:Texas Instruments
1790:Sunway TaihuLight
1399:, a triple-core,
1264:Espressif Systems
1052:
1051:
998:Hardware examples
957:Texas Instruments
595:frequency scaling
416:Technical factors
306:multi-chip module
16:(Redirected from
6049:
6032:Flynn's taxonomy
5998:
5997:
5972:Software lockout
5771:Computer cluster
5706:Vector processor
5661:Array processing
5646:Flynn's taxonomy
5553:Memory coherence
5328:Computer network
5267:
5260:
5253:
5244:
5243:
5202:Processor design
5094:Power management
4976:Instruction unit
4837:Branch predictor
4786:
4785:
4484:System on a chip
4426:
4425:
4266:Transistor count
4190:Flynn's taxonomy
4047:
4046:
3905:
3904:
3708:Addressing modes
3619:
3618:
3565:Memory hierarchy
3429:Hypercomputation
3347:Abstract machine
3324:
3317:
3310:
3301:
3300:
3257:
3240:
3215:
3214:
3207:
3201:
3200:
3182:
3176:
3175:
3173:
3171:
3165:
3154:
3148:
3147:
3139:
3133:
3132:
3130:
3129:
3115:
3109:
3108:
3106:
3105:
3091:
3085:
3084:
3082:
3081:
3066:
3060:
3059:
3057:
3056:
3041:
3035:
3034:
3023:
3017:
3016:
3005:
2999:
2998:
2983:
2977:
2976:
2964:
2958:
2957:
2955:
2954:
2945:. Archived from
2938:
2932:
2931:
2929:
2928:
2914:
2908:
2907:
2905:
2904:
2890:
2884:
2883:
2881:
2880:
2866:
2860:
2859:
2857:
2856:
2842:
2836:
2835:
2833:
2832:
2818:
2812:
2811:
2809:
2808:
2794:
2788:
2787:
2785:
2784:
2770:
2764:
2763:
2761:
2759:
2744:
2738:
2737:
2735:
2734:
2720:
2714:
2713:
2711:
2710:
2696:
2690:
2689:
2687:
2685:
2666:
2660:
2659:
2652:
2646:
2645:
2643:
2642:
2626:
2620:
2619:
2617:
2615:
2589:
2583:
2582:
2580:
2560:
2554:
2553:
2548:. Archived from
2542:
2536:
2535:
2533:
2531:
2516:
2510:
2509:
2495:
2489:
2488:
2477:
2471:
2470:
2458:
2452:
2451:
2449:
2447:
2427:
2421:
2420:
2418:
2416:
2407:. Archived from
2400:
2394:
2391:
2385:
2384:
2382:
2380:
2365:
2359:
2358:
2356:
2354:
2338:
2332:
2331:
2311:
2305:
2304:
2296:
2290:
2289:
2287:
2286:
2271:
2265:
2264:
2262:
2261:
2247:
2241:
2240:
2238:
2237:
2223:
2217:
2216:
2196:
2190:
2189:
2187:
2185:
2169:
2163:
2162:
2160:
2159:
2145:
2139:
2138:
2136:
2135:
2120:
2111:
2110:
2108:
2102:. Archived from
2101:
2095:Schauer, Bryan.
2092:
2086:
2085:
2083:
2081:
2065:
2042:
2020:
1734:Sun Microsystems
1654:, an eight-core
1596:Charles H. Moore
1320:Hewlett-Packard
1302:hx3100 Processor
1067:Aeroflex Gaisler
1047:
1044:
1038:
1010:
1009:
1002:
677:entropy encoding
662:Software effects
587:power management
552:operating system
362:microelectronics
293:system on a chip
133:are multi-core.
107:CPU instructions
21:
6057:
6056:
6052:
6051:
6050:
6048:
6047:
6046:
6037:Microprocessors
6012:
6011:
6010:
6005:
5986:
5930:
5836:Coarray Fortran
5792:
5776:Beowulf cluster
5632:
5582:
5573:Synchronization
5558:Cache coherence
5548:Multiprocessing
5536:
5500:
5481:Cost efficiency
5476:Gustafson's law
5444:
5388:
5337:
5313:Multiprocessing
5303:Cloud computing
5276:
5271:
5241:
5236:
5222:Tick–tock model
5180:
5136:
5125:
5065:
5049:Address decoder
5003:
4957:
4953:Program counter
4928:Status register
4909:
4864:
4824:Load–store unit
4791:
4784:
4711:
4680:
4581:
4538:Image processor
4513:
4506:
4476:
4470:
4446:Microcontroller
4436:Embedded system
4424:
4324:
4257:
4246:
4184:
4134:
4032:
4009:
3993:Re-order buffer
3964:
3945:Data dependency
3931:
3890:
3720:
3714:
3613:
3612:Instruction set
3606:
3592:Multiprocessing
3560:Cache hierarchy
3553:Register/memory
3477:
3377:Queue automaton
3333:
3328:
3264:
3223:
3221:Further reading
3218:
3209:
3208:
3204:
3183:
3179:
3169:
3167:
3163:
3155:
3151:
3140:
3136:
3127:
3125:
3117:
3116:
3112:
3103:
3101:
3093:
3092:
3088:
3079:
3077:
3069:PDF, Download.
