547:
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.
873:
5985:
47:
27:
997:
35:
664:). 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
767:
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.
766:
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
679:
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
530:
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
482:
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
380:
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
546:
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
255:
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
675:
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
392:
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
887:
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
799:
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
790:
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
527:(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.
659:
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
243:
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
522:
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
895:, 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.
617:
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
251:
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.
372:(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.
923:
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
745:
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
609:
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,
676:
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.
434:; 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.
800:
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.
808:
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
573:
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
781:
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
3185:
1925:
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.
2040:(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.
560:
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
452:; 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 "
353:
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
660:
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
2623:
2596:
495:
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
543:(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.
2469:
629:
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.
2013:(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
3175:
2335:
976:
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.
2021:. 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
833:
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.
932:
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.
511:
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.
569:(or asymmetric) cores promise further performance and efficiency gains, especially in processing multimedia, recognition and networking applications. For example, a
2113:
98:(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
423:
For general-purpose processors, much of the motivation for multi-core processors comes from greatly diminished gains in processor performance from increasing the
5311:
4338:
177:
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
1880:
2058:
1874:
863:
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.
389:
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.
2619:
952:
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
3019:
3001:
2983:
2961:
2582:
3220:. 17th International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA-11). Las Vegas, Nevada, USA. pp. 551–557.
2393:
2162:
1052:
Epiphany, a many-core processor architecture which allows up to 4096 processors on-chip, although only a 16-core version has been commercially produced.
1055:
413:(SMP) designs using discrete CPUs, the issues regarding implementing multi-core processor architecture and supporting it with software are well known.
2382:
Patterson, David A. "Future of computer architecture." Berkeley EECS Annual
Research Symposium (BEARS), College of Engineering, UC Berkeley, US. 2006.
460:
poses manufacturing, system design and deployment problems that have not been justified in the face of the diminished gains in performance due to the
3059:
3310:
2428:
366:
are suitable for many applications, but are inefficient for others that contain difficult-to-predict code. Many applications are better suited to
2239:
646:, took the opposing view. He said multi-core chips need to be homogeneous collections of general-purpose cores to keep the software model simple.
2487:
2451:
309:
refers to multiple physically separate processing-units (which often contain special circuitry to facilitate communication between each other).
4449:
3632:
3107:
636:, generally agreed. He suggested the cellphone's use of many specialty cores working in concert is a good model for future multi-core designs.
2473:
597:
Chips designed from the outset for a large number of cores (rather than having evolved from single core designs) are sometimes referred to as
5401:
4151:
3083:
129:
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
5253:
3429:
2906:
2762:
2736:
4308:
3874:
3691:
2858:
953:
4662:
2712:
2688:
2331:
1597:
916:
1011:
672:
are impossible to parallelize because each result generated is used to help create the next result of the entropy decoding algorithm.
3654:
2085:
3262:
2552:
John
Darlinton; Moustafa Ghanem; Yike Guo; Hing Wing To (1996). "Guided Resource Organisation in Heterogeneous Parallel Computing".
4303:
1196:
334:
4375:
1867:
727:
5382:
4657:
4128:
2882:
2361:
190:
99:
3033:
2931:
2314:
1531:
231:
algorithms used and their implementation. In particular, possible gains are limited by the fraction of the software that can
5422:
2264:
2025:. In other applications, a growing number of companies have developed multi-core DSPs with very large numbers of processors.
5649:
5072:
4196:
3459:
3303:
3147:
1007:
5672:
5082:
4223:
3237:. The 20th IEEE International Conference on High Performance Computer Architecture (HPCA-14) workshop. Orlando, FL, USA.
817:, multi-core processors are ideal because they allow many users to connect to a site simultaneously and have independent
6015:
5561:
3350:
3132:
2534:
1560:(Polaris), a 3.16 GHz, 80-core processor prototype, which the company originally stated would be released by 2011.
1089:
474:
In order to continue delivering regular performance improvements for general-purpose processors, manufacturers such as
5417:
2834:
2496:
Galaxy smartphones run on either Octa-core (2.3GHz Quad + 1.6GHz Quad) or Quad-core (2.15GHz + 1.6GHz Dual) processors
5667:
5644:
4390:
4218:
4191:
3570:
2014:
774:
acquires a central role in developing parallel applications. The basic steps in designing parallel applications are:
382:
322:
are sometimes used to describe multi-core architectures with an especially high number of cores (tens to thousands).
3282:
2289:
1600:, 256-core processor, released 2012 (256 usable VLIW cores, Network-on-Chip (NoC), 32/64-bit IEEE 754 compliant FPU)
5246:
5205:
4768:
3661:
3627:
3622:
3541:
3506:
1982:
539:
Maximizing the usage of the computing resources provided by multi-core processors requires adjustments both to the
508:
5639:
5454:
5180:
5077:
4478:
4385:
4186:
3407:
3296:
2810:
2666:
2062:
1863:
1100:
531:
performance with less energy. A challenge in this, however, is the additional overhead of writing parallel code.
174:
2786:
6010:
5746:
5660:
5609:
4206:
3925:
3360:
3015:
2997:
2975:
2957:
2304:
1977:
1715:
core and eight specialized SPUs (Synergistic
Processing Unit) optimized for vector operations used in the Sony
972:, where a system uses more than one kind of processor or cores, multi-core solutions are becoming more common:
759:
2397:
1566:
dual-, quad-, 6-, 8-, 10-, 12-, 14-, 15-, 16-, 18-, 20-, 22-, 24-, 26-, 28-, 32-, 48-, and 56-core processors.
625:
The article "CPU designers debate multi-core future" by Rick
Merritt, EE Times 2008, includes these comments:
5970:
5804:
5655:
5342:
4380:
4228:
4201:
4062:
3676:
3637:
3494:
2166:
735:
731:
719:
442:
355:
330:
284:(SoC). The terms are generally used only to refer to multi-core microprocessors that are manufactured on the
6020:
4817:
4579:
4055:
4016:
3671:
3666:
3600:
3412:
170:
1161:, dual-, quad-, 6-, 8-, 12-, 16-, 24-, 32-, and 64-core desktop, mobile, and embedded platform processors.
6030:
5989:
5935:
5395:
5239:
4444:
4141:
3839:
3536:
1897:
1534:, a family of dual-, quad-, 6-, 8-, 10-, 12-, 14-, 16-, and 18-core processors, and the successor of the
1389:
747:
562:
256:
cores) started entering the mainstream and since the early 2020s has overtaken quad-core in many spaces.
30:
Diagram of a generic dual-core processor with CPU-local level-1 caches and a shared, on-die level-2 cache
2420:
891:
In addition, embedded software is typically developed for a specific hardware release, making issues of
5914:
5709:
5594:
5556:
5406:
5296:
5094:
4741:
4158:
3649:
3617:
3387:
3375:
3355:
1439:
1415:
739:
695:
655:
An outdated version of an anti-virus application may create a new thread for a scan process, while its
178:
2114:"NVIDIA Announces the GeForce RTX 30 Series: Ampere For Gaming, Starting With RTX 3080 & RTX 3090"
622:, or similar styling, as opposed to being merely two sets of quad-cores each with fixed clock speeds.
420:
Using a proven processing-core design without architectural changes reduces design risk significantly.
292:; separate microprocessor dies in the same package are generally referred to by another name, such as
5930:
5909:
5854:
5741:
5731:
5704:
5566:
5185:
5148:
5138:
3526:
3271:
2491:
2037:
1993:
941:
925:
818:
723:
715:
583:
519:
410:
209:
3256:
711:
514:
Assuming that the die can physically fit into the package, multi-core CPU designs require much less
227:
The improvement in performance gained by the use of a multi-core processor depends very much on the
6025:
5884:
5510:
5449:
5362:
5200:
4607:
4543:
4520:
4370:
4332:
4168:
4118:
4113:
3590:
3484:
3392:
2645:"6WINDGATE Software: Network Optimization Software – SDN Software – Control Plane Software | 6WIND"
2566:
2010:
1887:
1822:
1791:
1433:
1427:
1421:
1303:
1281:
751:
722:
can be used on multi-core platforms. Intel introduced a new abstraction for C++ parallelism called
656:
368:
345:
While manufacturing technology improves, reducing the size of individual gates, physical limits of
289:
277:
213:
107:
2509:
5945:
5940:
5799:
5390:
5153:
4936:
4830:
4794:
4711:
4695:
4537:
4326:
4285:
4273:
4136:
4050:
3971:
3736:
3397:
3340:
1557:
1296:
1079:
1023:
969:
908:
611:
579:
566:
273:
240:
162:
75:
20:
1907:
5684:
5616:
5520:
5412:
5367:
4959:
4931:
4841:
4806:
4555:
4549:
4531:
4265:
4259:
4163:
4067:
3958:
3897:
3759:
3402:
2740:
2561:
1483:
1095:
363:
1019:
5776:
5736:
5689:
5679:
5474:
5337:
5276:
5133:
5042:
4788:
4500:
4318:
4077:
4045:
4003:
3915:
3716:
3531:
3521:
3511:
3501:
3471:
3454:
3319:
2188:
Duran, A (2011). "Ompss: a proposal for programming heterogeneous multi-core architectures".
1966:
1961:
1708:
1489:
763:
591:
515:
182:
5716:
5604:
5599:
5589:
5576:
5372:
5163:
5099:
4407:
4297:
4244:
3776:
3489:
3345:
3327:
1659:
HAL series tightly coupled 16-256 cores, L1 shared memory, hardware synchronized processor.
1290:
1204:
892:
850:
771:
691:
2092:
1666:
KC256, a 257-core microcontroller with a PowerPC core and 256 8-bit "processing elements".
1498:, the first dual-core (and, later, quad-core) processor for the budget/entry-level market.
8:
5879:
5834:
5634:
5500:
5210:
5195:
5015:
4866:
4848:
4812:
4800:
4454:
4401:
4178:
4094:
3976:
3831:
3726:
3585:
1656:
956:
with 40 and 80 general purpose ALUs per chip, all programmable in C as a SIMD engine and
854:
326:
821:
of execution. This allows for Web servers and application servers that have much better
298:. This article uses the terms "multi-core" and "dual-core" for CPUs manufactured on the
216:(GPU). Core count goes up to even dozens, and for specialized chips over 10,000, and in
5904:
5753:
5726:
5551:
5515:
5505:
5464:
5306:
5286:
5281:
5262:
5067:
5059:
4911:
4886:
4690:
4565:
4089:
4030:
3910:
3642:
3370:
3218:
Predicting CPU Availability of a Multi-core
Processor Executing Concurrent Java Threads
1901:
1640:
1207:
processor cores, intended for high-performance embedded and entertainment applications.
904:
884:
860:
314:
232:
103:
71:
5950:
5626:
5584:
5479:
5020:
4987:
4903:
4835:
4736:
4726:
4716:
4647:
4642:
4637:
4560:
4489:
4395:
4355:
3988:
3938:
3888:
3864:
3746:
3686:
3681:
3563:
3479:
2310:
1786:
1778:
1628:
1385:
1278:
1266:
CSX700, 192-core processor, released in 2008 (32/64-bit floating point; Integer ALU).
