Meta tags:

Headings (most frequently used words):

graphics, core, next, chips, command, processor, compute, video, schedulers, engine, precision, performance, contents, instruction, set, microarchitectures, generations, comparison, of, gcn, gpus, see, also, external, links, references, processing, geometric, units, audio, and, acceleration, blocks, unified, virtual, memory, heterogeneous, system, architecture, hsa, hardware, primitive, discard, accelerator, asynchronous, cu, scheduler, wavefront, simd, vector, unit, coding, zerocore, power,

Text of the page (most frequently used words):
the (230), and (120), amd (104), graphics (89), radeon (66), retrieved (63), for (56), with (54), gcn (54), gpu (53), vega (48), next (46), core (44), from (44), edit (39), polaris (37), 2017 (37), #compute (36), november (35), january (30), are (25), 2016 (25), support (24), architecture (24), instruction (24), this (23), com (23), original (23), memory (22), archived (22), cards (22), 2018 (21), gpus (21), techpowerup (21), video (20), 2022 (20), performance (20), units (20), one (19), anandtech (19), generation (19), simd (19), 2014 (18), unit (18), specs (17), hardware (17), engine (17), also (17), set (17), all (16), july (16), which (16), used (16), june (15), new (15), not (15), chips (15), that (15), was (14), articles (14), apus (14), shader (13), finfet (13), september (12), pro (12), series (12), other (12), driver (12), asynchronous (12), command (12), processing (12), precision (12), process (12), found (12), processor (12), each (12), vector (12), may (11), october (11), 2026 (11), unified (11), 2011 (11), but (11), islands (11), 256 (11), per (11), branded (11), has (11), queues (11), wavefronts (11), mobile (10), wikipedia (10), december (10), xbox (10), 2015 (10), engines (10), microarchitecture (10), schedulers (10), data (10), wavefront (10), august (9), 2012 (9), org (9), hsa (9), uvd (9), shaders (9), release (9), pdf (9), samsung (9), introduced (9), tasks (9), registers (9), microarchitectures (8), playstation (8), terascale (8), system (8), discrete (8), power (8), drivers (8), high (8), ryan (8), kernel (8), single (8), cache (8), die (8), can (8), number (8), page (7), integrated (7), apu (7), features (7), first (7), 2019 (7), primitive (7), march (7), based (7), higher (7), 2013 (7), kaveri (7), processors (7), into (7), geometry (7), when (7), windows (7), history (7), version (7), main (7), scheduler (7), use (6), needing (6), different (6), bandwidth (6), vce (6), 7000 (6), hawaii (6), comes (6), nvidia (6), preview (6), smith (6), more (6), purpose (6), aces (6), work (6), its (6), date (6), family (6), coding (6), glofo (6), developed (6), shot (6), schedule (6), successor (6), predecessor (6), part (6), over (6), article (6), threads (6), kib (6), hide (6), available (5), failed (5), verification (5), unsourced (5), statements (5), computing (5), instinct (5), cpu (5), powertune (5), api (5), time (5), software (5), encoding (5), 500 (5), ellesmere (5), tonga (5), review (5), stream (5), april (5), zerocore (5), cus (5), ace (5), such (5), display (5), 4096 (5), 1024 (5), 512 (5), 128 (5), jet (5), tsmc (5), globalfoundries (5), including (5), fp32 (5), were (5), generations (5), bit (5), numbers (5), there (5), two (5), supported (5), status (5), years (5), ago (5), refreshed (5), third (5), scheduling (5), since (5), virtual (5), priority (5), while (5), some (5), blocks (5), latency (5), toggle (4), contents (4), search (4), code (4), cs1 (4), deprecated (4), references (4), context (4), 2020 (4), rdna (4), tools (4), 200 (4), vertex (4), lexa (4), baffin (4), fiji (4), bonaire (4), tahiti (4), 7nm (4), techreport (4), need (4), group (4), dispatch (4), 480 (4), drm (4), changes (4), links (4), decoder (4), chelsea (4), texture (4), contains (4), function (4), accelerator (4), codenamed (4), design (4), technology (4), address (4), instructions (4), update (4), than (4), hevc (4), color (4), controllers (4), multiple (4), functional (4), they (4), these (4), file (4), general (4), scalar (4), move (4), sidebar (4), table (3), view (3), about (3), additional (3), you (3), maint (3), archival (3), service (3), parallel (3), gpgpu (3), computer (3), second (3), neo (3), current (3), optimization (3), refresh (3), trueaudio (3), technologies (3), level (3), developer (3), rocm (3), decoding (3), acceleration (3), 400 (3), wonder (3), before (3), fixed (3), pipeline (3), list (3), hainan (3), oland (3), cape (3), verde (3), pitcairn (3), press (3), world (3), ryzen (3), updated (3), opencl (3), kampman (3), jeff (3), draw (3), binning (3), carrizo (3), 290x (3), link (3), cite (3), operations (3), llvm (3), end (3), community (3), see (3), listed (3), launch (3), oem (3), southern (3), jun (3), 2048 (3), render (3), output (3), 640 (3), 384 (3), malta (3), 000 (3), size (3), only (3), half (3), vii (3), floating (3), point (3), ridge (3), electronics (3), added (3), increase (3), their (3), include (3), controller (3), lower (3), example (3), space (3), clock (3), less (3), regular (3), lpp (3), slightly (3), fourth (3), discard (3), frames (3), etc (3), tessellation (3), independent (3), called (3), audio (3), pcie (3), dma (3), citation (3), needed (3), opengl (3), specific (3), run (3), allowing (3), heterogeneous (3), copy (3), copied (3), distinct (3), succeeded (3), lane (3), four (3), vgpr (3), program (3), between (3), consists (3), operation (3), cycle (3), share (3), every (3), read (3), message (3), improve (3), learn (3), help (3), risc (3), add (2), languages (2), contact (2), privacy (2), policy (2), under (2), terms (2), using (2), wikimedia (2), commons (2), categories (2), url (2), pages (2), dates (2), short (2), description (2), wikidata (2), related (2), lenovo (2), legion (2), asus (2), rog (2), ally (2), project (2), oberon (2), 2021 (2), scorpio (2), durango (2), liverpool (2), 360 (2), gamecube (2), obsolete (2), products (2), multi (2), rate (2), freesync (2), codexl (2), sdk (2), mantle (2), legacy (2), gpuopen (2), open (2), source (2), amdgpu (2), vcn (2), multimedia (2), ray (2), tracing (2), cores (2), 6000 (2), 5000 (2), 300 (2), 8000 (2), 2000 (2), fragment (2), template (2), february (2), topaz (2), radeonfeature (2), unveils (2), coming (2), gaming (2), announces (2), mi60 (2), mi50 (2), accelerators (2), runtime (2), will (2), curtain (2), revealed (2), teaser (2), rasterizer (2), tile (2), rendering (2), shrout (2), perspective (2), 590 (2), www (2), 285 (2), freedesktop (2), a10 (2), 7850k (2), idle (2), off (2), details (2), pci (2), express (2), amdkfd (2), 2024 (2), usage (2), thread (2), mantor (2), michael (2), deep (2), miaow (2), offloading (2), gcc (2), back (2), architectural (2), card (2), external (2), treasure (2), initial (2), variant (2), arctic (2), pirate (2), volcanic (2), sea (2), nov (2), feb (2), aug (2), mar (2), 768 (2), mapping (2), 1280 (2), unknown (2), transistor (2), 550 (2), 123 (2), greenland (2), comparison (2), 5th (2), double (2), fp64 (2), allows (2), full (2), decode (2), edition (2), licensed (2), done (2), maxwell (2), efficiency (2), due (2), expected (2), improvements (2), heavily (2), modified (2), previous (2), math (2), significant (2), same (2), speeds (2), cdna (2), addition (2), clocks (2), 560 (2), 460 (2), 470 (2), newer (2), enabling (2), 3rd (2), bits (2), feature (2), have (2), delta (2), compression (2), quality (2), mesa (2), beema (2), mullins (2), item (2), dispatching (2), 2nd (2), well (2), rops (2), advantages (2), desktop (2), 290 (2), unsupported (2), mid (2), final (2), long (2), through (2), events (2), several (2), enter (2), triangles (2), includes (2), without (2), possible (2), resources (2), essentially (2), purposes (2), perform (2), until (2), least (2), queue (2), assign (2), device (2), existing (2), very (2), linux (2), special (2), zero (2), bus (2), supports (2), needs (2), note (2), shows (2), adding (2), encode (2), yuv420 (2), asic (2), many (2), limited (2), takes (2), cycles (2), register (2), most (2), basic (2), width (2), sgprs (2), vgprs (2), waiting (2), attributed (2), vus (2), thus (2), warp (2), small (2), written (2), performs (2), processes (2), load (2), similar (2), calls (2), block (2), confused (2), local (2), inputs (2), get (2), both (2), operate (2), handles (2), tesselator (2), assembler (2), geometric (2), execution (2), commands (2), handle (2), compiler (2), sets (2), section (2), how (2), remove (2), please (2), citations (2), talk (2), those (2), vliw (2), switch (2), appearance (2), upload (2), subsection (2), log (2), create (2), account (2), donate (2), menu (2), topic, cookie, statement, statistics, developers, conduct, legal, safety, contacts, disclaimers, text, apply, site, agree, registered, trademark, non, profit, organization, foundation, inc, creative, attribution, sharealike, license, rendered, parsoid, last, edited, utc, hidden, unfit, clarification, disputed, accuracy, disputes, cleanup, mdy, introductions, https, index, php, title, graphics_core_next, oldid, 1378337360, msi, claw, microsoft, ayaneo, valve, steam, deck, onexplayer, aya, handheld, pcs, viola, scarlett, lockhart, slim, plus, atari, vcs, wii, hollywood, xenos, flipper, consoles, firestream, firemv, firegl, firepro, supercomputers, workstations, udna, hyperz, shared, hypermemory, hybrid, crossfire, dynamic, management, variable, monitor, eyefinity, stereo, low, ctm, analysis, perfstudio, profiling, debugging, app, catalyst, translation, hlsl2glsl, real, hair, fur, effects, tressfx, platform, control, 9000, 600, 4000, 3000, after, 1999, r500, r400, r300, r200, r100, rage, mach, brand, datacenter, powering, era, artificial, intelligence, cloud, hpc, ferreira, bruno, tech, report, laptop, near, las, vegas, nevada, delivering, exceptional, incredible, experiences, gamers, creators, enthusiasts, powered, lives, macbook, pros, burning, bright, github, libutils, cpp, master, radeonopencompute, trustedreviews, promises, 499, trusted, reviews, raevenlord, 2700u, 2500u, redesigned, nate, feat, xfx, powercolor, returns, again, 3dcenter, fast, vollständige, daten, amds, serie, geleakt, whycry, videocardz, confirms, ipc, tiling, guru3d, cutress, ian, 7th, dcc, png, images, overview, tested, managing, introducing, web, serve, principal, accept, architected, tony, chen, jason, greaves, 2023, gen, compliant, makes, mainstream, mark, promise, dissecting, angelini, chris, tom, 8gb, lkml, patch, airlie, dave, merge, semiaccurate, just, too, white, paper, unifiedvideodecoder, optimizing, occupancy, resource, large, groups, houston, mike, news, whitepaper, khronos, conformant, dives, shading, fixes, forums, gulati, abheek, medium, dive, architectures, launches, fastest, 7970, central, slideshare, net, 4106, crash, course, layla, mah, official, website, old, names, notes, apr, oct, jan, dec, stencil, ops, 144, 176, 1536, 2304, 2816, 896, 320, density, mtr, 230, 700, 900, 080, 690, 950, 800, 313, transistors, million, 331, 250, 495, 208, 103, 232, 596, 366, 125, 438, 160, 212, 352, 365, fab, capsaicin, amethyst, antigua, meso, weston, vesuvius, grenada, saturn, tobago, strato, emerald, sun, exo, banks, mars, opal, litho, heathrow, venus, tropo, wimbledon, curaçao, neptune, trinidad, zealand, moonshot, iceland, ibiza, tiran, name, vegam, igp, console, socs, except, sixteenth, quarter, calculations, twice, fp16, cezanne, renoir, picasso, raven, came, supersedes, vp9, frontier, v340, 9100, 8200, uses, stage, provide, flexible, replace, because, required, yet, rasterization, big, reason, assumed, would, finally, catch, regarding, energy, optimizations, dsbr, began, releasing, termed, larger, chipsets, hbcc, additionally, packed, rapid, pack, advantage, acceptable, rasterisation, hbm2, 4th, dedicated, utilized, titled, respectively, 3200, 540x, kaby, lake, 570, 580, 540, produced, built, node, compatible, resolutions, channel, gca, array, gfx, identical, stoney, bristol, hbm, m295x, improved, lossless, reduce, efficient, scaler, encoder, however, worse, compared, nano, godavari, kabini, temash, summit, presented, entity, comprises, chip, rasterizers, interface, revised, puma, steamroller, 390, 390x, 260, 260x, 295x2, 8770, 7790, saving, shutting, supplements, controlling, computation, vulkan, dynamically, adjusts, stay, within, tdp, partially, resident, textures, enable, extensions, directx, addressing, sends, various, faults, interrupt, requests, x86, 7950, ghz, discards, cover, any, fragments, degenerate, performed, prioritized, allow, critical, requiring, preempted, task, therefore, concurrently, scheduled, hog, much, letting, lack, present, internal, testing, enabled, parts, production, offload, assignment, buffering, empty, causes, hws, immediately, buffered, safely, implemented, operating, subsystems, published, patches, merged, supporting, resides, directory, reside, augment, already, implementation, focuses, works, alongside, kgd, mainline, gain, conserving, scarce, solutions, suffer, allocated, exclusively, partition, partitioned, hence, instead, passed, paramount, pointers, classical, physical, spaces, entire, diagram, bandwidths, wrote, subsequently, formed, scaling, 