Field-programmable gate array: Difference between revisions

Revision as of 04:03, 15 July 2025

Latest revision as of 19:07, 25 August 2025

(6 intermediate revisions by 6 users not shown)

Line 12:

}}

A '''field-programmable gate array''' ('''FPGA''') is a type of configurable [[integrated circuit]] that can be repeatedly programmed after manufacturing. FPGAs are a subset of logic devices referred to as [[programmable logic devices]] (PLDs). They consist of ana grid-connected array of [[programmable logic device|programmable]] [[logic block|logic blocks]] with a connecting grid, that can be configured "in the field" to interconnect with other logic blocks to perform various digital functions. FPGAs are often used in limited (low) quantity production of custom-made products, and in research and development, where the higher cost of individual FPGAs is not as important, and where creating and manufacturing a custom circuit would not be feasible. Other applications for FPGAs include the telecommunications, automotive, aerospace, and industrial sectors, which benefit from their flexibility, high signal processing speed, and parallel processing abilities.

A FPGA configuration is generally written using a [[hardware description language]] (HDL) e.g. [[VHDL]], similar to the ones used for [[application-specific integrated circuit]]s (ASICs). [[Circuit diagram|Circuit diagrams]] were formerly used to write the configuration.

Line 40:

Companies like Microsoft have started to use FPGAs to accelerate high-performance, computationally intensive systems (like the [[data center]]s that operate their [[Bing search engine]]), due to the [[performance per watt]] advantage FPGAs deliver.<ref>{{cite magazine|url=https://www.wired.com/2014/06/microsoft-fpga/|title=Microsoft Supercharges Bing Search With Programmable Chips|date=16 June 2014|magazine=WIRED}}</ref> Microsoft began using FPGAs to [[Hardware acceleration|accelerate]] Bing in 2014, and in 2018 began deploying FPGAs across other data center workloads for their [[Microsoft Azure|Azure]] [[cloud computing]] platform.<ref name="ProjCatapult" />

Since 2019, modern generation of FPGAs have been integrated with other architectures like [[AI engine|AI engines]] to target workloads in artificial intelligence ___domain.<ref>{{Cite journalbook |lastlast1=Gaide |firstfirst1=Brian |last2=Gaitonde |first2=Dinesh |last3=Ravishankar |first3=Chirag |last4=Bauer |first4=Trevor |chapter=Xilinx Adaptive Compute Acceleration Platform: Versal ^TM Architecture |date=2019-02-20 |title=XilinxProceedings Adaptiveof Computethe Acceleration2019 Platform:ACM/SIGDA VersalInternational TMSymposium Architectureon Field-Programmable Gate Arrays |chapter-url=https://dl.acm.org/doi/10.1145/3289602.3293906 |journal=ACM/SIGDA International Symposium on Field-Programmable Gate Arrays |language=en |publisher=ACM |pages=84–93 |doi=10.1145/3289602.3293906 |isbn=978-1-4503-6137-8|chapter-url-access=subscription }}</ref>.

===Growth===

Line 81:

"An application circuit must be mapped into an FPGA with adequate resources. While the number of logic blocks and I/Os required is easily determined from the design, the number of routing channels needed may vary considerably even among designs with the same amount of logic. For example, a [[crossbar switch]] requires much more routing than a [[systolic array]] with the same gate count. Since unused routing channels increase the cost (and decrease the performance) of the FPGA without providing any benefit, FPGA manufacturers try to provide just enough channels so that most designs that will fit in terms of [[Lookup table#Hardware LUTs|lookup tables]] (LUTs) and I/Os can be [[Routing (electronic design automation)|routed]]. This is determined by estimates such as those derived from [[Rent's rule]] or by experiments with existing designs."<ref>{{Cite journal |last1=M.b |first1=Swami |last2=V.p |first2=Pawar |date=2014-07-31 |title=VLSI DESIGN: A NEW APPROACH |url=https://bioinfopublication.org/pages/article.php?id=BIA0002301 |journal=Journal of Intelligence Systems |language=En |volume=4 |issue=1 |pages=60–63 |issn=2229-7057}}</ref>