1252:
945:
424:
294:
205:
119:
2357:
2215:
5960:
5759:
5694:
5541:
5357:
5352:
5347:
5316:
5190:
5123:
5109:
4964:
4871:
4825:
4632:
4627:
4622:
4617:
4612:
4602:
4472:
4439:
4350:
4345:
4254:
4106:
4101:
4084:
4072:
4011:
3575:
3553:
3439:
3417:
3335:
3238:
3235:
A New
Composite CPU/Memory Model for Predicting Efficiency of Multi-core Processing
3221:
2197:
2033:
1722:
1584:
1193:
872:
791:
proceed. This information flow is specified in the communication phase of a design.
665:
575:
540:
397:
has produced a 48-core processor for research in cloud computing; each core has an
350:
281:
186:
142:
3232:
3215:
2935:
1066:
5824:
5764:
5699:
5546:
5536:
5469:
5459:
5301:
5291:
5104:
5089:
5037:
4941:
4916:
4753:
4746:
4597:
4592:
4587:
4526:
4434:
4424:
4146:
3981:
3933:
3696:
3580:
3548:
3449:
3444:
3365:
3176:"Meet KiloCore, a 1,000-core processor so efficient it could run on a AA battery"
2551:
2240:"Intel taking its six-core processors mainstream in 2018 with Coffee Lake family"
1951:
1674:
1644:
1274:
912:
504:
492:
236:
201:
196:
Multi-core processors are widely used across many application domains, including
158:
154:
134:
126:
95:
91:
1238:
BCM2836, BCM2837, BCM2710 and BCM2711 quad-core ARM SoC (designed for different
1131:, single-, dual-, quad-, 6-, 8-, 12-, and 16-core server/workstation processors.
1058:
5955:
5771:
5428:
5321:
5215:
5049:
5032:
5025:
4921:
4778:
4515:
4429:
4360:
3943:
3905:
3854:
3849:
3844:
3558:
3382:
2018:
1988:
1774:
1492:, single-core, dual-core, and quad-core processors for smartphones and tablets.
1452:
1403:
1174:
980:
880:
on a plug-in card with processor, memory, power supply, and external interfaces
570:
524:
496:
166:
67:
3242:
2201:
2138:"Sunway TaihuLight - Sunway MPP, Sunway SW26010 260C 1.45GHz, Sunway | TOP500"
1569:
6004:
5844:
5721:
5010:
4926:
3966:
3948:
3741:
3434:
3276:
1936:
1751:
1745:
1739:
1733:
1716:
661:
639:
507:
between different CPUs travel shorter distances, and therefore those signals
346:
217:
138:
3869:
2592:
1167:, quad-, 8-, 12-, 16-, 24-, 32-, and 64-core server and embedded processors.
960:
with 300 processors on a single die, focused on communication applications.
224:
up to 20 million processing elements total in addition to host processors).
5444:
5220:
5158:
4974:
4951:
4763:
4484:
3422:
3199:
2644:
2587:
2036:
are able to use a dual-CPU multiprocessor: partitioned multiprocessing and
2022:
1916:, Kilocore, a 1000 core 1.78 GHz processor on a 32 nm IBM process
1458:
1379:
1239:
1216:
1200:
1110:
1106:
500:
150:
115:
51:
5965:
5005:
4969:
4680:
4652:
4510:
4365:
3288:
1554:, single, dual-core, and quad-core processors for the entry-level market.
1539:
1513:
1375:
814:
669:
360:
245:
146:
106:
techniques. Manufacturers typically integrate the cores onto a single IC
3233:
Khondker S. Hasan; John
Antonio; Sridhar Radhakrishnan (February 2014).
3225:
2538:
2137:
4891:
4881:
4876:
4858:
4758:
4731:
3993:
3826:
3796:
3516:
3034:"Intel shows off Xeon Platinum CPU with up to 56 cores and 112 threads"
1891:
1857:
1535:
1507:
1471:
1261:
1152:
1134:
842:
822:
453:
386:
39:
2658:
46:
5839:
5814:
5231:
4982:
4979:
4721:
3791:
3769:
1835:
1300:
1248:
1140:
1116:
1103:, dual-, triple-, and quad-core of Accelerated Processor Units (APU).
949:
838:
699:
681:
381:
be used in the design, which increased functionality, especially for
130:
3216:
Khondker S. Hasan; Nicolas G. Grounds; John K. Antonio (July 2011).
1474:, single, dual-core, quad-core, 8-, 12-, and 16-core processors for
26:
5889:
5869:
5794:
4997:
3816:
3180:
2583:"Windows Server 2016 moving to per core, not per socket, licensing"
2513:
1974:(Open Computing Language) – a framework for heterogeneous execution
1956:
1913:
1847:
1763:
1757:
1680:
1663:
1650:
1501:
1447:
1409:
1393:
1187:
1049:
957:
920:
758:, which perform math computations faster than newer languages like
707:
694:
techniques can benefit from multiple cores directly. Some existing
685:
598:
228:
197:
3285:, published on Feb 19, 2010 (more than one dead link in the slide)
2265:"Six-core CPUs are now more popular than quad-core chips on Steam"
220:(i.e. clusters of chips) the count can go over 10 million (and in
5894:
5874:
5849:
5484:
3806:
3764:
2332:"Intel Shows 48-core x86 Processor as Single-chip Cloud Computer"
1813:
1712:
1704:
1551:
1545:
1527:
1523:
1519:
1495:
1475:
1399:
1369:
1333:
1318:
1314:
1310:
1146:
1128:
846:
755:
633:
221:
1287:
456:" the processor by using smaller traces for the same logic. The
5864:
5859:
3821:
3786:
3751:
3108:"Intel® Xeon Phi™ 72x5 Processor Family Product Specifications"
1971:
1807:
1802:
1686:
1670:
1623:
1609:
1592:
1479:
1363:
1357:
1351:
1345:
1339:
1329:
1170:
1122:
1085:
973:
703:
643:
587:
491:
The proximity of multiple CPU cores on the same die allows the
161:. Homogeneous multi-core systems include only identical cores;
235:
simultaneously on multiple cores; this effect is described by
4279:
3811:
3781:
3084:"Intel® Xeon Phi™ x100 Product Family Product Specifications"
2979:
1946:
1653:
PC200 series 200–300 cores per device for DSP & wireless.
1632:
1466:
1256:
1210:
1178:
1158:
1075:
1072:
1062:
929:
475:
394:
2472:(Press release). MediaTek. November 20, 2013. Archived from
2452:"These 5 Myths About the Octa-Core Phones Are Actually True"
1271:
Cradle
Technologies CT3400 and CT3600, both multi-core DSPs.
5899:
5829:
5819:
5143:
4291:
4211:
3801:
2907:"Intel® Pentium® Processor D Series Product Specifications"
2763:"Intel® Celeron® Processor J Series Product Specifications"
2622:. OMT-CO Operations Management Technology Consulting GmbH.
1941:
1727:
1696:
1563:
1442:, an eight-core z/Architecture processor, released in 2021.
1436:, a twelve-core z/Architecture processor, released in 2019.
1424:, an eight-core z/Architecture processor, released in 2015.
1164:
165:
multi-core systems have cores that are not identical (e.g.
34:
1711:
processor, a nine-core processor with one general purpose
948:
has the three-core TMS320C6488 and four-core TMS320C5441,
5809:
5786:
3731:
3721:
2859:"Intel® Core™ X-series Processors Product Specifications"
2470:"MediaTek Launches MT6592 True Octa-Core Mobile Platform"
1853:
1700:
1412:, a quad-core z/Architecture processor, released in 2010.
1324:
479:
398:
118:. As of 2024, the microprocessors used in almost all new
2713:"Intel® Atom™ Processor Z Series Product Specifications"
2689:"Intel® Atom™ Processor C Series Product Specifications"
1430:, a ten-core z/Architecture processor, released in 2017.
1418:, a six-core z/Architecture processor, released in 2012.
1354:, a 4, 6 and 8-core PowerPC processor, released in 2010.
3133:"40-core processor with Forth-based IDE tools unveiled"
1392:-capable, PowerPC microprocessor used in the Microsoft
1149:, single-, dual-, and quad-core entry level processors.
1143:, dual-, triple-, quad-, and 6-core desktop processors.
1683:"SiCortex node" has six MIPS64 cores on a single chip.
1548:, single, dual-core, quad-core, and 8-core processors.
1372:, a 15 or 30-core PowerPC processor, released in 2021.
1366:, a 12 or 24-core PowerPC processor, released in 2017.
1251:
Xtensa LX6, available in a dual-core configuration in
445:
to keep a high-performance single-core processor busy.
1378:
MP, a dual-core PowerPC processor, used in the Apple
845:
licensing system. However, for some software such as
2358:"Intel unveils 48-core cloud computing silicon chip"
19:"Dual Core" redirects here. Not to be confused with
3200:"COSMIC:Statistical Multiprocessor Benchmark Suite"
1618:
XLS, an eight-core, quad-threaded MIPS64 processor.
1615:
XLR, an eight-core, quad-threaded MIPS64 processor.
1516:, 2 dual-core dies packaged in a multi-chip module.
1119:, dual-, triple-, and quad-core desktop processors.
409:Since computer manufacturers have long implemented
2883:"Intel® Itanium® Processor Product Specifications"
2394:"Q & A: Do multicores save energy? Not really"
2107:
2105:
1766:, a sixteen-core, 128-concurrent-thread processor.
1754:, a sixteen-core, 128-concurrent-thread processor.
1348:, a dual-core PowerPC processor, released in 2007.
1342:, a dual-core PowerPC processor, released in 2004.
1199:is a fully synthesizable multi-core container for
16:Microprocessor with more than one processing unit
6002:
3060:"2nd Gen Intel® Xeon® Scalable Processors Brief"
2659:"Sempron™ 3850 APU with Radeon™ R3 Series | AMD"
2290:"The 2,048-core PEZY-SC2 sets a Green500 record"
1760:, an eight-core, 64-concurrent-thread processor.
1748:, an eight-core, 64-concurrent-thread processor.
1572:57-, 60-, 61-, 64-, 68-, and 72-core processors.
1360:, a 12-core PowerPC processor, released in 2013.
837:Initially, for some of its enterprise software,
2574:
2102:
1213:ModemX, up to 128 cores, wireless applications.
907:have become mainstream, with companies such as
3283:Architecting solutions for the Manycore future
1227:Vega 3, a 54-core processor, released in 2008.
1224:Vega 2, a 48-core processor, released in 2006.
1221:Vega 1, a 24-core processor, released in 2005.
857:, Microsoft has shifted to per-core licensing.
601:designs, emphasising qualitative differences.
