Meta tags:

Headings (most frequently used words):

maxwell, generation, chips, microarchitecture, contents, first, gm10x, second, gm20x, performance, successor, see, also, references,

Text of the page (most frequently used words):
the (90), maxwell (52), nvidia (48), and (38), geforce (35), gtx (26), gpus (22), #generation (19), from (17), was (14), gpu (14), memory (13), 2014 (13), for (13), kepler (12), edit (12), units (12), also (12), archived (11), video (11), cuda (11), power (10), first (10), graphics (9), 2015 (9), 750 (9), with (8), decoding (8), original (8), series (8), speed (8), used (8), microarchitecture (7), wikipedia (7), this (7), retrieved (7), smith (7), ryan (7), 980 (7), chips (7), support (7), performance (7), bit (7), second (7), pascal (6), multi (6), review (6), architecture (6), based (6), february (6), move (6), per (6), that (6), gm107 (6), smm (6), shared (6), code (5), page (5), technologies (5), products (5), nvenc (5), 970 (5), september (5), new (5), efficiency (5), successor (5), which (5), compared (5), controllers (5), bus (5), both (5), gm20x (5), gm10x (5), introduced (5), were (5), smx (5), toggle (4), contents (4), search (4), microarchitectures (4), tegra (4), shield (4), software (4), quadro (4), other (4), tools (4), encoding (4), cores (4), march (4), has (4), 2025 (4), processing (4), low (4), core (4), each (4), gddr5 (4), read (4), can (4), segment (4), one (4), supported (4), gm108 (4), are (4), set (4), more (4), hide (4), sidebar (4), view (3), about (3), use (3), copy (3), title (3), rtx (3), switch (3), reality (3), cards (3), super (3), resolution (3), system (3), rate (3), api (3), ray (3), tracing (3), purevideo (3), 800m (3), 700 (3), unified (3), shaders (3), pixel (3), 128 (3), list (3), native (3), rop (3), allocation (3), gm204 (3), mark (3), not (3), www (3), com (3), anti (3), aliasing (3), direct (3), pdf (3), dynamic (3), tile (3), makes (3), its (3), hardware (3), features (3), precision (3), cache (3), but (3), upgraded (3), supports (3), 60fps (3), compute (3), capability (3), gk110 (3), gk208 (3), gpc (3), however (3), reducing (3), times (3), design (3), execution (3), released (3), chip (3), numbers (3), history (3), main (3), languages (2), table (2), mobile (2), contact (2), privacy (2), policy (2), additional (2), terms (2), may (2), you (2), categories (2), cs1 (2), maint (2), short (2), description (2), wikidata (2), articles (2), david (2), jetson (2), nintendo (2), tesla (2), protocol (2), nvlink (2), module (2), socket (2), nvdec (2), acceleration (2), 900 (2), 256 (2), riva (2), official (2), blog (2), updates (2), fp64 (2), december (2), anandtech (2), titan (2), january (2), crossbar (2), most (2), tech (2), here (2), than (2), csaa (2), frame (2), sampled (2), hdmi (2), hevc (2), gaming (2), link (2), cite (2), 2017 (2), rendering (2), 2016 (2), real (2), ganesh (2), get (2), october (2), references (2), see (2), bandwidth (2), theoretical (2), single (2), being (2), operations (2), number (2), clock (2), gm200 (2), after (2), consumer (2), rops (2), cost (2), high (2), time (2), because (2), unit (2), write (2), simultaneous (2), peak (2), still (2), only (2), executing (2), operation (2), while (2), have (2), 264 (2), gk10x (2), version (2), improved (2), tessellation (2), some (2), sampling (2), feature (2), die (2), eight (2), claims (2), increases (2), when (2), atomic (2), streaming (2), multiprocessor (2), partitioned (2), warp (2), schedulers (2), special (2), had (2), resources (2), 192 (2), 745 (2), these (2), increased (2), card (2), until (2), appearance (2), upload (2), file (2), changes (2), links (2), article (2), subsection (2), log (2), create (2), account (2), donate (2), menu (2), add, topic, cookie, statement, statistics, developers, conduct, legal, safety, contacts, disclaimers, text, available, under, apply, using, site, agree, registered, trademark, non, profit, organization, wikimedia, foundation, inc, creative, commons, attribution, sharealike, license, rendered, parsoid, last, edited, july, 2026, utc, hidden, different, https, org, index, php, maxwell_, oldid, 1365695556, stexar, mediaq, exluna, portalplayer, mental, images, mellanox, icera, deepmap, cumulus, networks, bright, computing, uli, ageia, 