In general, a logic block consists of a few logical cells. A typical cell consists of a 4-input LUT, a [[Adder (electronics)|full adder]] (FA) and a [[D-type flip-flop]]. The LUT might be split into two 3-input LUTs. In ''normal mode'' those are combined into a 4-input LUT through the first [[multiplexer]] (mux). In ''arithmetic'' mode, their outputs are fed to the adder. The selection of mode is programmed into the second mux. The output can be either [[Synchronous circuit|synchronous]] or [[Asynchronous circuit|asynchronous]], depending on the programming of the third mux. In practice, the entire adder or parts of it are [[Shannon expansion|stored as functions]] into the LUTs in order to save [[Circuit utilization|space]].<ref>[http://www.altera.com/literature/hb/cyc2/cyc2_cii51002.pdf 2. CycloneII Architecture] {{Webarchive|url=https://web.archive.org/web/20101214055643/http://www.altera.com/literature/hb/cyc2/cyc2_cii51002.pdf |date=2010-12-14 }}. Altera. February 2007</ref><ref>{{cite web |url=http://www.altera.com/literature/hb/stratix-iv/stx4_5v1_01.pdf |title=Documentation: Stratix IV Devices |publisher=Altera.com |date=2008-06-11 |access-date=2013-05-01 |archive-url=https://web.archive.org/web/20110926214034/http://www.altera.com/literature/hb/stratix-iv/stx4_5v1_01.pdf |archive-date=2011-09-26 |url-status=dead}}</ref><ref>[http://www.xilinx.com/support/documentation/user_guides/ug070.pdf Virtex-4 FPGA User Guide] (December 1st, 2008). Xilinx, Inc.</ref>

=== Hard blocks ===

Line 153:

** [[SiliconBlue Technologies]] provides extremely low-power SRAM-based FPGAs with optional integrated [[Non-volatile memory|nonvolatile]] configuration memory; acquired by Lattice in 2011

* [[Microchip Technology|Microchip]]:

** [[Microsemi]] (previously [[Actel]]), producing [[antifuse]], flash-based, [[mixed-signal]] FPGAs; acquired by Microchip in 2018

** [[Atmel]], a second source of some Altera-compatible devices; also FPSLIC{{Clarify|reason=|date=December 2018}} mentioned above;<ref>{{Cite news|url=http://sourcetech411.com/2013/04/top-fpga-companies-for-2013/|title=Top FPGA Companies For 2013|date=2013-04-28|work=SourceTech411|access-date=2018-12-01|language=en-US|archive-date=2018-08-24|archive-url=https://web.archive.org/web/20180824135219/https://sourcetech411.com/2013/04/top-fpga-companies-for-2013/|url-status=dead}}</ref> acquired by Microchip in 2016

* QuickLogic manufactures ultra-low-power sensor hubs, extremely-low-powered, low-density SRAM-based FPGAs, with display bridges MIPI and RGB inputs; MIPI, RGB and LVDS outputs.<ref>{{Cite web|url=http://www.quicklogic.com/|title=QuickLogic — Customizable Semiconductor Solutions for Mobile Devices|website=www.quicklogic.com|publisher=QuickLogic Corporation|language=en|access-date=2018-10-07}}{{better source needed|date=September 2024}}</ref>

Line 173:

* Space (with [[radiation hardening]]<ref>{{Cite web|url=https://www.militaryaerospace.com/articles/2016/06/radiation-hardened-space-fpga.html|title=FPGA development devices for radiation-hardened space applications introduced by Microsemi|website=www.militaryaerospace.com|access-date=2018-11-02|date=2016-06-03}}</ref>)

* [[Hardware security module]]s<ref name="auto">{{cite web|title=CrypTech: Building Transparency into Cryptography t |url=https://cryptech.is/wp-content/uploads/2016/02/CrypTech_Building_Transparency.pdf |archive-url=https://web.archive.org/web/20160807180252/https://cryptech.is/wp-content/uploads/2016/02/CrypTech_Building_Transparency.pdf |archive-date=2016-08-07 |url-status=live}}</ref>