441:; the increasing difficulty of finding enough
82:to emphasize their multiplicity (for example,
5247:
3304:
2958:"Intel releases 15-core Xeon E7 v2 processor"
2620:"The Licensing Of Oracle Technology Products"
632:Atsushi Hasegawa, a senior chief engineer at
169:have heterogeneous cores that share the same
1583:SEAforth24, a 24-core processor designed by
935:
305:In contrast to multi-core systems, the term
302:integrated circuit, unless otherwise noted.
4309:Computer performance by orders of magnitude
2262:
1137:, dual-, triple-, and quad-core processors.
1125:, quad-, 6-, and 8-core desktop processors.
5254:
5240:
3318:
3311:
3297:
3272:"Multicore Is Bad News for Supercomputers"
2737:"Intel Preps Dual-Core Celeron Processors"
2421:"Intel: Why a 1,000-core chip is feasible"
1635:cores; plus other more specialized cores).
1155:, single- and dual-core laptop processors.
928:(SMP) operating system. Companies such as
503:) operations. Put simply, this means that
443:parallelism in a single instruction stream
2955:
2565:
1794:, a five-core multimedia video processor.
1736:and UltraSPARC IV+, dual-core processors.
1113:single- and dual-core desktop processors.
750:to access code written in languages like
642:, founder and chief executive of startup
276:(CPU), but are sometimes also applied to
141:inter-core communication methods. Common
114:(CMP), or onto multiple dies in a single
2510:"CPU designers debate multi-core future"
2449:
1299:QorIQ series processors, up to 8 cores,
1190:BF561, a symmetrical dual-core processor
1022:by adding descriptive text and removing
963:
871:
867:
427:. This is due to three primary factors:
375:
335:semiconductor intellectual property core
45:
33:
25:
3145:
2929:
2507:
2391:
2163:"What makes parallel programming hard?"
2160:
1868:Asynchronous array of simple processors
6003:
5261:
3173:
2956:Kowaliski, Cyril (February 18, 2014).
2580:
2329:
2028:
565:, memory-on-chip, and special-purpose
5235:
3292:
2554:Journal of High Performance Computing
2364:from the original on December 6, 2012
2302:
2287:
2187:
2056:
1742:, an eight-core, 32-thread processor.
898:
726:. Other research efforts include the
4280:Floating-point operations per second
3146:Hammond, Lance; et al. (1999).
3131:Cole, Bernard (September 24, 2008).
3130:
2930:Zazaian, Mike (September 26, 2006).
2599:from the original on 4 December 2015
2338:from the original on January 5, 2016
2086:"Multicore Processors – A Necessity"
1797:TMS320TMS320C66, 2-, 4-, 8-core DSP.
1580:SEAforth 40C18, a 40-core processor.
990:
986:
404:
272:most commonly refer to some sort of
3188:from the original on June 23, 2016.
3016:"Intel Xeon Processor E3 v2 Family"
2998:"Intel Xeon Processor E7 v2 Family"
2976:"Intel Xeon Processor E7 v3 Family"
2535:"Multicore Packet Processing Forum"
2083:
1777:, a 260-core processor used in the
650:
333:. Each "core" can be considered a "
145:used to interconnect cores include
13:
3209:
2508:Merritt, Rick (February 6, 2008).
2431:from the original on 6 August 2015
2059:"Definition: multi-core processor"
2057:Rouse, Margaret (March 27, 2007).
1090:massively parallel processor array
125:A multi-core processor implements
14:
6042:
3250:
2739:. 11 October 2007. Archived from
2581:Bright, Peter (4 December 2015).
2450:Kudikala, Chakri (Aug 27, 2016).
2418:
2330:Shrout, Ryan (December 2, 2009).
2111:
1457:Danube, a dual-core, MIPS-based,
383:complex instruction set computing
5984:
5983:
5206:Semiconductor device fabrication
3022:from the original on 2015-07-07.
3004:from the original on 2015-07-07.
2986:from the original on 2015-07-07.
2964:from the original on 2014-10-11.
2488:"What is an Octa-core processor"
2006:
1983:Partitioned global address space
995:
534:
175:AMD Accelerated Processing Units
90:). Each core reads and executes
5455:Analysis of parallel algorithms
5181:History of general-purpose CPUs
3408:Nondeterministic Turing machine
3192:
3167:
3139:
3124:
3100:
3076:
3057:
3051:
3026:
3008:
2990:
2968:
2949:
2923:
2899:
2875:
2851:
2827:
2803:
2779:
2755:
2729:
2705:
2681:
2669:from the original on 4 May 2019
2651:
2637:
2626:from the original on 2014-03-21
2611:
2545:
2527:
2501:
2480:
2462:
2443:
2412:
2392:Suleman, Aater (May 19, 2011).
2385:
2376:
2350:
2323:
2296:
2281:
2161:Suleman, Aater (May 20, 2011).
1864:University of California, Davis
1293:, a 100-core DSP/GPP processor.
680:that were based on proprietary
604:
518:(PCB) space than do multi-chip
74:(IC) with two or more separate
3361:Deterministic finite automaton
3174:Chacos, Brad (June 20, 2016).
2835:"Products formerly Kentsfield"
2288:Schor, David (November 2017).
2256:
2232:
2208:
2181:
2154:
2130:
2077:
2050:
1978:Parallel random access machine
1730:5200, two-core VLIW processor.
1608:XLP, a 32-core, quad-threaded
340:
259:
239:. In the best case, so-called
1:
5402:Simultaneous and heterogenous
4152:Simultaneous and heterogenous
3263:"Embedded moves to multicore"
2165:. FutureChips. Archived from
2044:
1920:
1816:, a 72-core 64-bit processor.
1810:, a 64-core 32-bit processor.
1482:, embedded applications, and
1235:SiByte SB1250, SB1255, SB1455
1043:
486:
356:instruction-level parallelism
5990:Category: Parallel computing
4836:Integrated memory controller
4818:Translation lookaside buffer
4017:Memory dependence prediction
3460:Random-access stored program
3413:Probabilistic Turing machine
2303:Vajda, András (2011-06-10).
2061:. TechTarget. Archived from
1406:processor, released in 2008.
1336:processor, released in 2001.
828:
7:
4292:Synaptic updates per second
3257:"What Is a Processor Core?"
2306:Programming Many-Core Chips
2190:Parallel Processing Letters
1929:
1898:University of Texas, Austin
1841:
1490:Atom SoC (system on a chip)
979:Mobile devices may use the
696:parallel programming models
563:simultaneous multithreading
550:
10:
6047:
5297:High-performance computing
4696:Heterogeneous architecture
3618:Orthogonal instruction set
3388:Alternating Turing machine
3376:Quantum cellular automaton
2811:"Products formerly Conroe"
2263:Alan Dexter (2022-04-05).
813:On the other hand, on the
18:
6016:Digital signal processing
5979:
5931:Automatic parallelization
5923:
5785:
5625:
5575:
5567:Application checkpointing
5529:
5493:
5437:
5381:
5330:
5269:
5186:Microprocessor chronology
5173:
5149:Dynamic frequency scaling
5122:
5058:
4996:
4950:
4902:
4857:
4777:
4704:
4673:
4578:
4499:
4463:
4417:
4317:
4304:Cache performance metrics
4243:
4177:
4127:
4038:
4029:
4002:
3957:
3924:
3896:
3887:
3707:
3610:
3599:
3470:
3326:
3267:Embedded Computing Design
3243:10.13140/RG.2.1.3051.9207
2932:"Intel: 80 Cores by 2011"
2787:"Products formerly Yonah"
2490:. Samsung. Archived from
2360:. BBC. December 3, 2009.
2216:"Definition of dual core"
2202:10.1142/S0129626411000151
2038:symmetric multiprocessing
2011:Digital signal processors
1910:, Sweden, ePUMA processor
942:digital signal processing
936:Digital signal processing
926:symmetric multiprocessing
555:
483:functions into the chip.
411:symmetric multiprocessing
337:" as well as a CPU core.
329:cores placed on a single
278:digital signal processors
210:digital signal processing
133:, and they may implement
42:E6750 dual-core processor
5201:Hardware security module
4544:Digital signal processor
4521:Graphics processing unit
4333:Graphics processing unit
2000:
1888:University of Washington
1850:, 4-core Hydra processor
1823:Software Defined Silicon
1641:Parallax Propeller P8X32
1510:, a dual-core processor.
1504:, a dual-core processor.
944:the same trend applies:
762:. Intel's MKL and AMD's
369:thread-level parallelism
94:, specifically ordinary
76:central processing units
5946:Embarrassingly parallel
5941:Deterministic algorithm
5154:Dynamic voltage scaling
4937:Memory address register
4831:Branch target predictor
4795:Address generation unit
4538:Physics processing unit
4327:Central processing unit
4286:Transactions per second
4274:Instructions per second
4197:Array processing (SIMT)
3341:Stored-program computer
2309:. Springer. p. 3.
1829:
1558:Teraflops Research Chip
1484:mobile internet devices
1297:Freescale Semiconductor
1247:Cadence Design Systems
1080:physics processing unit
1024:less pertinent examples
970:heterogeneous computing
909:Freescale Semiconductor
903:As of 2010, multi-core
841:continued to use a per-
274:central processing unit
241:embarrassingly parallel
21:Dual Core (hip hop duo)
5661:Associative processing
5617:Non-blocking algorithm
5423:Clustered multi-thread
4960:Hardwired control unit
4842:Memory management unit
4807:Memory management unit
4556:Secure cryptoprocessor
4550:Tensor Processing Unit
4532:Vision processing unit
4266:Cycles per instruction
4260:Instructions per cycle
4207:Associative processing
3898:Instruction pipelining
3320:Processor technologies
3149:The Stanford Hydra CMP
1879:167-core 1.2 GHz
1605:NetLogic Microsystems
1067:fault-tolerant version
1065:that also exists in a
954:Stream Processors, Inc
881:
648:
592:portable media players
385:(CISC) architectures.
359:(ILP) methods such as
325:Some systems use many
55:
43:
31:
6011:Computer architecture
5777:Hardware acceleration
5690:Superscalar processor
5680:Dataflow architecture
5277:Distributed computing
5043:Sum-addressed decoder
4789:Arithmetic logic unit
3916:Classic RISC pipeline
3870:Epiphany architecture
3717:Motorola 68000 series
2494:on January 17, 2022.
1962:Multicore Association
1873:36-core 610 MHz
964:Heterogeneous systems
875:
868:Embedded applications
728:Codeplay Sieve System
627:
516:printed circuit board
376:Commercial incentives
49:
37:
29:
5656:Pipelined processing
5605:Explicit parallelism
5600:Implicit parallelism
5590:Dataflow programming
5164:Performance per watt
4742:replacement policies
4408:Package on a package
4298:Performance per watt
4202:Pipelined processing
3972:Tomasulo's algorithm
3777:Clipper architecture
3633:Application-specific
3346:Finite-state machine
2476:on October 29, 2020.