265, codec, pixel, format, along, sve, temporal, mode, frame, yuv444, 264, implementations, typically, accompanied, buffers, execute, issues, buffer, logical, counters, integer, alu, arithmetic, logic, conjunction, concept, parallelism, often, characterized, total, sse, tables, relationship, 104, solution, attribute, distributes, result, lies, assigns, another, exchange, maximum, don, chose, approaches, unavoidable, grouping, whereas, implement, approach, minimum, processed, fashion, smallest, executable, way, glsl, choosing, executes, picks, claimed, lds, room, private, where, stores, types, hold, bytes, represent, byte, correspond, pixels, them, wired, 16kib, 32kib, operates, elements, branching, fetch, store, filter, gpr, files, wide, branch, combines, tmus, separate, feed, following, capable, doing, defined, then, latest, semiconductor, intellectual, property, ati, truform, direct3d, given, minimize, iteration, during, latter, helps, executing, underutilized, speed, functionality, known, async, serving, whose, ambiguous, gcp, among, responsible, handling, parse, incoming, manage, items, complete, streams, interleaved, seen, five, iterations, differences, rather, minimal, fifth, simultaneous, place, shares, encodings, isa, discuss, dubious, threaded, stand, alone, programs, via, openacc, openmp, gnu, collection, referred, documentation, owned, specifically, division, micro, relevant, discussion, material, challenged, removed, reliable, sources, consumer, center, selected, models, separately, released, portion, accelerated, cmos, fabricated, contrasting, mandates, considerable, alleviates, complexity, involved, generating, fuller, simpler, easier, compilers, count, word, reduced, computers, product, featuring, launched, circuit, providing, reader, provides, insufficient, unfamiliar, subject, nintendo, redirected, free, encyclopedia, projects, printable, download, print, export, parser, shortened, information, permanent, what, here, actions, english, português, norsk, bokmål, 한국어, 日本語, italiano, français, suomi, فارسی, español, català, top, personal, recent, portal, contribute, random, navigation, jump, content,

Text of the page (random words):
ast one empty queue in at least one ace this causes the hws to immediately assign buffered queues to the aces until all queues are full or there are no more queues to safely assign 24 part of the scheduling work performed includes prioritized queues which allow critical tasks to run at a higher priority than other tasks without requiring the lower priority tasks to be preempted to run the high priority task therefore allowing the tasks to run concurrently with the high priority tasks scheduled to hog the gpu as much as possible while letting other tasks use the resources that the high priority tasks are not using 23 these are essentially asynchronous compute engines that lack dispatch controllers 23 they were first introduced in the fourth generation gcn microarchitecture 23 but were present in the third generation gcn microarchitecture for internal testing purposes 25 a driver update has enabled the hardware schedulers in third generation gcn parts for production use 23 primitive discard accelerator edit this unit discards degenerate triangles before they enter the vertex shader and triangles that do not cover any fragments before they enter the fragment shader 26 this unit was introduced with the fourth generation gcn microarchitecture 26 generations edit graphics core next 1 edit amd graphics core next 1 release date january 2012 14 years ago january 2012 history predecessor terascale 3 successor graphics core next 2 support status unsupported since mid 2022 final windows driver version 22 6 1 for windows 7 and 10 the gcn 1 microarchitecture was used in several radeon hd 7000 series graphics cards die shot of the tahiti gpu used in radeon hd 7950 ghz edition graphics cards support for 64 bit addressing x86 64 address space with unified address space for cpu and gpu 18 support for pcie 3 0 27 gpu sends interrupt requests to cpu on various events such as page faults support for partially resident textures 28 which enable virtual