* High-speed financial transactions<ref>{{Cite web |last=Mann |first=Tobias |date=2023-03-08 |title=While Intel XPUs are delayed, here's some more FPGAs to tide you over |url=https://www.theregister.com/2023/03/08/intel_fpga_agilex/ |website=The Register}}</ref><ref>{{Cite conference |url=https://ieeexplore.ieee.org/document/6044837 |title=High Frequency Trading Acceleration Using FPGAs |last1=Leber |first1=Christian |last2=Geib |first2=Benjamin |last3=Litz |first3=Heiner |doi=10.1109/FPL.2011.64 |publisher=IEEE |date=September 2011 |conference=International Conference on Field Programmable Logic and Applications|url-access=subscription }}</ref>

* [[Retrocomputing]] (e.g. the MARS and [[MiSTer]] FPGA projects)<ref>{{cite web |url=https://www.retrorgb.com/the-diy-mister-handheld.html |title=The DIY MiSTer Handheld |date=16 December 2024 |access-date=}}</ref>

* Large scale integrated [[digital differential analyzer]]s, a form of an [[analog computer]] based on digital computing elements<ref>[https://people.ece.cornell.edu/land/courses/ece5760/DDA/index.htm DDA on FPGA - A modern Analog Computer]</ref>

Line 181:

== Security ==

Concerning [[hardware security]], FPGAs have both advantages and disadvantages as compared to ASICs or secure microprocessors. FPGAs' flexibility makes malicious modifications during [[Semiconductor device fabrication|fabrication]] a lower risk.<ref name="paper">{{Cite journal |doi=10.1109/MDT.2008.166 |title=Managing Security in FPGA-Based Embedded Systems |journal=IEEE Design & Test of Computers |volume=25 |issue=6 |pages=590–598 |year=2008 |last1=Huffmire |first1=Ted |last2=Brotherton |first2=Brett |last3=Sherwood |first3=Timothy |last4=Kastner |first4=Ryan |last5=Levin |first5=Timothy |last6=Nguyen |first6=Thuy D. |last7=Irvine |first7=Cynthia|bibcode=2008IDTC...25..590H |s2cid=115840 |hdl=10945/7159 |hdl-access=free }}</ref> Previously, for many FPGAs, the design [[bitstream]] was exposed while the FPGA loads it from external memory, typically during powerup. All major FPGA vendors now offer a spectrum of security solutions to designers such as bitstream [[encryption]] and [[authentication]]. For example, [[Altera]] and [[Xilinx]] offer [[Advanced Encryption Standard|AES]] encryption (up to 256-bit) for bitstreams stored in an external flash memory. [[Physical unclonable function]]s (PUFs) are integrated circuits that have their own unique signatures and can be used to secure FPGAs while taking up very little hardware space.<ref>{{Cite journal |last1=Babaei |first1=Armin |last2=Schiele |first2=Gregor |last3=Zohner |first3=Michael |date=2022-07-26 |title=Reconfigurable Security Architecture (RESA) Based on PUF for FPGA-Based IoT Devices |journal=Sensors |language=en |volume=22 |issue=15 |pages=5577 |doi=10.3390/s22155577 |issn=1424-8220 |pmc=9331300 |pmid=35898079 |bibcode=2022Senso..22.5577B |doi-access=free }}</ref>

FPGAs that store their configuration internally in nonvolatile flash memory, such as [[Microsemi]]'s ProAsic&nbsp;3 or [[Lattice Semiconductor|Lattice]]'s XP2 programmable devices, do not expose the bitstream and do not need [[encryption]]. Customers wanting a higher guarantee of tamper resistance can use write-once, antifuse FPGAs from vendors such as [[Microsemi]].