3067:
3063:
3054:
3052:
3043:
3042:
3038:
3025:
3024:
3020:
3007:
3006:
3002:
2985:
2984:
2980:
2965:
2961:
2952:
2950:
2939:
2935:
2926:
2924:
2916:
2915:
2911:
2902:
2900:
2892:
2891:
2887:
2878:
2876:
2868:
2867:
2863:
2854:
2852:
2844:
2843:
2839:
2830:
2828:
2820:
2819:
2815:
2806:
2804:
2796:
2795:
2791:
2782:
2780:
2772:
2771:
2767:
2757:
2755:
2746:
2745:
2741:
2732:
2730:
2722:
2721:
2717:
2708:
2706:
2698:
2697:
2693:
2683:
2681:
2668:
2667:
2663:
2654:
2653:
2649:
2640:
2638:
2629:
2627:
2623:
2613:
2611:
2590:
2586:
2561:
2557:
2544:
2543:
2539:
2529:
2527:
2517:
2513:
2497:
2496:
2492:
2479:
2478:
2474:
2459:
2455:
2445:
2443:
2428:
2424:
2414:
2412:
2401:
2397:
2392:
2388:
2378:
2376:
2367:
2366:
2362:
2352:
2350:
2339:
2335:
2328:
2312:
2308:
2297:
2293:
2284:
2282:
2272:
2268:
2259:
2257:
2249:
2248:
2244:
2235:
2233:
2225:
2224:
2220:
2197:
2193:
2183:
2181:
2180:on May 29, 2011
2170:
2166:
2157:
2155:
2147:
2146:
2142:
2133:
2131:
2121:
2114:
2106:
2099:
2093:
2089:
2079:
2077:
2066:
2062:
2058:
2014:
2009:
1963:Hyper-threading
1943:
1934:
1855:
1843:
1688:microcontroller
1656:microcontroller
1286:Cavium Networks
1197:Analog Devices
1089:, a multi-core
1072:, a multi-core
1057:
1048:
1042:
1039:
1028:
1011:
1007:
1000:
977:
949:
924:Cavium Networks
912:
889:embedded system
881:
862:SQL Server 2014
842:
664:
618:
578:"heterogeneous"
569:
564:
548:
504:cache coherency
500:
418:
389:
354:
273:
244:run in parallel
209:general-purpose
182:instruction set
146:message passing
138:multiprocessing
89:(CPUs), called
63:Athlon X2 6400+
35:
28:
23:
22:
15:
12:
11:
5:
6055:
6045:
6044:
6039:
6034:
6029:
6024:
6007:
6006:
6004:
6003:
5991:
5988:
5987:
5985:
5984:
5979:
5974:
5969:
5967:Race condition
5964:
5959:
5954:
5949:
5944:
5938:
5936:
5932:
5931:
5929:
5928:
5923:
5918:
5913:
5908:
5903:
5898:
5893:
5888:
5883:
5878:
5873:
5868:
5863:
5858:
5853:
5848:
5843:
5838:
5833:
5828:
5823:
5818:
5813:
5808:
5802:
5800:
5794:
5793:
5791:
5790:
5785:
5780:
5779:
5778:
5768:
5762:
5761:
5760:
5755:
5750:
5745:
5740:
5735:
5725:
5724:
5723:
5718:
5711:Multiprocessor
5708:
5703:
5698:
5693:
5688:
5687:
5686:
5681:
5676:
5675:
5674:
5669:
5664:
5653:
5642:
5640:
5634:
5633:
5631:
5630:
5625:
5624:
5623:
5618:
5613:
5603:
5598:
5592:
5590:
5584:
5583:
5581:
5580:
5575:
5570:
5565:
5560:
5555:
5550:
5544:
5542:
5538:
5537:
5535:
5534:
5529:
5524:
5519:
5514:
5508:
5506:
5502:
5501:
5499:
5498:
5493:
5488:
5483:
5478:
5473:
5468:
5463:
5458:
5452:
5450:
5446:
5445:
5443:
5442:
5440:Hardware scout
5437:
5431:
5426:
5421:
5415:
5410:
5404:
5398:
5396:
5394:Multithreading
5390:
5389:
5387:
5386:
5381:
5376:
5371:
5366:
5361:
5356:
5351:
5345:
5343:
5339:
5338:
5336:
5335:
5333:Systolic array
5330:
5325:
5320:
5315:
5310:
5305:
5300:
5295:
5290:
5284:
5282:
5278:
5277:
5270:
5269:
5262:
5255:
5247:
5238:
5237:
5235:
5234:
5229:
5227:Pin grid array
5224:
5219:
5214:
5209:
5204:
5199:
5194:
5188:
5186:
5182:
5181:
5179:
5178:
5172:
5167:
5162:
5157:
5152:
5147:
5141:
5139:
5131:
5130:
5127:
5126:
5124:
5123:
5118:
5113:
5108:
5103:
5098:
5097:
5096:
5091:
5086:
5075:
5073:
5067:
5066:
5064:
5063:
5061:Barrel shifter
5058:
5057:
5056:
5051:
5044:Binary decoder
5041:
5040:
5039:
5029:
5024:
5019:
5013:
5011:
5005:
5004:
5002:
5001:
4996:
4988:
4983:
4978:
4973:
4967:
4965:
4959:
4958:
4956:
4955:
4950:
4945:
4940:
4935:
4933:Stack register
4930:
4925:
4919:
4917:
4911:
4910:
4908:
4907:
4906:
4905:
4900:
4890:
4885:
4880:
4874:
4872:
4866:
4865:
4863:
4862:
4857:
4856:
4855:
4844:
4839:
4834:
4833:
4832:
4826:
4815:
4809:
4803:
4796:
4794:
4783:
4782:
4777:
4772:
4767:
4762:
4761:
4760:
4755:
4750:
4745:
4740:
4735:
4725:
4719:
4717:
4713:
4712:
4710:
4709:
4704:
4699:
4694:
4688:
4686:
4682:
4681:
4679:
4678:
4677:
4676:
4666:
4661:
4656:
4651:
4646:
4641:
4636:
4631:
4626:
4621:
4616:
4611:
4606:
4601:
4595:
4593:
4587:
4586:
4583:
4582:
4580:
4579:
4574:
4569:
4564:
4558:
4552:
4546:
4540:
4535:
4529:
4527:AI accelerator
4524:
4518:
4516:
4508:
4507:
4505:
4504:
4498:
4493:
4490:Multiprocessor
4487:
4480:
4478:
4472:
4471:
4469:
4468:
4463:
4458:
4453:
4448:
4443:
4441:Microprocessor
4438:
4432:
4430:
4429:By application
4423:
4422:
4416:
4410:
4404:
4399:
4394:
4389:
4384:
4379:
4374:
4372:Tile processor
4369:
4364:
4359:
4354:
4353:
4352:
4341:
4334:
4332:
4326:
4325:
4323:
4322:
4317:
4312:
4306:
4300:
4294:
4288:
4282:
4281:
4280:
4268:
4262:
4260:
4252:
4251:
4248:
4247:
4245:
4244:
4243:
4242:
4232:
4227:
4226:
4225:
4220:
4215:
4210:
4200:
4194:
4192:
4186:
4185:
4183:
4182:
4177:
4172:
4167:
4166:
4165:
4160:
4158:Hyperthreading
4150:
4144:
4142:
4140:Multithreading
4136:
4135:
4133:
4132:
4127:
4122:
4121:
4120:
4110:
4109:
4108:
4103:
4093:
4092:
4091:
4086:
4076:
4071:
4070:
4069:
4064:
4053:
4051:
4044:
4038:
4037:
4034:
4033:
4031:
4030:
4025:
4019:
4017:
4011:
4010:
4008:
4007:
4002:
3997:
3996:
3995:
3990:
3980:
3974:
3972:
3966:
3965:
3963:
3962:
3957:
3952:
3947:
3941:
3939:
3933:
3932:
3930:
3929:
3924:
3919:
3917:Pipeline stall
3913:
3911:
3902:
3896:
3895:
3892:
3891:
3889:
3888:
3883:
3878:
3873:
3870:
3869:
3868:
3866:z/Architecture
3863:
3858:
3853:
3845:
3840:
3835:
3830:
3825:
3820:
3815:
3810:
3805:
3800:
3795:
3790:
3785:
3784:
3783:
3778:
3773:
3765:
3760:
3755:
3750:
3745:
3740:
3735:
3730:
3724:
3722:
3716:
3715:
3713:
3712:
3711:
3710:
3700:
3695:
3690:
3685:
3680:
3675:
3670:
3669:
3668:
3658:
3657:
3656:
3646:
3641:
3636:
3631:
3625:
3623:
3616:
3608:
3607:
3605:
3604:
3599:
3594:
3589:
3584:
3579:
3578:
3577:
3572:
3570:Virtual memory
3562:
3557:
3556:
3555:
3550:
3545:
3540:
3530:
3525:
3520:
3515:
3510:
3509:
3508:
3498:
3493:
3487:
3485:
3479:
3478:
3476:
3475:
3474:
3473:
3468:
3463:
3458:
3448:
3443:
3438:
3437:
3436:
3431:
3426:
3421:
3416:
3411:
3406:
3401:
3394:Turing machine
3391:
3390:
3389:
3384:
3379:
3374:
3369:
3364:
3354:
3349:
3343:
3341:
3335:
3334:
3327:
3326:
3319:
3312:
3304:
3298:
3297:
3291:
3280:
3271:
3263:
3262:External links
3260:
3259:
3258:
3241:
3222:
3219:
3217:
3216:
3202:
3177:
3149:
3134:
3110:
3086:
3061:
3051:. 2 April 2019
3036:
3018:
3000:
2978:
2959:
2933:
2909:
2885:
2861:
2837:
2813:
2789:
2765:
2739:
2715:
2691:
2661:
2647:
2621:
2584:
2578:10.1.1.37.4309
2555:
2552:on 2009-12-21.