3dfx, interactive, acquisitions, jonah, alben, ranga, jayaraman, debora, shoquist, colette, kress, bill, dally, kirk, curtis, priem, chris, malachowsky, jensen, huang, key, people, company, nforce, computer, chipsets, project, denver, cpus, spark, drive, goforce, embedded, socs, now, android, tablet, portable, t239, drake, erista, playstation, rsx, synthesizer, xbox, nv2a, components, console, dgx, gpgpu, plex, workstation, feynman, rubin, blackwell, ada, lovelace, hopper, ampere, turing, volta, fermi, curie, rankine, kelvin, celsius, live, upscaling, framebuffer, turbocache, sli, sxm, mxm, monitor, surround, switching, optimus, variable, refresh, sync, stereo, vision, decode, vdpau, platform, omniverse, physics, sdk, physx, optix, gameworks, shading, language, multimedia, tensor, 600, numa, 500, 400, 300, 200, 100, vertex, tnt2, tnt, nv2, nv1, pre, fixed, pipeline, roadmap, announces, puny, just, diving, deeper, partitions, correcting, specs, exploring, geoffrey, tim, lazygamer, net, gtx970, rather, serious, bug, parallel, forall, advanced, ever, made, report, another, reason, slower, introducing, laptops, fast, delivers, better, gamers, display, matters, how, brings, virtual, closer, guides, news, featured, stories, batteryboost, ditch, brick, web, improves, your, games, quality, monitors, 2020, kanter, august, april, world, rasterization, designed, energy, things, should, know, kampman, jeffrey, tom, confirms, end, game, ready, driver, affected, will, optimized, drivers, through, eponyms, codenamed, higher, double, been, noted, very, previous, computed, fma, instruction, cycle, flops, gm206, complaints, revealed, able, disable, individual, containing, 256kb, without, disabling, whole, comes, dividing, into, segments, cannot, accessed, same, reads, managing, shares, return, channel, data, between, reading, writing, impossible, therefore, described, having, 224, 512, such, attained, figure, reachable, adds, resolutions, 1440p, 1080p, polymorph, engine, responsible, resulting, controller, ratio, changed, generally, idle, there, enough, enabled, smms, give, them, work, maximum, fill, several, third, delta, color, compression, programming, vxgi, voxel, projection, mfaa, coverage, removed, direct3d12, level, 12_1, added, global, illumination, shot, found, inside, since, fully, uefi, output, thought, they, actually, tiled, caching, encoder, faster, meaning, encode, six, playback, ten, increase, due, decoder, paired, full, relying, mix, cpu, state, gc5, conserve, 265, parallelism, hyperq, two, across, entire, product, line, provides, integers, compare, swap, cas, implement, functions, retooled, renamed, structure, scheduler, inherited, texture, layout, controls, fp32, load, store, function, contrast, where, scheduled, pool, latter, necessitated, wide, unnecessary, allow, all, conversely, modular, allows, finer, grained, efficient, saving, workload, isn, optimal, factor, cluster, contains, 850m, 860m, 830m, 840m, few, facing, instead, focused, increasing, kib, mib, need, accordingly, reduced, gk106, area, draw, soc, application, processor, named, discoverer, classical, electromagnetism, james, clerk, sixth, seventh, palit, branded, oem, earlier, gk1xx, kxxx, followed, 960, june, final, lowest, spec, gtx950, aug, 20th, codename, developed, later, models, mxxx, well, limited, november, security, 2028, status, predecessor, tsmc, fabrication, process, release, date, free, encyclopedia, item, projects, printable, download, print, export, legacy, parser, shortened, url, information, permanent, related, what, general, actions, english, talk, українська, русский, norsk, bokmål, 한국어, français, suomi, فارسی, español, català, top, personal, pages, recent, community, portal, learn, help, contribute, random, current, events, navigation, jump, content,

Text of the page (random words):
maxwell microarchitecture wikipedia jump to content main menu main menu move to sidebar hide navigation main page contents current events random article about wikipedia contact us contribute help learn to edit community portal recent changes upload file special pages search search appearance donate create account log in personal tools donate create account log in contents move to sidebar hide top 1 first generation maxwell gm10x toggle first generation maxwell gm10x subsection 1 1 chips 2 second generation maxwell gm20x toggle second generation maxwell gm20x subsection 2 1 chips 3 performance 4 successor 5 see also 6 references toggle the table of contents maxwell microarchitecture 10 languages català español فارسی suomi français 한국어 norsk bokmål русский українська 中文 edit links article talk english read edit view