Line 189:

In 2012 researchers Sergei Skorobogatov and Christopher Woods demonstrated that some FPGAs can be vulnerable to hostile intent. They discovered a critical [[Backdoor (computing)|backdoor vulnerability]] had been manufactured in silicon as part of the Actel/Microsemi ProAsic&nbsp;3 making it vulnerable on many levels such as reprogramming crypto and [[access key]]s, accessing unencrypted bitstream, modifying [[low-level]] silicon features, and extracting [[Computer configuration|configuration]] data.<ref>{{cite book |volume=7428|pages=23–40|doi=10.1007/978-3-642-33027-8_2|series = Lecture Notes in Computer Science|year = 2012|last1 = Skorobogatov|first1 = Sergei|title=Cryptographic Hardware and Embedded Systems – CHES 2012|last2=Woods|first2=Christopher|isbn=978-3-642-33026-1|chapter=Breakthrough Silicon Scanning Discovers Backdoor in Military Chip}}</ref>

In 2020 a critical vulnerability (named Starbleed) was discovered in all Xilinx&nbsp;7 series FPGAs that rendered bitstream encryption useless. There is no workaround. Xilinx did not produce a hardware revision. Ultrascale and later devices, already on the market at the time, were not affected.{{cn|date=July 2025}}<!--[[User:Kvng/RTH]]-->

== Similar technologies ==

Historically, FPGAs have been slower, less energy efficient and generally achieved less functionality than their fixed ASIC counterparts. A study from 2006 showed that designs implemented on FPGAs need on average 40 times as much area, draw 12 times as much dynamic power, and run at one third the speed of corresponding ASIC implementations.<ref name="FPGA-ASIC-comparison">{{cite conference|doi=10.1145/1117201.1117205|chapter=Measuring the gap between FPGAs and ASICs|title=Proceedings of the international symposium on Field programmable gate arrays – FPGA'06|pages=21–30|year=2006|last1=Kuon|first1=Ian|last2=Rose|first2=Jonathan|isbn=1-59593-292-5|publisher=ACM|___location=New York, NY|chapter-url=http://ece.gmu.edu/coursewebpages/ECE/ECE448/S09/viewgraphs/Gap_between_FPGAs_and_ASICs.pdf|conference=|access-date=2017-10-25|archive-date=2010-06-22|archive-url=https://web.archive.org/web/20100622170541/http://ece.gmu.edu/coursewebpages/ECE/ECE448/S09/viewgraphs/Gap_between_FPGAs_and_ASICs.pdf|url-status=dead}}</ref>

Advantages of FPGAs include the ability to re-program whenreprogram alreadyequipment deployed (i.e. "in the field") to fix [[Bug (computer programming)|bugs]], andor oftenmake other improvements. Some FPGAs have the capability of [[partial re-configuration]] that lets one portion of the device be re-programmed while other portions continue running.<ref>{{Cite web|url=https://www.intel.com/content/www/us/en/programmable/documentation/ekx1496870149834.html|title=AN 818: Static Update Partial Reconfiguration Tutorial: for Intel Stratix 10 GX FPGA Development Board|website=www.intel.com|access-date=2018-12-01}}</ref><ref>{{Cite web|url=https://electronics.stackexchange.com/questions/45115/can-fpgas-dynamically-modify-their-logic|title=Can FPGAs dynamically modify their logic?|website=Electrical Engineering Stack Exchange|access-date=2018-12-01}}</ref> Other advantages may include shorter [[time to market]] and lower [[non-recurring engineering]] costs. Vendors can also take a middle road via [[FPGA prototyping]]: developing their prototype hardware on FPGAs, but manufacturemanufacturing their final version as an ASIC so that it can no longer be modified after the design has been committed. This is often also the case with new processor designs.<ref>{{cite web|url=https://www.anandtech.com/show/14798/xilinx-announces-world-largest-fpga-virtex-ultrascale-vu19p-with-9m-cells|archive-url=https://web.archive.org/web/20190827160514/https://www.anandtech.com/show/14798/xilinx-announces-world-largest-fpga-virtex-ultrascale-vu19p-with-9m-cells|url-status=dead|archive-date=August 27, 2019|title=Xilinx Announces World Largest FPGA: Virtex Ultrascale+ VU19P with 9m Cells|first=Ian|last=Cutress|date=August 27, 2019|website=[[AnandTech]]}}</ref> Some FPGAs have the capability of [[partial re-configuration]] that lets one portion of the device be re-programmed while other portions continue running.<ref>{{Cite web|url=https://www.intel.com/content/www/us/en/programmable/documentation/ekx1496870149834.html|title=AN 818: Static Update Partial Reconfiguration Tutorial: for Intel Stratix 10 GX FPGA Development Board|website=www.intel.com|access!-date=2018-12-01}}</ref><ref>{{Cite web|url=https[[User:Kvng//electronics.stackexchange.com/questions/45115/canRTH]]-fpgas-dynamically-modify-their-logic|title=Can FPGAs dynamically modify their logic?|website=Electrical Engineering Stack Exchange|access-date=2018-12-01}}</ref>