2537:
2511:
2490:
2472:
2453:
2422:
2395:
2386:
2360:
2333:
2326:
2306:
2291:
2266:
2242:
2218:
2207:(2): 173–193.
2191:
2164:
2153:www.top500.org
2140:
2112:
2109:on 2011-11-25.
2087:
2059:
2057:
2054:
2053:
2052:
2037:
2028:CPU and a DSP
2013:
2010:
2008:
2007:
2002:
2000:Race condition
1997:
1991:
1986:
1980:
1975:
1970:
1965:
1960:
1955:
1950:
1944:
1942:
1939:
1933:
1930:
1929:
1928:
1922:
1916:
1906:
1896:
1895:
1894:
1888:
1872:
1862:
1854:
1851:
1850:
1849:
1842:
1839:
1838:
1837:
1830:
1829:
1828:
1822:
1811:
1810:
1809:
1806:
1795:
1794:
1793:
1786:Sunway SW26010
1780:
1779:
1778:
1772:
1766:
1760:
1754:
1748:
1742:
1731:
1705:
1704:
1703:
1695:
1689:
1686:ARM Cortex-M0+
1678:
1671:
1665:
1659:
1649:
1648:
1647:
1632:
1631:
1630:
1627:
1624:
1614:
1613:
1612:
1601:
1600:
1599:
1592:
1586:
1585:
1584:
1578:
1572:
1566:
1560:
1554:
1528:
1522:
1516:
1510:
1504:
1498:
1475:
1474:
1473:
1466:
1456:
1455:
1454:
1448:
1442:
1436:
1430:
1424:
1418:
1415:z/Architecture
1413:, a quad-core
1408:
1394:
1384:
1378:
1372:
1366:
1360:
1354:
1348:
1343:, a dual-core
1333:
1318:
1305:
1299:Coherent Logix
1296:
1283:
1280:
1279:
1278:
1270:
1256:
1255:
1254:
1247:
1241:
1240:
1239:
1236:
1233:
1225:
1219:
1202:
1195:
1194:
1193:
1179:
1173:
1167:
1161:
1155:
1149:
1143:
1137:
1131:
1125:
1115:
1104:
1094:
1081:
1064:
1056:
1053:
1050:
1049:
1014:
1012:
1005:
999:
996:
994:architecture.
992:ARM big.LITTLE
976:
973:
948:
945:
911:
908:
880:
877:
876:
875:
869:
841:
838:
822:
821:
817:
813:
812:
808:
804:
803:
799:
795:
794:
790:
663:
660:
631:True Octa-core
617:
614:
601:computers and
568:
565:
563:
560:
547:
544:
536:front-side bus
510:(alternative:
499:
496:
483:
482:
481:
480:
457:
446:
432:
427:Additionally:
417:
414:
412:architecture.
388:
385:
353:
350:
272:
269:
229:supercomputers
202:multithreading
111:multithreading
79:microprocessor
26:
9:
6:
4:
3:
2:
6054:
6043:
6040:
6038:
6035:
6033:
6030:
6028:
6025:
6023:
6020:
6019:
6017:
6002:
5993:
5992:
5989:
5983:
5980:
5978:
5975:
5973:
5970:
5968:
5965:
5963:
5960:
5958:
5955:
5953:
5950:
5948:
5945:
5943:
5940:
5939:
5937:
5933:
5927:
5924:
5922:
5919:
5917:
5914:
5912:
5909:
5907:
5904:
5902:
5899:
5897:
5894:
5892:
5889:
5887:
5884:
5882:
5879:
5877:
5874:
5872:
5869:
5867:
5864:
5862:
5859:
5857:
5856:Global Arrays
5854:
5852:
5849:
5847:
5844:
5842:
5839:
5837:
5834:
5832:
5829:
5827:
5824:
5822:
5819:
5817:
5814:
5812:
5809:
5807:
5804:
5803:
5801:
5799:
5795:
5789:
5786:
5784:
5783:Grid computer
5781:
5777:
5774:
5773:
5772:
5769:
5766:
5763:
5759:
5756:
5754:
5751:
5749:
5746:
5744:
5741:
5739:
5736:
5734:
5731:
5730:
5729:
5726:
5722:
5719:
5717:
5714:
5713:
5712:
5709:
5707:
5704:
5702:
5699:
5697:
5694:
5692:
5689:
5685:
5682:
5680:
5677:
5673:
5670:
5668:
5665:
5662:
5659:
5658:
5657:
5654:
5652:
5649:
5648:
5647:
5644:
5643:
5641:
5639:
5635:
5629:
5626:
5622:
5619:
5617:
5614:
5612:
5609:
5608:
5607:
5604:
5602:
5599:
5597:
5594:
5593:
5591:
5589:
5585:
5579:
5576:
5574:
5571:
5569:
5566:
5564:
5561:
5559:
5556:
5554:
5551:
5549:
5546:
5545:
5543:
5539:
5533:
5530:
5528:
5525:
5523:
5520:
5518:
5515:
5513:
5510:
5509:
5507:
5503:
5497:
5494:
5492:
5489:
5487:
5484:
5482:
5479:
5477:
5474:
5472:
5469:
5467:
5464:
5462:
5459:
5457:
5454:
5453:
5451:
5447:
5441:
5438:
5435:
5432:
5430:
5427:
5425:
5422:
5419:
5416:
5414:
5411:
5408:
5405:
5403:
5400:
5399:
5397:
5395:
5391:
5385:
5382:
5380:
5377:
5375:
5372:
5370:
5367:
5365:
5362:
5360:
5357:
5355:
5352:
5350:
5347:
5346:
5344:
5340:
5334:
5331:
5329:
5326:
5324:
5321:
5319:
5316:
5314:
5311:
5309:
5306:
5304:
5301:
5299:
5296:
5294:
5291:
5289:
5286:
5285:
5283:
5279:
5275:
5268:
5263:
5261:
5256:
5254:
5249:
5248:
5245:
5233:
5230:
5228:
5225:
5223:
5220:
5218:
5215:
5213:
5210:
5208:
5205:
5203:
5200:
5198:
5195:
5193:
5190:
5189:
5187:
5183:
5176:
5173:
5171:
5168:
5166:
5163:
5161:
5158:
5156:
5153:
5151:
5148:
5146:
5143:
5142:
5140:
5138:
5132:
5122:
5119:
5117:
5114:
5112:
5109:
5107:
5104:
5102:
5099:
5095:
5092:
5090:
5087:
5085:
5082:
5081:
5080:
5077:
5076:
5074:
5072:
5068:
5062:
5059:
5055:
5052:
5050:
5047:
5046:
5045:
5042:
5038:
5035:
5034:
5033:
5030:
5028:
5025:
5023:
5022:Demultiplexer
5020:
5018:
5015:
5014:
5012:
5010:
5006:
5000:
4997:
4995:
4992:
4989:
4987:
4984:
4982:
4979:
4977:
4974:
4972:
4969:
4968:
4966:
4964:
4960:
4954:
4951:
4949:
4946:
4944:
4943:Memory buffer
4941:
4939:
4938:Register file
4936:
4934:
4931:
4929:
4926:
4924:
4921:
4920:
4918:
4916:
4912:
4904:
4901:
4899:
4896:
4895:
4894:
4891:
4889:
4886:
4884:
4881:
4879:
4878:Combinational
4876:
4875:
4873:
4871:
4867:
4861:
4858:
4854:
4851:
4850:
4848:
4845:
4843:
4840:
4838:
4835:
4830:
4827:
4825:
4822:
4821:
4819:
4816:
4813:
4810:
4807:
4804:
4801:
4798:
4797:
4795:
4793:
4787:
4781:
4778:
4776:
4773:
4771:
4768:
4766:
4763:
4759:
4756:
4754:
4751:
4749:
4746:
4744:
4741:
4739:
4736:
4734:
4731:
4730:
4729:
4726:
4724:
4721:
4720:
4718:
4714:
4708:
4705:
4703:
4700:
4698:
4695:
4693:
4690:
4689:
4687:
4683:
4675:
4672:
4671:
4670:
4667:
4665:
4662:
4660:
4657:
4655:
4652:
4650:
4647:
4645:
4642:
4640:
4637:
4635:
4632:
4630:
4627:
4625:
4622:
4620:
4617:
4615:
4612:
4610:
4607:
4605:
4602:
4600:
4597:
4596:
4594:
4592:
4588:
4578:
4575:
4573:
4570:
4568:
4565:
4562:
4559:
4556:
4553:
4550:
4547:
4544:
4541:
4539:
4536:
4533:
4530:
4528:
4525:
4523:
4520:
4519:
4517:
4515:
4509:
4502:
4499:
4497:
4494:
4491:
4488:
4485:
4482:
4481:
4479:
4473:
4467:
4464:
4462:
4459:
4457:
4454:
4452:
4449:
4447:
4444:
4442:
4439:
4437:
4434:
4433:
4431:
4427:
4420:
4417:
4414:
4411:
4408:
4405:
4403:
4400:
4398:
4395:
4393:
4390:
4388:
4385:
4383:
4380:
4378:
4375:
4373:
4370:
4368:
4365:
4363:
4360:
4358:
4355:
4351:
4348:
4347:
4345:
4342:
4339:
4336:
4335:
4333:
4331:
4327:
4321:
4318:
4316:
4313:
4310:
4307:
4304:
4301:
4298:
4295:
4292:
4289:
4286:
4283:
4278:
4275:
4274:
4272:
4269:
4267:
4264:
4263:
4261:
4259:
4253:
4241:
4238:
4237:
4236:
4233:
4231:
4228:
4224:
4221:
4219:
4216:
4214:
4211:
4209:
4206:
4205:
4204:
4201:
4199:
4196:
4195:
4193:
4191:
4187:
4181:
4178:
4176:
4173:
4171:
4168:
4164:
4161:
4159:
4156:
4155:
4154:
4151:
4149:
4146:
4145:
4143:
4141:
4137:
4131:
4128:
4126:
4123:
4119:
4116:
4115:
4114:
4111:
4107:
4104:
4102:
4099:
4098:
4097:
4094:
4090:
4087:
4085:
4082:
4081:
4080:
4077:
4075:
4072:
4068:
4065:
4063:
4060:
4059:
4058:
4055:
4054:
4052:
4048:
4045:
4043:
4039:
4029:
4026:
4024:
4021:
4020:
4018:
4016:
4012:
4006:
4003:
4001:
3998:
3994:
3991:
3989:
3986:
3985:
3984:
3981:
3979:
3978:Scoreboarding
3976:
3975:
3973:
3971:
3967:
3961:
3960:False sharing
3958:
3956:
3953:
3951:
3948:
3946:
3943:
3942:
3940:
3938:
3934:
3928:
3925:
3923:
3920:
3918:
3915:
3914:
3912:
3910:
3906:
3903:
3901:
3897:
3887:
3884:
3882:
3879:
3877:
3874:
3871:
3867:
3864:
3862:
3859:
3857:
3854:
3852:
3849:
3848:
3846:
3844:
3841:
3839:
3836:
3834:
3831:
3829:
3826:
3824:
3821:
3819:
3816:
3814:
3811:
3809:
3806:
3804:
3801:
3799:
3796:
3794:
3791:
3789:
3786:
3782:
3779:
3777:
3774:
3772:
3769:
3768:
3766:
3764:
3761:
3759:
3756:
3754:
3753:Stanford MIPS
3751:
3749:
3746:
3744:
3741:
3739:
3736:
3734:
3731:
3729:
3726:
3725:
3723:
3717:
3709:
3706:
3705:
3704:
3701:
3699:
3696:
3694:
3691:
3689:
3686:
3684:
3681:
3679:
3676:
3674:
3671:
3667:
3664:
3663:
3662:
3659:
3655:
3652:
3651:
3650:
3647:
3645:
3642:
3640:
3637:
3635:
3632:
3630:
3627:
3626:
3624:
3620:
3617:
3615:
3614:architectures
3609:
3603:
3600:
3598:
3595:
3593:
3590:
3588:
3585:
3583:
3582:Heterogeneous
3580:
3576:
3573:
3571:
3568:
3567:
3566:
3563:
3561:
3558:
3554:
3551:
3549:
3546:
3544:
3541:
3539:
3536:
3535:
3534:
3533:Memory access
3531:
3529:
3526:
3524:
3521:
3519:
3516:
3514:
3511:
3507:
3504:
3503:
3502:
3499:
3497:
3494:
3492:
3489:
3488:
3486:
3484:
3480:
3472:
3469:
3467:
3466:Random-access
3464:
3462:
3459:
3457:
3454:
3453:
3452:
3449:
3447:
3446:Stack machine
3444:
3442:
3439:
3435:
3432:
3430:
3427:
3425:
3422:
3420:
3417:
3415:
3412:
3410:
3407:
3405:
3402:
3400:
3397:
3396:
3395:
3392:
3388:
3385:
3383:
3380:
3378:
3375:
3373:
3370:
3368:
3365:
3363:
3362:with datapath
3360:
3359:
3358:
3355:
3353:
3350:
3348:
3345:
3344:
3342:
3340:
3336:
3332:
3325:
3320:
3318:
3313:
3311:
3306:
3305:
3302:
3295:
3292:
3290:
3289:
3288:IEEE Spectrum
3284:
3281:
3279:
3275:
3272:
3269:
3266:
3265:
3255:
3251:
3247:
3242:
3238:
3234:
3230:
3225:
3224:
3212:
3206:
3198:
3194:
3193:
3188:
3181:
3162:
3161:
3153:
3145:
3138:
3124:
3123:ark.intel.com
3120:
3114:
3100:
3099:ark.intel.com
3096:
3090:
3076:
3072:
3065:
3050:
3046:
3040:
3032:
3028:
3022:
3014:
3010:
3004:
2996:
2992:
2988:
2982:
2974:
2970:
2963:
2949:on 2006-11-09
2948:
2944:
2937:
2923:
2922:ark.intel.com
2919:
2913:
2899:
2898:ark.intel.com
2895:
2889:
2875:
2874:ark.intel.com
2871:
2865:
2851:
2850:ark.intel.com
2847:
2841:
2827:
2826:ark.intel.com
2823:
2817:
2803:
2802:ark.intel.com
2799:
2793:
2779:
2778:ark.intel.com
2775:
2769:
2753:
2749:
2743:
2729:
2728:ark.intel.com
2725:
2719:
2705:
2704:ark.intel.com
2701:
2695:
2679:
2675:
2671:
2665:
2657:
2651:
2636:
2632:
2625:
2609:
2605:
2601:
2600:
2595:
2588:
2579:
2574:
2570:
2566:
2559:
2551:
2547:
2541:
2526:
2522:
2515:
2508:
2504:
2500:
2494:
2486:
2482:
2476:
2468:
2464:
2457:
2441:
2437:
2433:
2430:Clark, Jack.
2426:
2410:
2406:
2399:
2390:
2374:
2370:
2364:
2348:
2344:
2337:
2329:
2323:
2319:
2318:
2310:
2302:
2295:
2281:
2277:
2270:
2256:
2252:
2246:
2232:
2228:
2222:
2214:
2210:
2206:
2202:
2195:
2179:
2175:
2168:
2154:
2150:
2144:
2130:
2126:
2123:Smith, Ryan.
2119:
2117:
2105:
2098:
2091:
2075:
2071:
2064:
2060:
2050:
2046:
2043:Two types of
2041:
2038:
2035:
2034:mobile phones
2031:
2027:
2023:
2019:
2016:
2015:
2006:
2003:
2001:
1998:
1995:
1992:
1990:
1987:
1984:
1981:
1979:
1976:
1974:
1971:
1969:
1966:
1964:
1961:
1959:
1956:
1954:
1951:
1949:
1948:CPU shielding
1946:
1945:
1938:
1926:
1923:
1920:
1917:
1914:
1910:
1907:
1904:
1900:
1897:
1893:
1889:
1887:
1883:
1882:
1880:
1876:
1873:
1870:
1866:
1863:
1860:
1857:
1856:
1848:
1845:
1844:
1835:
1831:
1826:
1823:
1820:
1817:
1816:
1815:
1812:
1807:
1804:
1803:TMS320C80 MVP
1801:
1800:
1799:
1796:
1791:
1787:
1784:
1783:
1781:
1776:
1773:
1770:
1767:
1764:
1763:UltraSPARC T3
1761:
1758:
1757:UltraSPARC T2
1755:
1752:
1751:UltraSPARC T1
1749:
1746:
1745:UltraSPARC IV
1743:
1740:
1737:
1736:
1735:
1732:
1729:
1728:PlayStation 3
1725:
1721:
1717:
1713:
1709:
1706:
1701:
1700:
1699:
1696:
1693:
1690:
1687:
1683:
1679:
1676:
1672:
1669:
1666:
1663:
1660:
1657:
1653:
1650:
1645:
1641:
1638:
1637:
1636:
1633:
1628:
1625:
1622:
1618:
1617:
1615:
1610:
1607:
1606:
1605:
1602:
1597:
1593:
1590:
1589:
1587:
1582:
1579:
1576:
1573:
1570:
1567:
1564:
1561:
1558:
1555:
1552:
1548:
1544:
1540:
1536:
1532:
1529:
1526:
1523:
1520:
1517:
1514:
1511:
1508:
1505:
1502:
1499:
1496:
1492:
1488:
1484:
1481:
1480:
1479:
1476:
1471:
1467:
1465:
1462:
1461:
1460:
1457:
1452:
1449:
1446:
1443:
1440:
1437:
1434:
1431:
1428:
1425:
1422:
1419:
1416:
1412:
1409:
1407:game console.