history tools tools move to sidebar hide actions read edit view history general what links here related changes upload file permanent link page information cite this page get shortened url switch to legacy parser print export download as pdf printable version in other projects wikidata item appearance move to sidebar hide from wikipedia the free encyclopedia gpu microarchitecture by nvidia nvidia maxwell release date february 2014 fabrication process tsmc 28 nm 20 nm 16 nm history predecessor kepler successor pascal support status limited support until november 2025 security updates until october 2028 1 maxwell is the codename for a gpu microarchitecture developed by nvidia as the successor to the kepler microarchitecture the maxwell architecture was introduced in later models of the geforce 700 series and is also used in the geforce 800m series geforce 900 series and quadro mxxx series as well as some jetson products the first maxwell based products were the geforce gtx 745 oem geforce gtx 750 and the geforce gtx 750 ti both were released on february 18 2014 both with the chip code number gm107 earlier geforce 700 series gpus had used kepler chips with the code numbers gk1xx first generation maxwell gpus code numbers gm10x are also used in the geforce 800m series and the quadro kxxx series a second generation of maxwell based products was introduced on september 18 2014 with the geforce gtx 970 and geforce gtx 980 followed by the geforce gtx 960 on january 22 2015 the geforce gtx titan x on march 17 2015 and the geforce gtx 980 ti on june 1 2015 the final and lowest spec maxwell 2 0 card was the gtx950 released on aug 20th 2015 these gpus have gm20x chip code numbers a palit branded gtx 750 graphics card maxwell introduced an improved streaming multiprocessor sm design that increased power efficiency 2 the sixth and seventh generation purevideo hd and cuda compute capability 5 2 the architecture is named after james clerk maxwell the discoverer of classical electromagnetism the maxwell architecture is used in the system on a chip soc mobile application processor tegra x1 first generation maxwell gm10x edit first generation maxwell gpus gm107 gm108 were released as geforce gtx 745 gtx 750 750 ti gtx 850m 860m gm107 and geforce 830m 840m gm108 these new chips introduced few consumer facing additional features as nvidia instead focused more on increasing gpu power efficiency the l2 cache was increased from 256 kib on kepler to 2 mib on maxwell reducing the need for more memory bandwidth accordingly the memory bus was reduced from 192 bit on kepler gk106 to 128 bit reducing die area cost and power draw 3 the smx streaming multiprocessor design from kepler was also retooled and partitioned being renamed smm for maxwell the structure of the warp scheduler was inherited from kepler with the texture units and fp64 cuda cores still shared but the layout of most execution units were partitioned so that each warp schedulers in an smm controls one set of 32 fp32 cuda cores one set of 8 load store units and one set of 8 special function units this is in contrast to kepler where each smx had 4 schedulers that scheduled to a shared pool of execution units 4 the latter necessitated an smx wide crossbar that used unnecessary power to allow all execution units to be shared 4 conversely maxwell s more modular design allows for a finer grained and more efficient allocation of resources saving power when the workload isn t optimal for shared resources nvidia claims a 128 cuda core smm has 90 of the performance of a 192 cuda core smx while efficiency increases by a factor of 2 3 also each graphics processing cluster or gpc contains up to 4 smx units in kepler and up to 5 smm units in first generation maxwell 3 gm107 also supports cuda compute capability 5 0 compared to 3 5 on gk110 gk208 gpus and 3 0 on gk10x gpus dynamic parallelism and hyperq two features in gk110 gk208 gpus are also supported across the entire maxwell product line maxwell also provides native shared memory atomic operations for 32 bit integers and native shared memory 32 bit and 64 bit compare and swap cas which can be used to implement other atomic functions nvidia s video encoder nvenc was upgraded to be 1 5 to 2 times faster than on kepler based gpus meaning it can encode video at six to eight times playback speed 3 nvidia also claims an eight to ten times performance increase in purevideo feature set e video decoding due to the video decoder