2400:on December 16, 2012
1908:Linköping University
1691:U74 includes 4 cores
1205:ARM Cortex-A9 MPCore
893:software portability
692:Parallel programming
393:cores. For example,
320:massively multi-core
92:program instructions
60:multi-core processor
5880:Parallel Extensions
5685:Pipelined processor
5196:Digital electronics
4849:Instruction decoder
4801:Floating-point unit
4455:Soft microprocessor
4402:System in a package
3977:Reservation station
3507:Transport-triggered
1631:(10496 CUDA cores,
1088:Am2045, a 336-core
1020:improve the article
855:Windows Server 2016
847:BizTalk Server 2013
748:numerical libraries
668:algorithms used in
425:operating frequency
327:soft microprocessor
288:integrated circuit
112:chip multiprocessor
54:dual-core processor
6031:Parallel computing
5754:Massively parallel
5732:distributed shared
5552:Cache invalidation
5516:Instruction window
5307:Manycore processor
5287:Massively parallel
5282:Parallel computing
5263:Parallel computing
5068:Integrated circuit
4912:Processor register
4566:Baseband processor
3911:Operand forwarding
3371:Cellular automaton
2743:on 4 November 2007
1669:Raspberry Pi Ltd.
1277:Octeon, a 32-core
905:network processors
899:Network processors
885:Embedded computing
882:
861:Oracle Corporation
153:, two-dimensional
143:network topologies
120:personal computers
104:parallel computing
72:integrated circuit
56:
44:
32:
5998:
5997:
5951:Parallel slowdown
5585:Stream processing
5475:Karp–Flatt metric
5229:
5228:
5118:
5117:
4737:Instruction cache
4727:Scratchpad memory
4574:
4573:
4561:Network processor
4490:Network on a chip
4445:Ultra-low-voltage
4396:Multi-chip module
4239:
4238:
4025:
4024:
4012:Branch prediction
3989:Register renaming
3883:
3882:
3865:VISC architecture
3687:Quantum computing
3682:VISC architecture
3564:Secondary storage
3480:Microarchitecture
3440:Register machines
2316:978-1-4419-9739-5
2118:www.anandtech.com
2065:on August 5, 2010
2034:operating systems
1825:quad-core XS1-G4.
1787:Texas Instruments
1779:Sunway TaihuLight
1388:, a triple-core,
1253:Espressif Systems
1041:
1040:
987:Hardware examples
946:Texas Instruments
584:frequency scaling
405:Technical factors
295:multi-chip module
6038:
6021:Flynn's taxonomy
5987:
5986:
5961:Software lockout
5760:Computer cluster
5695:Vector processor
5650:Array processing
5635:Flynn's taxonomy
5542:Memory coherence
5317:Computer network
5256:
5249:
5242:
5233:
5232:
5191:Processor design
5083:Power management
4965:Instruction unit
4826:Branch predictor
4775:
4774:
4473:System on a chip
4415:
4414:
4255:Transistor count
4179:Flynn's taxonomy
4036:
4035:
3894:
3893:
3697:Addressing modes
3608:
3607:
3554:Memory hierarchy
3418:Hypercomputation
3336:Abstract machine
3313:
3306:
3299:
3290:
3289:
3246:
3229:
3204:
3203:
3196:
3190:
3189:
3171:
3165:
3164:
3162:
3160:
3154:
3143:
3137:
3136:
3128:
3122:
3121:
3119:
3118:
3104:
3098:
3097:
3095:
3094:
3080:
3074:
3073:
3071:
3070:
3055:
3049:
3048:
3046:
3045:
3030:
3024:
3023:
3012:
3006:
3005:
2994:
2988:
2987:
2972:
2966:
2965:
2953:
2947:
2946:
2944:
2943:
2934:. Archived from
2927:
2921:
2920:
2918:
2917:
2903:
2897:
2896:
2894:
2893:
2879:
2873:
2872:
2870:
2869:
2855:
2849:
2848:
2846:
2845:
2831:
2825:
2824:
2822:
2821:
2807:
2801:
2800:
2798:
2797:
2783:
2777:
2776:
2774:
2773:
2759:
2753:
2752:
2750:
2748:
2733:
2727:
2726:
2724:
2723:
2709:
2703:
2702:
2700:
2699:
2685:
2679:
2678:
2676:
2674:
2655:
2649:
2648:
2641:
2635:
2634:
2632:
2631:
2615:
2609:
2608:
2606:
2604:
2578:
2572:
2571:
2569:
2549:
2543:
2542:
2537:. Archived from
2531:
2525:
2524:
2522:
2520:
2505:
2499:
2498:
2484:
2478:
2477:
2466:
2460:
2459:
2447:
2441:
2440:
2438:
2436:
2416:
2410:
2409:
2407:
2405:
2396:. Archived from
2389:
2383:
2380:
2374:
2373:
2371:
2369:
2354:
2348:
2347:
2345:
2343:
2327:
2321:
2320:
2300:
2294:
2293:
2285:
2279:
2278:
2276:
2275:
2260:
2254:
2253:
2251:
2250:
2236:
2230:
2229:
2227:
2226:
2212:
2206:
2205:
2185:
2179:
2178:
2176:
2174:
2158:
2152:
2151:
2149:
2148:
2134:
2128:
2127:
2125:
2124:
2109:
2100:
2099:
2097:
2091:. Archived from
2090:
2084:Schauer, Bryan.
2081:
2075:
2074:
2072:
2070:
2054:
2031:
2009:
1723:Sun Microsystems
1643:, an eight-core
1585:Charles H. Moore
1309:Hewlett-Packard
1291:hx3100 Processor
1056:Aeroflex Gaisler
1036:
1033:
1027:
999:
998:
991:
666:entropy encoding
651:Software effects
576:power management
541:operating system
351:microelectronics
282:system on a chip
122:are multi-core.
96:CPU instructions
6046:
6045:
6041:
6040:
6039:
6037:
6036:
6035:
6026:Microprocessors
6001:
6000:
5999:
5994:
5975:
5919:
5825:Coarray Fortran
5781:
5765:Beowulf cluster
5621:
5571:
5562:Synchronization
5547:Cache coherence
5537:Multiprocessing
5525:
5489:
5470:Cost efficiency
5465:Gustafson's law
5433:
5377:
5326:
5302:Multiprocessing
5292:Cloud computing
5265:
5260:
5230:
5225:
5211:Tick–tock model
5169:
5125:
5114:
5054:
5038:Address decoder
4992:
4946:
4942:Program counter
4917:Status register
4898:
4853:
4813:Load–store unit
4780:
4773:
4700:
4669:
4570:
4527:Image processor
4502:
4495:
4465:
4459:
4435:Microcontroller
4425:Embedded system
4413:
4313:
4246:
4235:
4173:
4123:
4021:
3998:
3982:Re-order buffer
3953:
3934:Data dependency
3920:
3879:
3709:
3703:
3602:
3601:Instruction set
3595:
3581:Multiprocessing
3549:Cache hierarchy
3542:Register/memory
3466:
3366:Queue automaton
3322:
3317:
3253:
3212:
3210:Further reading
3207:
3198:
3197:
3193:
3172:
3168:
3158:
3156:
3152:
3144:
3140:
3129:
3125:
3116:
3114:
3106:
3105:
3101:
3092:
3090:
3082:
3081:
3077:
3068:
3066:
3058:PDF, Download.
3056:
3052:
3043:
3041:
3032:
3031:
3027:
3014:
3013:
3009:
2996:
2995:
2991:
2974:
2973:
2969:
2954:
2950:
2941:
2939:
2928:
2924:
2915:
2913:
2905:
2904:
2900:
2891:
2889:
2881:
2880:
2876:
2867:
2865:
2857:
2856:
2852:
2843:
2841:
2833:
2832:
2828:
2819:
2817:
2809:
2808:
2804:
2795:
2793:
2785:
2784:
2780:
2771:
2769:
2761:
2760:
2756:
2746:
2744:
2735:
2734:
2730:
2721:
2719:
2711:
2710:
2706:
2697:
2695:
2687:
2686:
2682:
2672:
2670:
2657:
2656:
2652:
2643:
2642:
2638:
2629:
2627:
2618:
2616:
2612:
2602:
2600:
2579:
2575:
2550:
2546:
2533:
2532:
2528:
2518:
2516:
2506:
2502:
2486:
2485:
2481:
2468:
2467:
2463:
2448:
2444:
2434:
2432:
2417:
2413:
2403:
2401:
2390:
2386:
2381:
2377:
2367:
2365:
2356:
2355:
2351:
2341:
2339:
2328:
2324:
2317:
2301:
2297:
2286:
2282:
2273:
2271:
2261:
2257:
2248:
2246:
2238:
2237:
2233:
2224:
2222:
2214:
2213:
2209:
2186:
2182:
2172:
2170:
2169:on May 29, 2011
2159:
2155:
2146:
2144:
2136:
2135:
2131:
2122:
2120:
2110:
2103:
2095:
2088:
2082:
2078:
2068:
2066:
2055:
2051:
2047:
2003:
1998:
1952:Hyper-threading
1932:
1923:
1844:
1832:
1677:microcontroller
1645:microcontroller
1275:Cavium Networks
1186:Analog Devices
1078:, a multi-core
1061:, a multi-core
1046:
1037:
1031:
1028:
1017:
1000:
996:
989:
966:
938:
913:Cavium Networks
901:
878:embedded system
870:
851:SQL Server 2014
831:
653:
607:
567:"heterogeneous"
558:
553:
537:
493:cache coherency
489:
407:
378:
343:
262:
233:run in parallel
198:general-purpose
171:instruction set
135:message passing
127:multiprocessing
78:(CPUs), called
52:Athlon X2 6400+
24:
17:
12:
11:
5:
6044:
6034:
6033:
6028:
6023:
6018:
6013:
5996:
5995:
5993:
5992:
5980:
5977:
5976:
5974:
5973:
5968:
5963:
5958:
5956:Race condition
5953:
5948:
5943:
5938:
5933:
5927:
5925:
5921:
5920:
5918:
5917:
5912:
5907:
5902:
5897:
5892:
5887:
5882:
5877:
5872:
5867:
5862:
5857:
5852:
5847:
5842:
5837:
5832:
5827:
5822:
5817:
5812:
5807:
5802:
5797:
5791:
5789:
5783:
5782:
5780:
5779:
5774:
5769:
5768:
5767:
5757:
5751:
5750:
5749:
5744:
5739:
5734:
5729:
5724:
5714:
5713:
5712:
5707:
5700:Multiprocessor
5697:
5692:
5687:
5682:
5677:
5676:
5675:
5670:
5665:
5664:
5663:
5658:
5653:
5642:
5631:
5629:
5623:
5622:
5620:
5619:
5614:
5613:
5612:
5607:
5602:
5592:
5587:
5581:
5579:
5573:
5572:
5570:
5569:
5564:
5559:
5554:
5549:
5544:
5539:
5533:
5531:
5527:
5526:
5524:
5523:
5518:
5513:
5508:
5503:
5497:
5495:
5491:
5490:
5488:
5487:
5482:
5477:
5472:
5467:
5462:
5457:
5452:
5447:
5441:
5439:
5435:
5434:
5432:
5431:
5429:Hardware scout
5426:
5420:
5415:
5410:
5404:
5399:
5393:
5387:
5385:
5383:Multithreading
5379:
5378:
5376:
5375:
5370:
5365:
5360:
5355:
5350:
5345:
5340:
5334:
5332:
5328:
5327:
5325:
5324:
5322:Systolic array
5319:
5314:
5309:
5304:
5299:
5294:
5289:
5284:
5279:
5273:
5271:
5267:
5266:
5259:
5258:
5251:
5244:
5236:
5227:
5226:
5224:
5223:
5218:
5216:Pin grid array
5213:
5208:
5203:
5198:
5193:
5188:
5183:
5177:
5175:
5171:
5170:
5168:
5167:
5161:
5156:
5151:
5146:
5141:
5136:
5130:
5128:
5120:
5119:
5116:
5115:
5113:
5112:
5107:
5102:
5097:
5092:
5087:
5086:
5085:
5080:
5075:
5064:
5062:
5056:
5055:
5053:
5052:
5050:Barrel shifter
5047:
5046:
5045:
5040:
5033:Binary decoder
5030:
5029:
5028:
5018:
5013:
5008:
5002:
5000:
4994:
4993:
4991:
4990:
4985:
4977:
4972:
4967:
4962:
4956:
4954:
4948:
4947:
4945:
4944:
4939:
4934:
4929:
4924:
4922:Stack register
4919:
4914:
4908:
4906:
4900:
4899:
4897:
4896:
4895:
4894:
4889:
4879:
4874:
4869:
4863:
4861:
4855:
4854:
4852:
4851:
4846:
4845:
4844:
4833:
4828:
4823:
4822:
4821:
4815:
4804:
4798:
4792:
4785:
4783:
4772:
4771:
4766:
4761:
4756:
4751:
4750:
4749:
4744:
4739:
4734:
4729:
4724:
4714:
4708:
4706:
4702:
4701:
4699:
4698:
4693:
4688:
4683:
4677:
4675:
4671:
4670:
4668:
4667:
4666:
4665:
4655:
4650:
4645:
4640:
4635:
4630:
4625:
4620:
4615:
4610:
4605:
4600:
4595:
4590:
4584:
4582:
4576:
4575:
4572:
4571:
4569:
4568:
4563:
4558:
4553:
4547:
4541:
4535:
4529:
4524:
4518:
4516:AI accelerator
4513:
4507:
4505:
4497:
4496:
4494:
4493:
4487:
4482:
4479:Multiprocessor
4476:
4469:
4467:
4461:
4460:
4458:
4457:
4452:
4447:
4442:
4437:
4432:
4430:Microprocessor
4427:
4421:
4419:
4418:By application
4412:
4411:
4405:
4399:
4393:
4388:
4383:
4378:
4373:
4368:
4363:
4361:Tile processor
4358:
4353:
4348:
4343:
4342:
4341:
4330:
4323:
4321:
4315:
4314:
4312:
4311:
4306:
4301:
4295:
4289:
4283:
4277:
4271:
4270:
4269:
4257:
4251:
4249:
4241:
4240:
4237:
4236:
4234:
4233:
4232:
4231:
4221:
4216:
4215:
4214:
4209:
4204:
4199:
4189:
4183:
4181:
4175:
4174:
4172:
4171:
4166:
4161:
4156:
4155:
4154:
4149:
4147:Hyperthreading
4139:
4133:
4131:
4129:Multithreading
4125:
4124:
4122:
4121:
4116:
4111:
4110:
4109:
4099:
4098:
4097:
4092:
4082:
4081:
4080:
4075:
4065:
4060:
4059:
4058:
4053:
4042:
4040:
4033:
4027:
4026:
4023:
4022:
4020:
4019:
4014:
4008:
4006:
4000:
3999:
3997:
3996:
3991:
3986:
3985:
3984:
3979:
3969:
3963:
3961:
3955:
3954:
3952:
3951:
3946:
3941:
3936:
3930:
3928:
3922:
3921:
3919:
3918:
3913:
3908:
3906:Pipeline stall
3902:
3900:
3891:
3885:
3884:
3881:
3880:
3878:
3877:
3872:
3867:
3862:
3859:
3858:
3857:
3855:z/Architecture
3852:
3847:
3842:
3834:
3829:
3824:
3819:
3814:
3809:
3804:
3799:
3794:
3789:
3784:
3779:
3774:
3773:
3772:
3767:
3762:
3754:
3749:
3744:
3739:
3734:
3729:
3724:
3719:
3713:
3711:
3705:
3704:
3702:
3701:
3700:
3699:
3689:
3684:
3679:
3674:
3669:
3664:
3659:
3658:
3657:
3647:
3646:
3645:
3635:
3630:
3625:
3620:
3614:
3612:
3605:
3597:
3596:
3594:
3593:
3588:
3583:
3578:
3573:
3568:
3567:
3566:
3561:
3559:Virtual memory
3551:
3546:
3545:
3544:
3539:
3534:
3529:
3519:
3514:
3509:
3504:
3499:
3498:
3497:
3487:
3482:
3476:
3474:
3468:
3467:
3465:
3464:
3463:
3462:
3457:
3452:
3447:
3437:
3432:
3427:
3426:
3425:
3420:
3415:
3410:
3405:
3400:
3395:
3390:
3383:Turing machine
3380:
3379:
3378:
3373:
3368:
3363:
3358:
3353:
3343:
3338:
3332:
3330:
3324:
3323:
3316:
3315:
3308:
3301:
3293:
3287:
3286:
3280:
3269:
3260:
3252:
3251:External links
3249:
3248:
3247:
3230:
3211:
3208:
3206:
3205:
3191:
3166:
3138:
3123:
3099:
3075:
3050:
3040:. 2 April 2019
3025:
3007:
2989:
2967:
2948:
2922:
2898:
2874:
2850:
2826:
2802:
2778:
2754:
2728:
2704:
2680:
2650:
2636:
2610:
2573:
2567:10.1.1.37.4309
2544:
2541:on 2009-12-21.
2526:
2500:
2479:
2461:
2442:
2411:
2384:
2375:
2349:
2322:
2315:
2295:
2280:
2255:
2231:
2207:
2196:(2): 173–193.
2180:
2153:
2142:www.top500.org
2129:
2101:
2098:on 2011-11-25.
2076:
2048:
2046:
2043:
2042:
2041:
2026:
2017:CPU and a DSP
2002:
1999:
1997:
1996:
1991:
1989:Race condition
1986:
1980:
1975:
1969:
1964:
1959:
1954:
1949:
1944:
1939:
1933:
1931:
1928:
1922:
1919:
1918:
1917:
1911:
1905:
1895:
1885:
1884:
1883:
1877:
1861:
1851:
1843:
1840:
1839:
1838:
1831:
1828:
1827:
1826:
1819:
1818:
1817:
1811:
1800:
1799:
1798:
1795:
1784:
1783:
1782:
1775:Sunway SW26010
1769:
1768:
1767:
1761:
1755:
1749:
1743:
1737:
1731:
1720:
1694:
1693:
1692:
1684:
1678:
1675:ARM Cortex-M0+
1667:
1660:
1654:
1648:
1638:
1637:
1636:
1621:
1620:
1619:
1616:
1613:
1603:
1602:
1601:
1590:
1589:
1588:
1581:
1575:
1574:
1573:
1567:
1561:
1555:
1549:
1543:
1517:
1511:
1505:
1499:
1493:
1487:
1464:
1463:
1462:
1455:
1445:
1444:
1443:
1437:
1431:
1425:
1419:
1413:
1407:
1404:z/Architecture
1402:, a quad-core
1397:
1383:
1373:
1367:
1361:
1355:
1349:
1343:
1337:
1332:, a dual-core
1322:
1307:
1294:
1288:Coherent Logix
1285:
1272:
1269:
1268:
1267:
1259:
1245:
1244:
1243:
1236:
1230:
1229:
1228:
1225:
1222:
1214:
1208:
1191:
1184:
1183:
1182:
1168:
1162:
1156:
1150:
1144:
1138:
1132:
1126:
1120:
1114:
1104:
1093:
1083:
1070:
1053:
1045:
1042:
1039:
1038:
1003:
1001:
994:
988:
985:
983:architecture.
981:ARM big.LITTLE
965:
962:
937:
934:
900:
897:
869:
866:
865:
864:
858:
830:
827:
811:
810:
806:
802:
801:
797:
793:
792:
788:
784:
783:
779:
652:
649:
620:True Octa-core
606:
603:
590:computers and
557:
554:
552:
549:
536:
533:
525:front-side bus
499:(alternative:
488:
485:
472:
471:
470:
469:
446:
435:
421:
416:Additionally:
406:
403:
401:architecture.