1406:
1402:
1398:
1395:
1392:
1388:
1385:
1382:
1379:
1376:
1373:
1370:
1367:
1364:
1361:
1358:
1355:
1352:
1349:
1346:
1342:
1339:
1338:
1337:
1334:
1331:
1327:
1323:
1319:
1316:
1313:
1309:
1306:
1303:
1300:
1297:
1294:
1291:
1287:
1284:
1281:
1276:
1275:
1274:
1271:
1269:
1265:
1261:
1257:
1252:
1248:
1245:
1244:
1242:
1237:
1234:
1231:
1230:
1229:
1226:
1223:
1220:
1217:
1213:
1209:
1206:
1203:
1200:
1196:
1191:
1187:
1183:
1180:
1177:
1174:
1171:
1168:
1165:
1162:
1159:
1156:
1153:
1150:
1147:
1144:
1141:
1138:
1135:
1132:
1129:
1126:
1123:
1119:
1116:
1113:
1110:
1109:
1108:
1105:
1102:
1098:
1095:
1092:
1088:
1085:
1082:
1079:
1075:
1071:
1068:
1065:
1062:
1059:
1058:
1046:
1036:
1032:
1026:
1024:
1020:
1015:This article
1013:
1004:
1003:
995:
993:
988:
986:
982:
972:
970:
966:
962:
958:
954:
944:
942:
938:
933:
929:
925:
921:
917:
907:
905:
900:
897:
890:
885:
873:
870:
867:
863:
859:
855:
851:
847:
846:
845:
837:
835:
831:
827:
818:
815:
814:
809:
807:Agglomeration
806:
805:
800:
798:Communication
797:
796:
791:
788:
787:
786:
784:
779:
776:
772:
768:
764:
760:
754:
752:
748:
744:
740:
736:
732:
728:
724:
720:
716:
712:
708:
704:
700:
698:
694:
688:
684:
682:
678:
674:
673:thread-safety
669:
658:
656:
652:
651:Anant Agarwal
648:
646:
641:
637:
634:
632:
626:
624:
623:heterogeneous
613:
611:
606:
604:
600:
596:
592:
588:
583:
579:
575:
559:
555:
553:
546:Disadvantages
543:
539:
537:
532:
528:
523:
521:
517:
513:
509:
505:
495:
492:
488:
478:
474:
470:
466:
462:
458:
455:
451:
447:
444:
440:
439:
437:
433:
430:
429:
428:
425:
423:
413:
411:
407:
401:
399:
395:
384:
382:
381:
376:
373:
369:
368:
363:
359:
358:semiconductor
349:
347:
343:
339:
334:
332:
328:
327:
321:
319:
314:
312:
308:
307:
302:
298:
294:
290:
286:
282:
278:
268:
264:
260:
258:
253:
249:
245:
241:
236:
234:
230:
226:
222:
218:
214:
210:
205:
203:
199:
195:
191:
187:
183:
179:
175:
174:heterogeneous
171:
167:
163:
159:
155:
151:
150:shared-memory
147:
143:
139:
134:
132:
128:
124:
121:, known as a
120:
116:
112:
108:
104:
100:
96:
92:
88:
84:
80:
76:
72:
64:
59:
52:
47:
39:
33:
19:
5541:Coordination
5471:Amdahl's law
5407:Simultaneous
5232:Chip carrier
5170:Clock gating
5089:Mixed-signal
4986:Write buffer
4963:Control unit
4775:Clock signal
4696:
4514:accelerators
4496:Cypress PSoC
4153:Simultaneous
3970:Out-of-order
3602:Neuromorphic
3483:Architecture
3441:Belt machine
3434:Zeno machine
3367:Hierarchical
3286:
3277:
3245:
3237:10657.1/2440
3228:
3205:
3190:
3180:
3168:. Retrieved
3159:
3152:
3137:
3126:. Retrieved
3122:
3113:
3102:. Retrieved
3098:
3089:
3078:. Retrieved
3074:
3064:
3053:. Retrieved
3048:
3039:
3021:
3003:
2981:
2962:
2951:. Retrieved
2947:the original
2936:
2925:. Retrieved
2921:
2912:
2901:. Retrieved
2897:
2888:
2877:. Retrieved
2873:
2864:
2853:. Retrieved
2849:
2840:
2829:. Retrieved
2825:
2816:
2805:. Retrieved
2801:
2792:
2781:. Retrieved
2777:
2768:
2756:. Retrieved
2752:the original
2742:
2731:. Retrieved
2727:
2718:
2707:. Retrieved
2703:
2694:
2682:. Retrieved
2673:
2664:
2650:
2639:. Retrieved
2624:
2612:. Retrieved
2599:Ars Technica
2597:
2587:
2571:(1): 13–23.
2568:
2564:
2558:
2550:the original
2540:
2528:. Retrieved
2514:
2506:
2503:the original
2493:
2485:the original
2475:
2466:
2456:
2444:. Retrieved
2435:
2425:
2413:. Retrieved
2409:the original
2398:
2389:
2377:. Retrieved
2363:
2351:. Retrieved
2336:
2316:
2309:
2294:
2283:. Retrieved
2279:
2269:
2258:. Retrieved
2254:
2245:
2234:. Retrieved
2230:
2221:
2204:
2200:
2194:
2182:. Retrieved
2178:the original
2167:
2156:. Retrieved
2152:
2143:
2132:. Retrieved
2128:
2104:the original
2090:
2078:. Retrieved
2074:the original
2063:
2040:
2018:
1978:Multitasking
1935:
1470:home gateway
1391:Power Mac G5
1328:, dual core
1251:Raspberry Pi
1228:Azul Systems
1212:ARM11 MPCore
1122:Athlon 64 X2
1118:Athlon 64 FX
1040:
1029:Please help
1017:may contain
1016:
989:
978:
950:
913:
901:
899:processors.
894:
888:
843:
823:
789:Partitioning
780:
755:
749:, and IBM's
725:, Skandium,
701:
689:
685:
681:video codecs
665:
649:
642:
639:
635:
630:
627:
619:
616:Architecture
607:
589:and dynamic
570:
556:
549:
540:
524:
512:Bus snooping
501:
484:
476:
472:
468:
460:
449:
442:
426:
419:
402:
390:
378:
365:
355:
335:
330:
324:
322:
317:
315:
310:
304:
296:
280:
276:
274:
265:
261:
248:Amdahl's law
237:
206:
135:
127:chip package
122:
98:
94:
90:
81:on a single
74:
70:
68:
5977:Scalability
5738:distributed
5621:Concurrency
5588:Programming
5429:Cooperative
5418:Speculative
5354:Instruction
5017:Multiplexer
4981:Data buffer
4692:Single-core
4664:bit slicing
4522:Coprocessor
4377:Coprocessor
4258:performance
4180:Cooperative
4170:Speculative
4130:Distributed
4089:Superscalar
4074:Instruction
4042:Parallelism
4015:Speculative
3847:System/3x0
3719:Instruction
3496:Von Neumann
3409:Post–Turing
3166:. Hot Chips
2758:12 November
2530:October 21,
2303:. WikiChip.