memory support through directx and opengl extensions amd powertune support which dynamically adjusts performance to stay within a specific tdp 29 support for mantle api vulkan and opengl 4 6 there are asynchronous compute engines controlling computation and dispatching 15 30 zerocore power edit zerocore power is a long idle power saving technology shutting off functional units of the gpu when not in use 31 amd zerocore power technology supplements amd powertune chips edit discrete gpus southern islands family hainan oland cape verde pitcairn tahiti graphics core next 2 edit amd graphics core next 2 release date september 2013 13 years ago september 2013 history predecessor graphics core next 1 successor graphics core next 3 support status unsupported since mid 2022 final windows driver version 22 6 1 for windows 7 and 10 amd powertune bonaire die shot of the hawaii gpu used in radeon r9 290 graphics cards the 2nd generation of gcn was introduced with the radeon hd 7790 and is also found in the radeon hd 8770 r7 260 260x r9 290 290x r9 295x2 r7 360 and r9 390 390x as well as steamroller based desktop kaveri apus and mobile kaveri apus and in the puma based beema and mullins apus it has multiple advantages over the original gcn including freesync support amd trueaudio and a revised version of amd powertune technology gcn 2nd generation introduced an entity called shader engine se citation needed a shader engine comprises one geometry processor up to 44 cus hawaii chip rasterizers rops and l1 cache not part of a shader engine is the graphics command processor the 8 aces the l2 cache and memory controllers as well as the audio and video accelerators the display controllers the 2 dma controllers and the pcie interface the a10 7850k kaveri contains 8 cus compute units and 8 asynchronous compute engines for independent scheduling and work item dispatching 32 at amd developer summit apu in november 2013 michael mantor presented the radeon r9 290x 33 chips edit discrete gpus sea islands family bonaire hawaii integrated into apus temash kabini liverpool i e the apu found in the playstation 4 durango i e the apu found in the xbox one and xbox one s kaveri godavari mullins beema carrizo l graphics core next 3 edit amd graphics core next 3 release date june 2015 11 years ago june 2015 history predecessor graphics core next 2 successor graphics core next 4 support status supported with less regular windows driver update schedule die shot of the fiji gpu used in radeon r9 nano graphics cards gcn 3rd generation 34 was introduced in 2014 with the radeon r9 285 and r9 m295x which have the tonga gpu it features improved tessellation performance lossless delta color compression to reduce memory bandwidth usage an updated and more efficient instruction set a new high quality scaler for video hevc encoding vce 3 0 and hevc decoding uvd 6 0 and a new multimedia engine video encoder decoder delta color compression is supported in mesa 35 however its double precision performance is worse compared to previous generation 36 failed verification chips edit discrete gpus tonga volcanic islands family comes with uvd 5 0 unified video decoder fiji pirate islands family comes with uvd 6 0 and high bandwidth memory hbm 1 integrated into apus carrizo comes with uvd 6 0 bristol ridge 37 stoney ridge 37 graphics core next 4 edit amd graphics core next 4 release date june 2016 10 years ago june 2016 history predecessor graphics core next 3 successor graphics core next 5 support status supported with less regular windows driver update schedule die shot of the polaris 11 gpu used in radeon rx 460 graphics cards die shot of the polaris 10 gpu used in radeon rx 470 graphics cards gpus of the arctic islands family were introduced in q2 of 2016 with the amd radeon 400 series the 3d engine i e gca graphics and compute array or gfx is identical to that found in the tonga chips 38 but polaris feature a newer display controller engine uvd version 6 3 etc all polaris based chips other than the polaris 30 are produced on the 14 nm finfet process developed by samsung electronics and licensed to globalfoundries 39 the slightly newer refreshed polaris 30 is built on the 12 nm lp finfet process node developed by samsung and globalfoundries the fourth generation gcn instruction set architecture is compatible with the third