377:
374:
342:
339:
261:
258:
218:supercomputers
191:multithreading
100:multithreading
68:microprocessor
15:
9:
6:
4:
3:
2:
6043:
6032:
6029:
6027:
6024:
6022:
6019:
6017:
6014:
6012:
6009:
6008:
6006:
5991:
5982:
5981:
5978:
5972:
5969:
5967:
5964:
5962:
5959:
5957:
5954:
5952:
5949:
5947:
5944:
5942:
5939:
5937:
5934:
5932:
5929:
5928:
5926:
5922:
5916:
5913:
5911:
5908:
5906:
5903:
5901:
5898:
5896:
5893:
5891:
5888:
5886:
5883:
5881:
5878:
5876:
5873:
5871:
5868:
5866:
5863:
5861:
5858:
5856:
5853:
5851:
5848:
5846:
5845:Global Arrays
5843:
5841:
5838:
5836:
5833:
5831:
5828:
5826:
5823:
5821:
5818:
5816:
5813:
5811:
5808:
5806:
5803:
5801:
5798:
5796:
5793:
5792:
5790:
5788:
5784:
5778:
5775:
5773:
5772:Grid computer
5770:
5766:
5763:
5762:
5761:
5758:
5755:
5752:
5748:
5745:
5743:
5740:
5738:
5735:
5733:
5730:
5728:
5725:
5723:
5720:
5719:
5718:
5715:
5711:
5708:
5706:
5703:
5702:
5701:
5698:
5696:
5693:
5691:
5688:
5686:
5683:
5681:
5678:
5674:
5671:
5669:
5666:
5662:
5659:
5657:
5654:
5651:
5648:
5647:
5646:
5643:
5641:
5638:
5637:
5636:
5633:
5632:
5630:
5628:
5624:
5618:
5615:
5611:
5608:
5606:
5603:
5601:
5598:
5597:
5596:
5593:
5591:
5588:
5586:
5583:
5582:
5580:
5578:
5574:
5568:
5565:
5563:
5560:
5558:
5555:
5553:
5550:
5548:
5545:
5543:
5540:
5538:
5535:
5534:
5532:
5528:
5522:
5519:
5517:
5514:
5512:
5509:
5507:
5504:
5502:
5499:
5498:
5496:
5492:
5486:
5483:
5481:
5478:
5476:
5473:
5471:
5468:
5466:
5463:
5461:
5458:
5456:
5453:
5451:
5448:
5446:
5443:
5442:
5440:
5436:
5430:
5427:
5424:
5421:
5419:
5416:
5414:
5411:
5408:
5405:
5403:
5400:
5397:
5394:
5392:
5389:
5388:
5386:
5384:
5380:
5374:
5371:
5369:
5366:
5364:
5361:
5359:
5356:
5354:
5351:
5349:
5346:
5344:
5341:
5339:
5336:
5335:
5333:
5329:
5323:
5320:
5318:
5315:
5313:
5310:
5308:
5305:
5303:
5300:
5298:
5295:
5293:
5290:
5288:
5285:
5283:
5280:
5278:
5275:
5274:
5272:
5268:
5264:
5257:
5252:
5250:
5245:
5243:
5238:
5237:
5234:
5222:
5219:
5217:
5214:
5212:
5209:
5207:
5204:
5202:
5199:
5197:
5194:
5192:
5189:
5187:
5184:
5182:
5179:
5178:
5176:
5172:
5165:
5162:
5160:
5157:
5155:
5152:
5150:
5147:
5145:
5142:
5140:
5137:
5135:
5132:
5131:
5129:
5127:
5121:
5111:
5108:
5106:
5103:
5101:
5098:
5096:
5093:
5091:
5088:
5084:
5081:
5079:
5076:
5074:
5071:
5070:
5069:
5066:
5065:
5063:
5061:
5057:
5051:
5048:
5044:
5041:
5039:
5036:
5035:
5034:
5031:
5027:
5024:
5023:
5022:
5019:
5017:
5014:
5012:
5011:Demultiplexer
5009:
5007:
5004:
5003:
5001:
4999:
4995:
4989:
4986:
4984:
4981:
4978:
4976:
4973:
4971:
4968:
4966:
4963:
4961:
4958:
4957:
4955:
4953:
4949:
4943:
4940:
4938:
4935:
4933:
4932:Memory buffer
4930:
4928:
4927:Register file
4925:
4923:
4920:
4918:
4915:
4913:
4910:
4909:
4907:
4905:
4901:
4893:
4890:
4888:
4885:
4884:
4883:
4880:
4878:
4875:
4873:
4870:
4868:
4867:Combinational
4865:
4864:
4862:
4860:
4856:
4850:
4847:
4843:
4840:
4839:
4837:
4834:
4832:
4829:
4827:
4824:
4819:
4816:
4814:
4811:
4810:
4808:
4805:
4802:
4799:
4796:
4793:
4790:
4787:
4786:
4784:
4782:
4776:
4770:
4767:
4765:
4762:
4760:
4757:
4755:
4752:
4748:
4745:
4743:
4740:
4738:
4735:
4733:
4730:
4728:
4725:
4723:
4720:
4719:
4718:
4715:
4713:
4710:
4709:
4707:
4703:
4697:
4694:
4692:
4689:
4687:
4684:
4682:
4679:
4678:
4676:
4672:
4664:
4661:
4660:
4659:
4656:
4654:
4651:
4649:
4646:
4644:
4641:
4639:
4636:
4634:
4631:
4629:
4626:
4624:
4621:
4619:
4616:
4614:
4611:
4609:
4606:
4604:
4601:
4599:
4596:
4594:
4591:
4589:
4586:
4585:
4583:
4581:
4577:
4567:
4564:
4562:
4559:
4557:
4554:
4551:
4548:
4545:
4542:
4539:
4536:
4533:
4530:
4528:
4525:
4522:
4519:
4517:
4514:
4512:
4509:
4508:
4506:
4504:
4498:
4491:
4488:
4486:
4483:
4480:
4477:
4474:
4471:
4470:
4468:
4462:
4456:
4453:
4451:
4448:
4446:
4443:
4441:
4438:
4436:
4433:
4431:
4428:
4426:
4423:
4422:
4420:
4416:
4409:
4406:
4403:
4400:
4397:
4394:
4392:
4389:
4387:
4384:
4382:
4379:
4377:
4374:
4372:
4369:
4367:
4364:
4362:
4359:
4357:
4354:
4352:
4349:
4347:
4344:
4340:
4337:
4336:
4334:
4331:
4328:
4325:
4324:
4322:
4320:
4316:
4310:
4307:
4305:
4302:
4299:
4296:
4293:
4290:
4287:
4284:
4281:
4278:
4275:
4272:
4267:
4264:
4263:
4261:
4258:
4256:
4253:
4252:
4250:
4248:
4242:
4230:
4227:
4226:
4225:
4222:
4220:
4217:
4213:
4210:
4208:
4205:
4203:
4200:
4198:
4195:
4194:
4193:
4190:
4188:
4185:
4184:
4182:
4180:
4176:
4170:
4167:
4165:
4162:
4160:
4157:
4153:
4150:
4148:
4145:
4144:
4143:
4140:
4138:
4135:
4134:
4132:
4130:
4126:
4120:
4117:
4115:
4112:
4108:
4105:
4104:
4103:
4100:
4096:
4093:
4091:
4088:
4087:
4086:
4083:
4079:
4076:
4074:
4071:
4070:
4069:
4066:
4064:
4061:
4057:
4054:
4052:
4049:
4048:
4047:
4044:
4043:
4041:
4037:
4034:
4032:
4028:
4018:
4015:
4013:
4010:
4009:
4007:
4005:
4001:
3995:
3992:
3990:
3987:
3983:
3980:
3978:
3975:
3974:
3973:
3970:
3968:
3967:Scoreboarding
3965:
3964:
3962:
3960:
3956:
3950:
3949:False sharing
3947:
3945:
3942:
3940:
3937:
3935:
3932:
3931:
3929:
3927:
3923:
3917:
3914:
3912:
3909:
3907:
3904:
3903:
3901:
3899:
3895:
3892:
3890:
3886:
3876:
3873:
3871:
3868:
3866:
3863:
3860:
3856:
3853:
3851:
3848:
3846:
3843:
3841:
3838:
3837:
3835:
3833:
3830:
3828:
3825:
3823:
3820:
3818:
3815:
3813:
3810:
3808:
3805:
3803:
3800:
3798:
3795:
3793:
3790:
3788:
3785:
3783:
3780:
3778:
3775:
3771:
3768:
3766:
3763:
3761:
3758:
3757:
3755:
3753:
3750:
3748:
3745:
3743:
3742:Stanford MIPS
3740:
3738:
3735:
3733:
3730:
3728:
3725:
3723:
3720:
3718:
3715:
3714:
3712:
3706:
3698:
3695:
3694:
3693:
3690:
3688:
3685:
3683:
3680:
3678:
3675:
3673:
3670:
3668:
3665:
3663:
3660:
3656:
3653:
3652:
3651:
3648:
3644:
3641:
3640:
3639:
3636:
3634:
3631:
3629:
3626:
3624:
3621:
3619:
3616:
3615:
3613:
3609:
3606:
3604:
3603:architectures
3598:
3592:
3589:
3587:
3584:
3582:
3579:
3577:
3574:
3572:
3571:Heterogeneous
3569:
3565:
3562:
3560:
3557:
3556:
3555:
3552:
3550:
3547:
3543:
3540:
3538:
3535:
3533:
3530:
3528:
3525:
3524:
3523:
3522:Memory access
3520:
3518:
3515:
3513:
3510:
3508:
3505:
3503:
3500:
3496:
3493:
3492:
3491:
3488:
3486:
3483:
3481:
3478:
3477:
3475:
3473:
3469:
3461:
3458:
3456:
3455:Random-access
3453:
3451:
3448:
3446:
3443:
3442:
3441:
3438:
3436:
3435:Stack machine
3433:
3431:
3428:
3424:
3421:
3419:
3416:
3414:
3411:
3409:
3406:
3404:
3401:
3399:
3396:
3394:
3391:
3389:
3386:
3385:
3384:
3381:
3377:
3374:
3372:
3369:
3367:
3364:
3362:
3359:
3357:
3354:
3352:
3351:with datapath
3349:
3348:
3347:
3344:
3342:
3339:
3337:
3334:
3333:
3331:
3329:
3325:
3321:
3314:
3309:
3307:
3302:
3300:
3295:
3294:
3291:
3284:
3281:
3279:
3278:
3277:IEEE Spectrum
3273:
3270:
3268:
3264:
3261:
3258:
3255:
3254:
3244:
3240:
3236:
3231:
3227:
3223:
3219:
3214:
3213:
3201:
3195:
3187:
3183:
3182:
3177:
3170:
3151:
3150:
3142:
3134:
3127:
3113:
3112:ark.intel.com
3109:
3103:
3089:
3088:ark.intel.com
3085:
3079:
3065:
3061:
3054:
3039:
3035:
3029:
3021:
3017:
3011:
3003:
2999:
2993:
2985:
2981:
2977:
2971:
2963:
2959:
2952:
2938:on 2006-11-09
2937:
2933:
2926:
2912:
2911:ark.intel.com
2908:
2902:
2888:
2887:ark.intel.com
2884:
2878:
2864:
2863:ark.intel.com
2860:
2854:
2840:
2839:ark.intel.com
2836:
2830:
2816:
2815:ark.intel.com
2812:
2806:
2792:
2791:ark.intel.com
2788:
2782:
2768:
2767:ark.intel.com
2764:
2758:
2742:
2738:
2732:
2718:
2717:ark.intel.com
2714:
2708:
2694:
2693:ark.intel.com
2690:
2684:
2668:
2664:
2660:
2654:
2646:
2640:
2625:
2621:
2614:
2598:
2594:
2590:
2589:
2584:
2577:
2568:
2563:
2559:
2555:
2548:
2540:
2536:
2530:
2515:
2511:
2504:
2497:
2493:
2489:
2483:
2475:
2471:
2465:
2457:
2453:
2446:
2430:
2426:
2422:
2419:Clark, Jack.
2415:
2399:
2395:
2388:
2379:
2363:
2359:
2353:
2337:
2333:
2326:
2318:
2312:
2308:
2307:
2299:
2291:
2284:
2270:
2266:
2259:
2245:
2241:
2235:
2221:
2217:
2211:
2203:
2199:
2195:
2191:
2184:
2168:
2164:
2157:
2143:
2139:
2133:
2119:
2115:
2112:Smith, Ryan.