1588:IntellaSys
1551:Core 2 Quad
1525:Core 2 Quad
1387:PowerPC 970
1332:processors.
826:server side
820:scheduling.
783:concurrency
508:cache snoop
473:memory wall
443:memory wall
398:Clock rates
372:superscalar
352:Development
271:Terminology
257:refactoring
223:(DSP), and
194:superscalar
18:Triple-core
6016:Categories
5982:Starvation
5721:asymmetric
5456:PRAM model
5424:Preemptive
5137:management
5032:Multiplier
4893:Logic gate
4883:Sequential
4790:Functional
4770:Clock rate
4743:Data cache
4716:Components
4697:Multi-core
4685:Core count
4175:Preemptive
4079:Pipelining
4062:Bit-serial
4005:Wide-issue
3950:Structural
3872:Tilera ISA
3838:MicroBlaze
3808:ETRAX CRIS
3703:Comparison
3548:Load–store
3528:Endianness
3270:—MakeUseOf
3128:2019-05-04
3104:2019-05-04
3080:2019-05-04
3055:2019-05-04
2953:2006-09-28
2927:2019-05-04
2903:2019-05-04
2879:2019-05-04
2855:2019-05-04
2831:2019-05-04
2807:2019-05-04
2783:2019-05-04
2733:2019-05-04
2709:2019-05-04
2641:2014-03-04
2614:5 December
2604:Condé Nast
2285:2024-05-22
2260:2023-10-27
2236:2023-10-27
2158:2020-09-15
2134:2020-09-15
2056:References
1932:Benchmarks
1903:Wavescalar
1867:, 16-core
1623:processor.
1547:Core 2 Duo
1519:Core 2 Duo
1472:processor.
1273:ClearSpeed
1186:FireStream
1055:Commercial
1023:irrelevant
834:throughput
582:big.LITTLE
498:Advantages
469:power wall
461:power wall
375:pipelining
323:The terms
291:(DSP) and
277:multi-core
275:The terms
178:big.LITTLE
51:Core 2 Duo
5716:symmetric
5461:PEM model
5071:Circuitry
4991:Microcode
4915:Registers
4758:coherence
4733:CPU cache
4591:Word size
4256:Processor
3900:Execution
3803:DEC Alpha
3781:Power ISA
3597:Cognitive
3404:Universal
3029:. Intel.
3011:. Intel.
2628:Compare:
2573:CiteSeerX
1915:processor
1905:processor
1871:processor
1847:OpenSPARC
1684:, a dual
1668:Plurality
1312:Power ISA
1260:Tensilica
1243:Broadcom
1152:Phenom II
1134:FX-Series
1128:Athlon II
1043:July 2016
1019:excessive
961:Freescale
850:Microsoft
840:Licensing
781:Managing
741:, Cray's
711:Cilk Plus
693:microcode
465:shrinking
326:many-core
318:multi-CPU
281:dual-core
113:or other
99:quad-core
95:dual-core
49:An Intel
5947:Deadlock
5935:Problems
5901:pthreads
5881:OpenHMPP
5806:Ateji PX
5767:computer
5638:Hardware
5505:Elements
5491:Slowdown
5402:Temporal
5384:Pipeline
5009:Datapath
4702:Manycore
4674:variable
4512:Hardware
4148:Temporal
3828:OpenRISC
3523:Cellular
3513:Dataflow
3506:modified
3197:Archived
3192:PC World
3049:TechSpot
3031:Archived
3013:Archived
2995:Archived
2973:Archived
2678:Archived
2635:Archived
2608:Archived
2525:EE Times
2446:6 August
2440:Archived
2415:March 6,
2379:March 6,
2373:Archived
2347:Archived
2280:PC Gamer
2184:March 6,
2080:March 6,
1968:Manycore
1941:See also
1925:UC Davis
1859:Stanford
1853:Academic
1775:SPARC T5
1769:SPARC T4
1692:SiCortex
1675:Kilocore
1673:Rapport
1662:picoChip
1640:RTX 3090
1609:MPPA-256
1581:Xeon Phi
1549:and the
1513:Core Duo
1487:netbooks
1459:Infineon
1405:Xbox 360
1199:Blackfin
1112:A-Series
1061:Adapteva
1025:examples
969:Picochip
932:Broadcom
928:Wintegra
793:problem.
747:Fortress
745:, Sun's
723:FastFlow
719:OpenHMPP
709:such as
697:picocode
625:" role.
610:manycore
562:Hardware
477:ILP wall
450:ILP wall
240:software
233:one case
225:graphics
213:embedded
184:, while
170:crossbar
5906:RaftLib
5886:OpenACC
5861:GPUOpen
5851:C++ AMP
5826:Charm++
5568:Barrier
5512:Process
5496:Speedup
5281:General
5185:Related
5116:Quantum
5106:Digital
5101:Boolean
4999:Counter
4898:Quantum
4659:512-bit
4654:256-bit
4649:128-bit
4492:(MPSoC)
4477:on chip
4475:Systems
4293:(FLOPS)
4106:Process
3955:Control
3937:Hazards
3823:Itanium
3818:Unicore
3776:PowerPC
3501:Harvard
3461:Pointer
3456:Counter
3414:Quantum
3170:27 June
2467:Giz Bot
2353:May 17,
1881:(AsAP)
1825:TILE-Gx
1782:Sunway
1724:PowerPC
1716:Toshiba
1563:Pentium
1557:Itanium
1543:Core i9
1539:Core i7
1535:Core i5
1531:Core i3
1507:Celeron
1497:(MIDs).