generation it is an optimization for 14 nm finfet process enabling higher gpu clock speeds than with the 3rd gcn generation 40 architectural improvements include new hardware schedulers a new primitive discard accelerator a new display controller and an updated uvd that can decode hevc at 4k resolutions at 60 frames per second with 10 bits per color channel chips edit discrete gpus 41 polaris 10 also codenamed ellesmere found on radeon rx 470 and radeon rx 480 branded graphics cards polaris 11 also codenamed baffin found on radeon rx 460 branded graphics cards also radeon rx 560 d polaris 12 also codenamed lexa found on radeon rx 550 and radeon rx 540 branded graphics cards polaris 20 which is a refreshed 14 nm lpp samsung glofo finfet process polaris 10 with higher clocks used for radeon rx 570 and radeon rx 580 branded graphics cards 42 polaris 21 which is a refreshed 14 nm lpp samsung glofo finfet process polaris 11 used for radeon rx 560 branded graphics cards polaris 22 found on radeon rx vega m gh and radeon rx vega m gl branded graphics cards as part of kaby lake g polaris 23 which is a refreshed 14 nm lpp samsung glofo finfet process polaris 12 used for radeon pro wx 3200 and radeon rx 540x branded graphics cards also radeon rx 640 43 polaris 30 which is a refreshed 12 nm lp glofo finfet process polaris 20 with higher clocks used for radeon rx 590 branded graphics cards 44 in addition to dedicated gpus polaris is utilized in the apus of the playstation 4 pro and xbox one x titled neo and scorpio respectively precision performance edit fp64 performance of all gcn 4th generation gpus is 1 16 of fp32 performance graphics core next 5 edit amd graphics core next 5 release date june 2017 9 years ago june 2017 history predecessor graphics core next 4 successor cdna 1 rdna 1 support status supported with less regular windows driver update schedule see also radeon rx vega series die shot of the vega 10 gpu used in radeon rx vega 64 graphics cards amd began releasing details of their next generation of gcn architecture termed the next generation compute unit in january 2017 40 45 46 the new design was expected to increase instructions per clock higher clock speeds support for hbm2 a larger memory address space the discrete graphics chipsets also include hbcc high bandwidth cache controller but not when integrated into apus 47 additionally the new chips were expected to include improvements in the rasterisation and render output units the stream processors are heavily modified from the previous generations to support packed math rapid pack math technology for 8 bit 16 bit and 32 bit numbers with this there is a significant performance advantage when lower precision is acceptable for example processing two half precision numbers at the same rate as a single single precision number nvidia introduced tile based rasterization and binning with maxwell 48 and this was a big reason for maxwell s efficiency increase in january anandtech assumed that vega would finally catch up with nvidia regarding energy efficiency optimizations due to the new dsbr draw stream binning rasterizer to be introduced with vega 49 it also added support for a new shader stage primitive shaders 50 51 primitive shaders provide more flexible geometry processing and replace the vertex and geometry shaders in a rendering pipeline as of december 2018 the primitive shaders can t be used because required api changes are yet to be done 52 vega 10 and vega 12 use the 14 nm finfet process developed by samsung electronics and licensed to globalfoundries vega 20 uses the 7 nm finfet process developed by tsmc chips edit discrete gpus vega 10 14 nm samsung glofo finfet process also codenamed greenland 53 found on radeon rx vega 64 radeon rx vega 56 radeon vega frontier edition radeon pro v340 radeon pro wx 9100 and radeon pro wx 8200 graphics cards 54 vega 12 14 nm samsung glofo finfet process found on radeon pro vega 20 and radeon pro vega 16 branded mobile graphics cards 55 vega 20 7 nm tsmc finfet process found on radeon instinct mi50 and radeon instinct mi60 branded accelerator cards 56 radeon pro vega ii and radeon vii branded graphics cards 57 integrated into apus raven ridge 58 came with vcn 1 which supersedes vce and uvd and allows full fixed function vp9 decode picasso renoir cezanne precision performance