2108:
2106:
2094:
2087:
2080:
2064:
2060:
2053:
2049:
2039:
2035:
2032:Two types of
2030:
2027:
2024:
2023:mobile phones
2020:
2016:
2012:
2008:
2005:
2004:
1995:
1992:
1990:
1987:
1984:
1981:
1979:
1976:
1973:
1970:
1968:
1965:
1963:
1960:
1958:
1955:
1953:
1950:
1948:
1945:
1943:
1940:
1938:
1937:CPU shielding
1935:
1934:
1927:
1915:
1912:
1909:
1906:
1903:
1899:
1896:
1893:
1889:
1886:
1882:
1878:
1876:
1872:
1871:
1869:
1865:
1862:
1859:
1855:
1852:
1849:
1846:
1845:
1837:
1834:
1833:
1824:
1820:
1815:
1812:
1809:
1806:
1805:
1804:
1801:
1796:
1793:
1792:TMS320C80 MVP
1790:
1789:
1788:
1785:
1780:
1776:
1773:
1772:
1770:
1765:
1762:
1759:
1756:
1753:
1752:UltraSPARC T3
1750:
1747:
1746:UltraSPARC T2
1744:
1741:
1740:UltraSPARC T1
1738:
1735:
1734:UltraSPARC IV
1732:
1729:
1726:
1725:
1724:
1721:
1718:
1717:PlayStation 3
1714:
1710:
1706:
1702:
1698:
1695:
1690:
1689:
1688:
1685:
1682:
1679:
1676:
1672:
1668:
1665:
1661:
1658:
1655:
1652:
1649:
1646:
1642:
1639:
1634:
1630:
1627:
1626:
1625:
1622:
1617:
1614:
1611:
1607:
1606:
1604:
1599:
1596:
1595:
1594:
1591:
1586:
1582:
1579:
1578:
1576:
1571:
1568:
1565:
1562:
1559:
1556:
1553:
1550:
1547:
1544:
1541:
1537:
1533:
1529:
1525:
1521:
1518:
1515:
1512:
1509:
1506:
1503:
1500:
1497:
1494:
1491:
1488:
1485:
1481:
1477:
1473:
1470:
1469:
1468:
1465:
1460:
1456:
1454:
1451:
1450:
1449:
1446:
1441:
1438:
1435:
1432:
1429:
1426:
1423:
1420:
1417:
1414:
1411:
1408:
1405:
1401:
1398:
1396:game console.
1395:
1391:
1387:
1384:
1381:
1377:
1374:
1371:
1368:
1365:
1362:
1359:
1356:
1353:
1350:
1347:
1344:
1341:
1338:
1335:
1331:
1328:
1327:
1326:
1323:
1320:
1316:
1312:
1308:
1305:
1302:
1298:
1295:
1292:
1289:
1286:
1283:
1280:
1276:
1273:
1270:
1265:
1264:
1263:
1260:
1258:
1254:
1250:
1246:
1241:
1237:
1234:
1233:
1231:
1226:
1223:
1220:
1219:
1218:
1215:
1212:
1209:
1206:
1202:
1198:
1195:
1192:
1189:
1185:
1180:
1176:
1172:
1169:
1166:
1163:
1160:
1157:
1154:
1151:
1148:
1145:
1142:
1139:
1136:
1133:
1130:
1127:
1124:
1121:
1118:
1115:
1112:
1108:
1105:
1102:
1099:
1098:
1097:
1094:
1091:
1087:
1084:
1081:
1077:
1074:
1071:
1068:
1064:
1060:
1057:
1054:
1051:
1048:
1047:
1035:
1025:
1021:
1015:
1013:
1009:
1004:This article
1002:
993:
992:
984:
982:
977:
975:
971:
961:
959:
955:
951:
947:
943:
933:
931:
927:
922:
918:
914:
910:
906:
896:
894:
889:
886:
879:
874:
862:
859:
856:
852:
848:
844:
840:
836:
835:
834:
826:
824:
820:
816:
807:
804:
803:
798:
796:Agglomeration
795:
794:
789:
787:Communication
786:
785:
780:
777:
776:
775:
773:
768:
765:
761:
757:
753:
749:
743:
741:
737:
733:
729:
725:
721:
717:
713:
709:
705:
701:
697:
693:
689:
687:
683:
677:
673:
671:
667:
663:
662:thread-safety
658:
647:
645:
641:
640:Anant Agarwal
637:
635:
630:
626:
623:
621:
615:
613:
612:heterogeneous
602:
600:
595:
593:
589:
585:
581:
577:
572:
568:
564:
548:
544:
542:
535:Disadvantages
532:
528:
526:
521:
517:
512:
510:
506:
502:
498:
494:
484:
481:
477:
467:
463:
459:
455:
451:
447:
444:
440:
436:
433:
429:
428:
426:
422:
419:
418:
417:
414:
412:
402:
400:
396:
390:
388:
384:
373:
371:
370:
365:
362:
358:
357:
352:
348:
347:semiconductor
338:
336:
332:
328:
323:
321:
317:
316:
310:
308:
303:
301:
297:
296:
291:
287:
283:
279:
275:
271:
267:
257:
253:
249:
247:
242:
238:
234:
230:
225:
223:
219:
215:
211:
207:
203:
199:
194:
192:
188:
184:
180:
176:
172:
168:
164:
163:heterogeneous
160:
156:
152:
148:
144:
140:
139:shared-memory
136:
132:
128:
123:
121:
117:
113:
110:, known as a
109:
105:
101:
97:
93:
89:
85:
81:
77:
73:
69:
65:
61:
53:
48:
41:
36:
28:
22:
5530:Coordination
5460:Amdahl's law
5396:Simultaneous
5221:Chip carrier
5159:Clock gating
5078:Mixed-signal
4975:Write buffer
4952:Control unit
4764:Clock signal
4685:
4503:accelerators
4485:Cypress PSoC
4142:Simultaneous
3959:Out-of-order
3591:Neuromorphic
3472:Architecture
3430:Belt machine
3423:Zeno machine
3356:Hierarchical
3275:
3266:
3234:
3226:10657.1/2440
3217:
3194:
3179:
3169:
3157:. Retrieved
3148:
3141:
3126:
3115:. Retrieved
3111:
3102:
3091:. Retrieved
3087:
3078:
3067:. Retrieved
3063:
3053:
3042:. Retrieved
3037:
3028:
3010:
2992:
2970:
2951:
2940:. Retrieved
2936:the original
2925:
2914:. Retrieved
2910:
2901:
2890:. Retrieved
2886:
2877:
2866:. Retrieved
2862:
2853:
2842:. Retrieved
2838:
2829:
2818:. Retrieved
2814:
2805:
2794:. Retrieved
2790:
2781:
2770:. Retrieved
2766:
2757:
2745:. Retrieved
2741:the original
2731:
2720:. Retrieved
2716:
2707:
2696:. Retrieved
2692:
2683:
2671:. Retrieved
2662:
2653:
2639:
2628:. Retrieved
2613:
2601:. Retrieved
2588:Ars Technica
2586:
2576:
2560:(1): 13–23.
2557:
2553:
2547:
2539:the original
2529:
2517:. Retrieved
2503:
2495:
2492:the original
2482:
2474:the original
2464:
2455:
2445:
2433:. Retrieved
2424:
2414:
2402:. Retrieved
2398:the original
2387:
2378:
2366:. Retrieved
2352:
2340:. Retrieved
2325:
2305:
2298:
2283:
2272:. Retrieved
2268:
2258:
2247:. Retrieved
2243:
2234:
2223:. Retrieved
2219:
2210:
2193:
2189:
2183:
2171:. Retrieved
2167:the original
2156:
2145:. Retrieved
2141:
2132:
2121:. Retrieved
2117:
2093:the original
2079:
2067:. Retrieved
2063:the original
2052:
2029:
2007:
1967:Multitasking
1924:
1459:home gateway
1380:Power Mac G5
1317:, dual core
1240:Raspberry Pi
1217:Azul Systems
1201:ARM11 MPCore
1111:Athlon 64 X2
1107:Athlon 64 FX
1029:
1018:Please help
1006:may contain
1005:
978:
967:
939:
902:
890:
888:processors.
883:
877:
832:
812:
778:Partitioning
769:
744:
738:, and IBM's
714:, Skandium,
690:
678:
674:
670:video codecs
654:
638:
631:
628:
624:
619:
616:
608:
605:Architecture
596:
578:and dynamic
559:
545:
538:
529:
513:
501:Bus snooping
490:
473:
465:
461:
457:
449:
438:
431:
415:
408:
391:
379:
367:
354:
344:
324:
319:
313:
311:
306:
304:
299:
293:
285:
269:
265:
263:
254:
250:
237:Amdahl's law
226:
195:
124:
116:chip package
111:
87:
83:
79:
70:on a single
63:
59:
57:
5966:Scalability
5727:distributed
5610:Concurrency
5577:Programming
5418:Cooperative
5407:Speculative
5343:Instruction
5006:Multiplexer
4970:Data buffer
4681:Single-core
4653:bit slicing
4511:Coprocessor
4366:Coprocessor
4247:performance
4169:Cooperative
4159:Speculative
4119:Distributed
4078:Superscalar
4063:Instruction
4031:Parallelism
4004:Speculative
3836:System/3x0
3708:Instruction
3485:Von Neumann
3398:Post–Turing
3155:. Hot Chips
2747:12 November
2519:October 21,
2292:. WikiChip.
1577:IntellaSys
1540:Core 2 Quad
1514:Core 2 Quad
1376:PowerPC 970
1321:processors.
815:server side
809:scheduling.
772:concurrency
497:cache snoop
462:memory wall
432:memory wall
387:Clock rates
361:superscalar
341:Development
260:Terminology
246:refactoring
212:(DSP), and
183:superscalar
6005:Categories
5971:Starvation
5710:asymmetric
5445:PRAM model
5413:Preemptive
5126:management
5021:Multiplier
4882:Logic gate
4872:Sequential
4779:Functional
4759:Clock rate
4732:Data cache
4705:Components
4686:Multi-core
4674:Core count
4164:Preemptive
4068:Pipelining
4051:Bit-serial
3994:Wide-issue
3939:Structural
3861:Tilera ISA
3827:MicroBlaze
3797:ETRAX CRIS
3692:Comparison
3537:Load–store
3517:Endianness
3259:—MakeUseOf
3117:2019-05-04
3093:2019-05-04
3069:2019-05-04
3044:2019-05-04
2942:2006-09-28
2916:2019-05-04
2892:2019-05-04
2868:2019-05-04
2844:2019-05-04
2820:2019-05-04
2796:2019-05-04
2772:2019-05-04
2722:2019-05-04
2698:2019-05-04
2630:2014-03-04
2603:5 December
2593:Condé Nast
2274:2024-05-22
2249:2023-10-27
2225:2023-10-27
2147:2020-09-15
2123:2020-09-15
2045:References
1921:Benchmarks
1892:Wavescalar
1856:, 16-core
1612:processor.
1536:Core 2 Duo
1508:Core 2 Duo
1461:processor.