cache paired with increases in memory efficiency however h 265 is not supported for full hardware decoding in first generation maxwell gpus relying on a mix of hardware decoding and software decoding cpu decoding 3 when decoding video a new low power state gc5 is used on maxwell gpus to conserve power 3 maxwell gpus were thought to use tile based rendering 5 but they actually use tiled caching 6 since first generation maxwell uefi graphics output protocol is fully supported on nvidia gpus chips edit gm107 gm108 second generation maxwell gm20x edit die shot of the gm200 gpu found inside geforce gtx 980 ti cards second generation maxwell gpus introduced several new technologies dynamic super resolution 7 third generation delta color compression 8 multi pixel programming sampling 9 nvidia vxgi real time voxel global illumination 10 vr direct 10 11 12 multi projection acceleration 8 multi frame sampled anti aliasing mfaa 13 however support for coverage sampling anti aliasing csaa was removed 14 and direct3d12 api at feature level 12_1 hdmi 2 0 support was also added 15 16 the rop to memory controller ratio was changed from 8 1 to 16 1 17 however some of the rops are generally idle in the gtx 970 because there are not enough enabled smms to give them work to do reducing its maximum fill rate 18 the polymorph engine responsible for tessellation was upgraded to version 3 0 in second generation maxwell gpus resulting in improved tessellation performance per unit clock second generation maxwell also has up to 4 smm units per gpc compared to 5 smm units per gpc 17 gm204 supports cuda compute capability 5 2 compared to 5 0 on gm107 gm108 gpus 3 5 on gk110 gk208 gpus and 3 0 on gk10x gpus 8 17 19 gm20x gpus have an upgraded nvenc which supports hevc encoding and adds support for h 264 encoding resolutions at 1440p 60fps 4k 60fps compared to nvenc on maxwell first generation gm10x gpus which only supported h 264 1080p 60fps encoding 12 after consumer complaints 20 nvidia revealed that it is able to disable individual units each containing 256kb of l2 cache and 8 rops without disabling whole memory controllers 21 this comes at the cost of dividing the memory bus into high speed and low speed segments that cannot be accessed at the same time for reads because the l2 rop unit managing both of the gddr5 controllers shares the read return channel and the write data bus between the gddr5 controllers this makes simultaneous reading from both gddr5 controllers or simultaneous writing to both gddr5 controllers impossible 21 this is used in the geforce gtx 970 which therefore can be described as having 3 5 gb in a high speed segment on a 224 bit bus and 512 mb in a low speed segment on a 32 bit bus 21 the peak speed of such a gpu can still be attained but the peak speed figure is only reachable if one segment is executing a read operation while the other segment is executing a write operation 21 chips edit gm200 gm204 gm206 performance edit the theoretical single precision processing power of a maxwell gpu in flops is computed as 2 operations per fma instruction per cuda core per cycle number of cuda cores core clock speed in hz the theoretical double precision processing power of a maxwell gpu is 1 32 of the single precision performance which has been noted as being very low compared to the previous generation kepler 22 successor edit the successor to maxwell is codenamed pascal 23 the pascal architecture features higher bandwidth unified memory and nvlink 23 see also edit list of eponyms of nvidia gpu microarchitectures list of nvidia graphics processing units nvidia nvdec nvidia nvenc references edit kampman jeffrey 2025 07 31 nvidia confirms end of game ready driver support for maxwell and pascal gpus affected products will get optimized drivers through october 2025 tom s hardware retrieved 2025 08 21 5 things you should know about the new maxwell gpu architecture 2014 02 21 1 2 3 4 5 6 smith ryan t s ganesh 18 february 2014 the nvidia geforce gtx 750 ti and gtx 750 review maxwell makes its move anandtech archived from the original on 18 february 2014 retrieved 18 february 2014 1 2 ryan smith ganesh t s maxwell designed for energy efficiency the nvidia geforce gtx 750 ti and gtx 750 review maxwell makes its move archived from the original on february 21 2014 kanter david august 1 2016 tile based rasterization in nvidia gpus real world technologies retrieved april 1 2016 on nvidia s tile based rendering tech power