edit double precision floating point fp64 performance of all gcn 5th generation gpus except for vega 20 is one sixteenth of fp32 performance for vega 20 with radeon instinct this is half of fp32 performance for vega 20 with radeon vii this is a quarter of fp32 performance 59 all gcn 5th generation gpus support half precision floating point fp16 calculations which is twice of fp32 performance comparison of gcn gpus edit table contains only discrete gpus including mobile apu igp and console socs are not listed microarchitecture 60 gcn 1 gcn 2 gcn 3 gcn 4 gcn 5 die tahiti 61 pitcairn 62 cape verde 63 oland 64 hainan 65 bonaire 66 hawaii 67 topaz 68 tonga 69 fiji 70 ellesmere 71 baffin 72 lexa 73 vegam 74 vega 10 75 vega 12 76 vega 20 77 code name 1 tiran ibiza iceland polaris 10 polaris 11 polaris 12 polaris 22 greenland treasure refresh moonshot variant s new zealand malta wimbledon curaçao neptune trinidad chelsea heathrow venus tropo mars opal litho sun jet exo banks saturn tobago strato emerald vesuvius grenada meso weston polaris 24 amethyst antigua capsaicin polaris 21 polaris 23 n a n a n a n a n a fab tsmc 28 nm globalfoundries 14 nm 12 nm polaris 30 tsmc 7 nm die size mm 2 352 365 malta 212 123 77 56 160 438 125 366 596 232 123 103 208 495 250 331 transistors million 4 313 2 800 1 500 950 690 2 080 6 200 1 550 5 000 8 900 5 700 3 000 2 200 5 000 12 500 unknown 13 230 transistor density mtr mm 2 12 3 12 8 malta 13 2 12 2 12 3 13 0 14 2 12 4 13 7 14 9 24 6 24 4 21 4 24 0 25 3 unknown 40 0 asynchronous compute engines 2 8 8 4 geometry engines 2 1 2 n a n a 4 2 4 shader engines n a 4 4 2 4 n a hardware schedulers n a 2 compute units 32 20 10 8 chelsea 6 5 6 jet 14 44 6 32 64 36 16 10 24 64 20 64 stream processors 2048 1280 640 512 chelsea 384 320 384 jet 896 2816 384 2048 4096 2304 1024 640 1536 4096 1280 4096 texture mapping units 128 80 40 32 chelsea 24 20 24 jet 56 176 24 128 256 144 64 40 96 256 80 256 render output units 32 16 8 16 64 8 32 64 32 16 64 32 64 z stencil ops 128 64 16 64 256 16 128 256 n a l1 cache kb 16 per compute unit cu l2 cache kb 768 512 256 128 256 jet 256 1024 256 768 2048 1024 512 1024 4096 1024 4096 display core engine 6 0 6 4 n a 8 2 8 5 n a 10 0 11 2 12 0 12 1 unified video decoder 3 2 4 0 n a 4 2 n a 5 0 6 0 6 3 7 0 7 2 video coding engine 1 0 n a 2 0 n a 3 0 3 4 4 0 4 1 launch 2 dec 2011 mar 2012 feb 2012 jan 2013 may 2015 mar 2013 oct 2013 2014 aug 2014 jun 2015 jun 2016 aug 2016 apr 2017 feb 2018 jun 2017 nov 2018 nov 2018 series family southern islands sea islands volcanic islands pirate islands arctic islands vega vega ii notes mobile oem mobile oem mobile oem mobile 1 old code names such as treasure lexa or hawaii refresh ellesmere are not listed 2 initial launch date launch dates of variant chips such as polaris 20 april 2017 are not listed see also edit list of amd graphics processing units external links edit official amd com graphics core next gcn website references edit amd developer central january 31 2014 gs 4106 the amd gcn architecture a crash course by layla mah slideshare net amd launches world s fastest single gpu graphics card the amd radeon hd 7970 press release amd december 22 2011 archived from the original on january 20 2015 retrieved january 20 2015 gulati abheek november 11 2019 an architectural deep dive into amd s terascale gcn rdna gpu architectures medium retrieved december 12 2021 amd community forums community amd com july 15 2016 llvm back end amdgpu llvm org gcc 9 release series changes new features and fixes retrieved november 13 2019 amd gcn offloading support retrieved november 13 2019 miaow gpu retrieved september 21 2026 smith ryan january 5 2017 the amd vega gpu architecture preview anandtech com archived from the original on january 6 2017 retrieved july 11 2017 smith ryan amd dives deep on asynchronous shading anandtech com archived from the original on april 2 2015 conformant products khronos org december 16 2018 compute cores whitepaper pdf amd 2014 p 5 1 2 smith ryan december 21 2011 amd s graphics core next preview anandtech com retrieved april 18 2017 cite news cs1 maint deprecated archival service link amd s graphics core next gcn architecture pdf techpowerup retrieved february 26 2024 1 2 mantor michael houston mike june 15 2011 amd graphics core next pdf amd p 40 retrieved july ...