1491:nettops
1381:Power10
1345:PowerPC
1330:PA-RISC
1326:PA-8900
1322:PA-8800
1253:models)
1158:Sempron
1140:Opteron
830:threads
816:Mapping
767:Fortran
645:Renesas
591:voltage
520:degrade
516:signals
360:-based
217:network
77:) is a
61:An AMD
5999:
5876:OpenCL
5871:OpenMP
5816:Chapel
5733:shared
5728:Memory
5663:(SIMT)
5606:Models
5517:Thread
5449:Theory
5420:(SpMT)
5374:Memory
5359:Thread
5342:Levels
5121:Switch
5111:Analog
4849:(IMC)
4820:(MMU)
4669:others
4644:64-bit
4639:48-bit
4634:32-bit
4629:24-bit
4624:16-bit
4619:15-bit
4614:12-bit
4451:Mobile
4367:Stream
4362:Barrel
4357:Vector
4346:(GPU)
4305:(SUPS)
4273:(IPC)
4125:Memory
4118:Vector
4101:Thread
4084:Scalar
3886:Others
3833:RISC-V
3798:SuperH
3767:Power
3763:MIPS-X
3738:PDP-11
3587:Fabric
3339:Models
2575:
2324:
2005:Thread
1996:(PGAS)
1983:OpenCL
1819:TILE64
1814:Tilera
1698:SiFive
1682:RP2040
1635:Nvidia
1621:MIPS64
1604:Kalray
1375:POWER9
1369:POWER8
1363:POWER7
1357:POWER6
1351:POWER5
1341:POWER4
1208:MPCore
1182:Radeon
1164:Turion
1146:Phenom
1103:(MPPA)
1097:Ambric
985:Xilinx
864:, and
854:socket
743:Chapel
731:Erlang
729:, and
715:OpenMP
655:Tilera
599:laptop
597:(i.e.
567:Trends
198:vector
168:, and
142:caches
5846:Dryad
5811:Boost
5532:Array
5522:Fiber
5436:(CMT)
5409:(SMT)
5323:GPGPU
5177:(PPW)
5135:Power
5027:Adder
4903:Array
4870:Logic
4831:(TLB)
4814:(FPU)
4808:(AGU)
4802:(ALU)
4792:units
4728:Cache
4609:8-bit
4604:4-bit
4599:1-bit
4563:(TPU)
4557:(DSP)
4551:(PPU)
4545:(VPU)
4534:(GPU)
4503:(NoC)
4486:(SoC)
4421:(PoP)
4415:(SiP)
4409:(MCM)
4350:GPGPU
4340:(CPU)
4330:Types
4311:(PPW)
4299:(TPS)
4287:(IPS)
4279:(CPI)
4050:Level
3861:S/390
3856:S/370
3851:S/360
3793:SPARC
3771:POWER
3654:TRIPS
3622:Types
3164:(PDF)
3075:Intel
2991:Intel
2684:5 May
2436:ZDNet
2255:ZDNET
2231:PCMAG
2107:(PDF)
2100:(PDF)
2012:Notes
1958:GPGPU
1913:TRIPS
1892:AsAP2
1832:XMOS
1644:GPGPU
1478:Intel
1464:AURIX
1451:Telum
1427:zEC12
1397:Xenon
1268:ESP32
1222:ASOCS
1190:GPGPU
1170:Ryzen
1087:PhysX
1084:Ageia
1074:SPARC
1070:LEON3
941:6WIND
487:Intel
406:Intel
200:, or
91:cores
5911:ROCm
5841:CUDA
5831:Cilk
5798:APIs
5758:COMA
5753:NUMA
5684:MIMD
5679:MISD
5656:SIMD
5651:SISD
5379:Loop
5369:Data
5364:Task
5155:ACPI
4888:Glue
4780:FIFO
4723:Core
4461:ASIP
4402:CPLD
4397:FPOA
4392:FPGA
4387:ASIC
4240:SPMD
4235:MIMD
4230:MISD
4223:SWAR
4203:SIMD
4198:SISD
4113:Data
4096:Task
4067:Word
3813:M32R
3758:MIPS
3721:sets
3688:ZISC
3683:NISC
3678:OISC
3673:MISC
3666:EPIC
3661:VLIW
3649:EDGE
3639:RISC
3634:CISC
3543:HUMA
3538:NUMA
3172:2023
2760:2007
2686:2019
2616:2015
2532:2023
2448:2015
2417:2013
2381:2013
2355:2015
2322:ISBN
2186:2013
2082:2013
2026:RISC
1953:CUDA
1886:AsAP
1841:Free
1739:MAJC
1720:Cell
1708:Sony
1575:Xeon
1541:and
1483:Atom
1421:z196
1324:and
1290:MIPS
1214:and
1188:GPU/
1184:and
1176:Epyc
1120:and
930:and
775:ACML
765:and
593:and
489:and
475:and
459:The
448:The
441:The
342:FPGA
329:and
311:same
297:same
279:and
190:VLIW
166:mesh
162:ring
5926:ZPL
5921:TBB
5916:UPC
5896:PVM
5866:MPI
5821:HPX
5748:UMA
5349:Bit
5150:APM
5145:PMU
5037:CPU
4994:ROM
4765:Bus
4382:PAL
4057:Bit
3843:LMC
3748:ARM
3743:x86
3733:VAX
3250:doi
3233:hdl
2674:AMD
2209:doi
2030:MPU
1869:RAW
1865:MIT
1718:'s
1712:IBM
1445:z15
1439:z14
1433:z13
1411:z10
1401:SMT
1336:IBM
1315:MPU
1293:MPU
1266:'s
1205:ARM
1107:AMD
1021:or
979:In
951:In
887:An
751:X10
735:TBB
727:MPI
695:or
668:GUI
605:).
531:SMP
491:AMD
410:x86
301:die
158:bus
148:or
119:die
97:or
75:MCP
6018::
5084:3D
3195:.
3189:.
3121:.
3097:.
3073:.
3047:.
2993:.
2989:.
2971:.
2920:.
2896:.
2872:.
2848:.
2824:.
2800:.
2776:.
2726:.
2702:.
2676:.
2672:.
2606:.
2602:.
2596:.
2567:.
2523:.
2465:.
2438:.
2434:.
2345:.
2278:.
2253:.
2229:.
2205:21
2203:.
2151:.
2127:.
2115:^
1911:,
1901:,
1877:,
1537:,
1533:,
1489:,
926:,
922:,
860:,
836:.
771:C#
753:.
721:,
717:,
713:,
699:.
259:.
219:,
215:,
211:,
204:.
196:,
192:,
160:,
69:A
5266:e
5259:t
5252:v
3323:e
3316:t
3309:v
3285:—
3276:—
3256:.
3252::
3239:.
3235::
3213:.
3174:.
3146:.
3131:.
3107:.
3083:.
3058:.
2956:.
2930:.
2906:.
2882:.
2858:.
2834:.
2810:.
2786:.
2762:.
2736:.
2712:.
2688:.
2658:.
2644:.
2618:.
2581:.
2569:4
2534:.
2469:.
2450:.
2419:.
2383:.
2357:.
2330:.
2288:.
2263:.
2239:.
2215:.
2211::
2188:.
2161:.
2137:.
2084:.
1792:.
1730:.
1714:/
1710:/
1658:.
1598:.
1553:.
1393:.
1317:.
1295:.
1192:.
1093:.
1080:.
1045:)
1041:(
1037:.
1027:.
763:C
621:"
479:.
73:(
34:.
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.