The primary differences between [[complex programmable logic device]]s (CPLDs) and FPGAs are [[Computer architecture|architectural]]. A CPLD has a comparatively restrictive structure consisting of one or more programmable [[Canonical normal form|sum-of-products]] logic arrays feeding a relatively small number of clocked [[Register (computing)|registers]]. As a result, CPLDs are less flexible but have the advantage of more predictable [[Latency (engineering)|timing delays]] and {{Citation needed span|text=a higher logic-to-interconnect ratio.|date=December 2018|reason=}} FPGA architectures, on the other hand, are dominated by [[Communications subsystem|interconnect]]. This makes them far more flexible (in terms of the range of designs that are practical for implementation on them) but also far more complex to design for, or at least requiring more complex [[electronic design automation]] (EDA) software. In practice, the distinction between FPGAs and CPLDs is often one of size as FPGAs are usually much larger in terms of resources than CPLDs. Typically only FPGAs contain more complex [[Functional unit|embedded functions]] such as [[Adder (electronics)|adders]], [[Binary multiplier|multipliers]], [[Computer memory|memory]], and [[SerDes|serializer/deserializers]]. Another common distinction is that CPLDs contain embedded [[flash memory]] to store their configuration while FPGAs usually require external [[non-volatile memory]] (but not always). When a design requires simple instant-on [[glue logic|(logic is already configured at power-up)]] CPLDs are generally preferred. For most other applications FPGAs are generally preferred. Sometimes both CPLDs and FPGAs are used in a single system design. In those designs, CPLDs generally perform glue logic functions and are responsible for "[[booting]]" the FPGA as well as controlling [[Reset (computing)|reset]] and boot sequence of the complete circuit board. Therefore, depending on the application it may be judicious to use both FPGAs and CPLDs in a single design.<ref>{{cite web|url=https://numato.com/kb/cpld-vs-fpga-differences-one-use/|title=CPLD vs FPGA: Differences between them and which one to use? – Numato Lab Help Center|website=numato.com|date=2017-11-29}}</ref>

Line 211 ⟶ 210:

* {{cite book |first1=Hartmut F.-W. |last1=Sadrozinski |first2=Jinyuan |last2=Wu |title=Applications of Field-Programmable Gate Arrays in Scientific Research |year=2010 |publisher=Taylor & Francis |isbn=978-1-4398-4133-4}}

* {{cite book|title=Digital Circuit Design An Introduction Textbook |first=Niklaus |last=Wirth |publisher=Springer |year=1995 |isbn=978-3-540-58577-0}}

* {{cite journal|title=An FPGA-Based Phase Measurement System |journal=IEEE Transactions on Very Large Scale Integration (VLSI) Systems |volume=26 |pages=133–142 |first=Jubin |last=Mitra |publisher=IEEE |year=2018 |issue=1 |doi=10.1109/TVLSI.2017.2758807|bibcode=2018ITVL...26..133M |s2cid=4920719 |doi-access=free }}

* Mencer, Oskar et al. (2020). "The history, status, and future of FPGAs". Communications of the ACM. ACM. Vol. 63, No. 10. [[doi:10.1145/3410669]]