up march 1 2017 retrieved may 9 2020 dynamic super resolution improves your games with 4k quality graphics on hd monitors www nvidia com 1 2 3 archived copy pdf archived from the original pdf on 2017 07 21 retrieved 2014 09 19 cite web cs1 maint archived copy as title link nvidia batteryboost ditch the brick nvidia 1 2 geforce articles guides gaming news featured stories www nvidia com how maxwell s vr direct brings virtual reality gaming closer to reality the official nvidia blog 1 2 ryan smith display matters hdmi 2 0 hevc vr direct the nvidia geforce gtx 980 review maxwell mark 2 archived from the original on september 22 2014 multi frame sampled anti aliasing delivers better performance to maxwell gamers www nvidia com new nvidia maxwell chips do not support fast csaa geforce rtx 20 series graphics cards and laptops nvidia ryan smith the nvidia geforce gtx 980 review maxwell mark 2 archived from the original on september 22 2014 1 2 3 ryan smith maxwell 2 architecture introducing gm204 the nvidia geforce gtx 980 review maxwell mark 2 archived from the original on september 22 2014 here s another reason the geforce gtx 970 is slower than the gtx 980 the tech report maxwell the most advanced cuda gpu ever made parallel forall 19 september 2014 geoffrey tim 3 december 2015 nvidia s gtx970 has a rather serious memory allocation bug lazygamer net 1 2 3 4 ryan smith diving deeper the maxwell 2 memory crossbar rop partitions geforce gtx 970 correcting the specs exploring memory allocation archived from the original on january 28 2015 smith ryan 17 march 2015 the nvidia geforce gtx titan x review anandtech p 2 archived from the original on march 19 2015 retrieved 6 december 2015 puny native fp64 rate of just 1 32 1 2 nvidia updates gpu roadmap announces pascal the official nvidia blog v t e nvidia geforce list of gpus fixed pixel pipeline pre geforce nv1 nv2 riva 128 riva tnt tnt2 geforce 256 2 4 mx vertex and pixel shaders geforce 3 4 ti fx 6 7 unified shaders geforce 8 9 100 200 300 400 500 unified shaders numa geforce 600 700 800m 900 10 16 ray tracing tensor cores geforce 20 30 40 50 software and technologies multimedia acceleration nvenc video encoding nvdec video decoding purevideo video decoding software cg shading language cuda nvidia gameworks optix ray tracing api physx physics sdk nvidia omniverse 3d graphics nvidia rtx ray tracing platform nvidia system tools vdpau video decode api technologies nvidia 3d vision stereo 3d nvidia g sync variable refresh rate nvidia optimus gpu switching nvidia surround multi monitor mxm module socket sxm module socket nvlink protocol sli multi gpu turbocache framebuffer in system memory video super resolution live video upscaling gpu microarchitectures celsius kelvin rankine curie tesla fermi kepler maxwell pascal volta turing ampere hopper ada lovelace blackwell rubin feynman other products graphics workstation cards nvidia quadro quadro plex gpgpu software nvidia tesla dgx console components nv2a xbox rsx reality synthesizer playstation 3 tegra x1 erista nintendo switch tegra t239 drake nintendo switch 2 nvidia shield shield portable shield tablet shield android tv geforce now socs and embedded goforce drive jetson tegra n1 nvidia rtx spark cpus project denver computer chipsets nforce company key people jensen huang chris malachowsky curtis priem david kirk bill dally colette kress debora shoquist ranga jayaraman jonah m alben acquisitions 3dfx interactive ageia uli bright computing cumulus networks deepmap icera mellanox technologies mental images portalplayer exluna mediaq stexar retrieved from https en wikipedia org w index php title maxwell_ microarchitecture oldid 1365695556 categories nvidia microarchitectures graphics microarchitectures hidden categories articles with short description short description is different from wikidata cs1 maint archived copy as title this page was last edited on 23 july 2026 at 22 44 utc page was rendered with parsoid text is available under the creative commons attribution sharealike 4 0 license additional terms may apply by using this site you agree to the terms of use and privacy policy wikipedia is a registered trademark of the wikimedia foundation inc a non profit organization privacy policy about wikipedia disclaimers contact wikipedia legal safety contacts code of conduct developers statistics cookie statement mobile view search search toggle the table of contents maxwell microarchitecture 10 languages add topic