1262:ClearSpeed
1175:FireStream
1044:Commercial
1012:irrelevant
823:throughput
571:big.LITTLE
487:Advantages
458:power wall
450:power wall
364:pipelining
312:The terms
280:(DSP) and
266:multi-core
264:The terms
167:big.LITTLE
40:Core 2 Duo
5705:symmetric
5450:PEM model
5060:Circuitry
4980:Microcode
4904:Registers
4747:coherence
4722:CPU cache
4580:Word size
4245:Processor
3889:Execution
3792:DEC Alpha
3770:Power ISA
3586:Cognitive
3393:Universal
3018:. Intel.
3000:. Intel.
2617:Compare:
2562:CiteSeerX
1904:processor
1894:processor
1860:processor
1836:OpenSPARC
1673:, a dual
1657:Plurality
1301:Power ISA
1249:Tensilica
1232:Broadcom
1141:Phenom II
1123:FX-Series
1117:Athlon II
1032:July 2016
1008:excessive
950:Freescale
839:Microsoft
829:Licensing
770:Managing
730:, Cray's
700:Cilk Plus
682:microcode
454:shrinking
315:many-core
307:multi-CPU
270:dual-core
102:or other
88:quad-core
84:dual-core
38:An Intel
5936:Deadlock
5924:Problems
5890:pthreads
5870:OpenHMPP
5795:Ateji PX
5756:computer
5627:Hardware
5494:Elements
5480:Slowdown
5391:Temporal
5373:Pipeline
4998:Datapath
4691:Manycore
4663:variable
4501:Hardware
4137:Temporal
3817:OpenRISC
3512:Cellular
3502:Dataflow
3495:modified
3186:Archived
3181:PC World
3038:TechSpot
3020:Archived
3002:Archived
2984:Archived
2962:Archived
2667:Archived
2624:Archived
2597:Archived
2514:EE Times
2435:6 August
2429:Archived
2404:March 6,
2368:March 6,
2362:Archived
2336:Archived
2269:PC Gamer
2173:March 6,
2069:March 6,
1957:Manycore
1930:See also
1914:UC Davis
1848:Stanford
1842:Academic
1764:SPARC T5
1758:SPARC T4
1681:SiCortex
1664:Kilocore
1662:Rapport
1651:picoChip
1629:RTX 3090
1598:MPPA-256
1570:Xeon Phi
1538:and the
1502:Core Duo
1476:netbooks
1448:Infineon
1394:Xbox 360
1188:Blackfin
1101:A-Series
1050:Adapteva
1014:examples
958:Picochip
921:Broadcom
917:Wintegra
782:problem.
736:Fortress
734:, Sun's
712:FastFlow
708:OpenHMPP
698:such as
686:picocode
614:" role.
599:manycore
551:Hardware
466:ILP wall
439:ILP wall
229:software
222:one case
214:graphics
202:embedded
173:, while
159:crossbar
5895:RaftLib
5875:OpenACC
5850:GPUOpen
5840:C++ AMP
5815:Charm++
5557:Barrier
5501:Process
5485:Speedup
5270:General
5174:Related
5105:Quantum
5095:Digital
5090:Boolean
4988:Counter
4887:Quantum
4648:512-bit
4643:256-bit
4638:128-bit
4481:(MPSoC)
4466:on chip
4464:Systems
4282:(FLOPS)
4095:Process
3944:Control
3926:Hazards
3812:Itanium
3807:Unicore
3765:PowerPC
3490:Harvard
3450:Pointer
3445:Counter
3403:Quantum
3159:27 June
2456:Giz Bot
2342:May 17,
1870:(AsAP)
1814:TILE-Gx
1771:Sunway
1713:PowerPC
1705:Toshiba
1552:Pentium
1546:Itanium
1532:Core i9
1528:Core i7
1524:Core i5
1520:Core i3
1496:Celeron
1486:(MIDs).
1480:nettops
1370:Power10
1334:PowerPC
1319:PA-RISC
1315:PA-8900
1311:PA-8800
1242:models)
1147:Sempron
1129:Opteron
819:threads
805:Mapping
756:Fortran
634:Renesas
580:voltage
509:degrade
505:signals
349:-based
206:network
66:) is a
50:An AMD
5988:
5865:OpenCL
5860:OpenMP
5805:Chapel
5722:shared
5717:Memory
5652:(SIMT)
5595:Models
5506:Thread
5438:Theory
5409:(SpMT)
5363:Memory
5348:Thread
5331:Levels
5110:Switch
5100:Analog
4838:(IMC)
4809:(MMU)
4658:others
4633:64-bit
4628:48-bit
4623:32-bit
4618:24-bit
4613:16-bit
4608:15-bit
4603:12-bit
4440:Mobile
4356:Stream
4351:Barrel
4346:Vector
4335:(GPU)
4294:(SUPS)
4262:(IPC)
4114:Memory
4107:Vector
4090:Thread
4073:Scalar
3875:Others
3822:RISC-V
3787:SuperH
3756:Power
3752:MIPS-X
3727:PDP-11
3576:Fabric
3328:Models
2564:
2313:
1994:Thread
1985:(PGAS)
1972:OpenCL
1808:TILE64
1803:Tilera
1687:SiFive
1671:RP2040
1624:Nvidia
1610:MIPS64
1593:Kalray
1364:POWER9
1358:POWER8
1352:POWER7
1346:POWER6
1340:POWER5
1330:POWER4
1197:MPCore
1171:Radeon
1153:Turion
1135:Phenom
1092:(MPPA)
1086:Ambric
974:Xilinx
853:, and
843:socket
732:Chapel
720:Erlang
718:, and
704:OpenMP
644:Tilera
588:laptop
586:(i.e.
556:Trends
187:vector
157:, and
131:caches
5835:Dryad
5800:Boost
5521:Array
5511:Fiber
5425:(CMT)
5398:(SMT)
5312:GPGPU
5166:(PPW)
5124:Power
5016:Adder
4892:Array
4859:Logic
4820:(TLB)
4803:(FPU)
4797:(AGU)
4791:(ALU)
4781:units
4717:Cache
4598:8-bit
4593:4-bit
4588:1-bit
4552:(TPU)
4546:(DSP)
4540:(PPU)
4534:(VPU)
4523:(GPU)
4492:(NoC)
4475:(SoC)
4410:(PoP)
4404:(SiP)
4398:(MCM)
4339:GPGPU
4329:(CPU)
4319:Types
4300:(PPW)
4288:(TPS)
4276:(IPS)
4268:(CPI)
4039:Level
3850:S/390
3845:S/370
3840:S/360
3782:SPARC
3760:POWER
3643:TRIPS
3611:Types
3153:(PDF)
3064:Intel
2980:Intel
2673:5 May
2425:ZDNet
2244:ZDNET
2220:PCMAG
2096:(PDF)
2089:(PDF)
2001:Notes
1947:GPGPU
1902:TRIPS
1881:AsAP2
1821:XMOS
1633:GPGPU
1467:Intel
1453:AURIX
1440:Telum
1416:zEC12
1386:Xenon
1257:ESP32
1211:ASOCS
1179:GPGPU
1159:Ryzen
1076:PhysX
1073:Ageia
1063:SPARC
1059:LEON3
930:6WIND
476:Intel
395:Intel
189:, or
80:cores
5900:ROCm
5830:CUDA
5820:Cilk
5787:APIs
5747:COMA
5742:NUMA
5673:MIMD
5668:MISD
5645:SIMD
5640:SISD
5368:Loop
5358:Data
5353:Task
5144:ACPI
4877:Glue
4769:FIFO
4712:Core
4450:ASIP
4391:CPLD
4386:FPOA
4381:FPGA
4376:ASIC
4229:SPMD
4224:MIMD
4219:MISD
4212:SWAR
4192:SIMD
4187:SISD
4102:Data
4085:Task
4056:Word
3802:M32R
3747:MIPS
3710:sets
3677:ZISC
3672:NISC
3667:OISC
3662:MISC
3655:EPIC
3650:VLIW
3638:EDGE
3628:RISC
3623:CISC
3532:HUMA
3527:NUMA
3161:2023
2749:2007
2675:2019
2605:2015
2521:2023
2437:2015
2406:2013
2370:2013
2344:2015
2311:ISBN
2175:2013
2071:2013
2015:RISC
1942:CUDA
1875:AsAP
1830:Free
1728:MAJC
1709:Cell
1697:Sony
1564:Xeon
1530:and
1472:Atom
1410:z196
1313:and
1279:MIPS
1203:and
1177:GPU/
1173:and
1165:Epyc
1109:and
919:and
764:ACML
754:and
582:and
478:and
464:and
448:The
437:The
430:The
331:FPGA
318:and
300:same
286:same
268:and
179:VLIW
155:mesh
151:ring
5915:ZPL
5910:TBB
5905:UPC
5885:PVM
5855:MPI
5810:HPX
5737:UMA
5338:Bit
5139:APM
5134:PMU
5026:CPU
4983:ROM
4754:Bus
4371:PAL
4046:Bit
3832:LMC
3737:ARM
3732:x86
3722:VAX
3239:doi
3222:hdl
2663:AMD
2198:doi
2019:MPU
1858:RAW
1854:MIT
1707:'s
1701:IBM
1434:z15
1428:z14
1422:z13
1400:z10
1390:SMT
1325:IBM
1304:MPU
1282:MPU
1255:'s
1194:ARM
1096:AMD
1010:or
968:In
940:In
876:An
740:X10
724:TBB
716:MPI
684:or
657:GUI
594:).
520:SMP
480:AMD
399:x86
290:die
147:bus
137:or
108:die
86:or
64:MCP
6007::
5073:3D
3184:.
3178:.
3110:.
3086:.
3062:.
3036:.
2982:.
2978:.
2960:.
2909:.
2885:.
2861:.
2837:.
2813:.
2789:.
2765:.
2715:.
2691:.
2665:.
2661:.
2595:.
2591:.
2585:.
2556:.
2512:.
2454:.
2427:.
2423:.
2334:.
2267:.
2242:.
2218:.
2194:21
2192:.
2140:.
2116:.
2104:^
1900:,
1890:,
1866:,
1526:,
1522:,
1478:,
915:,
911:,
849:,
825:.
760:C#
742:.
710:,
706:,
702:,
688:.
248:.
208:,
204:,
200:,
193:.
185:,
181:,
149:,
58:A
5255:e
5248:t
5241:v
3312:e
3305:t
3298:v
3274:—
3265:—
3245:.
3241::
3228:.
3224::
3202:.
3163:.
3135:.
3120:.
3096:.
3072:.
3047:.
2945:.
2919:.
2895:.
2871:.
2847:.
2823:.
2799:.
2775:.
2751:.
2725:.
2701:.
2677:.
2647:.
2633:.
2607:.
2570:.
2558:4
2523:.
2458:.
2439:.
2408:.
2372:.
2346:.
2319:.
2277:.
2252:.
2228:.
2204:.
2200::
2177:.
2150:.
2126:.
2073:.
1781:.
1719:.
1703:/
1699:/
1647:.
1587:.
1542:.
1382:.
1306:.
1284:.
1181:.
1082:.
1069:.
1034:)
1030:(
1026:.
1016:.
752:C
610:"
468:.
62:(
23:.
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.