Nadolazeće Kepler i Maxwell arhitekture za 2011. i 2013. godinu

[McG]

On the first day of the GPU Technology Conference (GTC), nVidia has unveiled brief details about its next-generation graphics processing units (GPUs). Quite logically, the future processor architectures by the company - code-named Kepler and Maxwell - will be massively focused on general-purpose computing and will offer substantial performance increases compared to existing chips.

Next year - sometimes in the second half of 2011 - nVidia plans to release GPUs based on the Kepler architecture. The chips will be made using 28nm process technology and will bring tangible performance improvements compared to currently shipping graphics chips based on Fermi architecture. Kepler architecture will increase performance per watt by about 3 or 4 times, hence, it is possible to expect something like 1.125TFLOPS - 1.50TFLOPS of DP performance by a chip with 250W thermal design power. Although math performance with DP accuracy does not have a direct correlation with graphics performance it is possible to expect Kepler to be at least two or three times faster than Fermi in games.

nVidia Maxwell will be launched in 2013, according to chief executive of nVidia Jen-Hsun Huang. Given the timeframe, it is logical to expect 20nm process technology to be used for manufacturing of Maxwell. The architecture due in three years from now will offer whopping 16GFLOPS of DP performance per watt, a massive improvement over current-generation hardware.

Izvor: X-Bit Labs

[Ivo_Strojnica]

Rekla-kazala...

[McG]

Ne znam koji dio u "On the first day of the GPU Technology Conference (GTC), nVidia has unveiled brief details about its next-generation graphics processing units (GPUs)", te tolko vuče na rekla-kazala.
Osim toga, ta konferencija je već dva dana jedna od glavnih tema svih važnijih IT portala. Ovaj put to nisu kojekakve glasine i nagađanja s izraelsko-tursko-rumunjskh foruma.

If Intel has IDF and Apple has the World Wide Developers Conference, then GTC is nVidia’s annual powwow to rally their developers and discuss their forthcoming plans.

[Astral]

zgleda dobro, al IMO pre kasno...

da je Kepler spreman za hd 6xxx onda bi bilo OK, ali ovak ce doc skoro godinu dana pre kasno - kako ja to vidim

[McG]

Pitanje je dal bude Kepler spreman za 2011., jer nVidia je tak i Fermi zacrtala za 2009., pa je ipak kasnil pol godine i izašel u proljeće 2010. Uglavnom, sudeći po ovome kaj su rekli na svojoj konferenciji i dalje idu s HPC GPGPU rješenjima koja budu (kad za to dođe vrijeme) baš kao i Fermi kasnije prilagodili gamerskoj i sveopćoj populaciji.

[Ivo_Strojnica]

To sve super zvuci, no tako je zvucao i Fermi. Ne zelim ja sad nista lose time reci, nego uzimam sve njihove izjave sa "trunkom" nevjerice.

[McG]

Zvuči super, ako se napokon okane monolitnih čipova i ako zahvaljujući manjem proizvodnom procesu, slože efikasnije (performanse, potrošnja i zagrijavanje) kartice koje budu opravdavale omjer uloženog i dobivenog. Ne zvuči (bar IMHO) super, jer/ako budu opet na sva vrata gurali GPGPU i na tome temeljili arhitekture koje budu kasnije prilagođavali svim tržišnim segmentima. Istu stvar su napravili s Fermi arhitekturom - napravili i dizajnirali su GPU koji se odlično snalazi u profesionalnim namjenama, napucali više od tri milijarde tranzistora u jedan ogroman čip, te ubili Boga u omjeru uloženog i dobivenog. Na kraju su pomoću drivera (GTX 480 i GTX 470), te s tweakanjem arhitekture (GTX 460, GTS 450 i sve kaj slijedi nakon njih), "pripitomili" moćan, rastrošan i monolitan GF100 za tržišta koja ciljaju na obične korisnike, gamere i entuzijaste.

[stef]

Ništa novo kod zelenih. Bliži se kraj godine, a financjski rezultat izgleda neće biti prema očekivanjima. Čim pričaju od ogromnim DP FP performansama, odmah znaš da će novi čip biti veći od bilo čega drugog. Usprkos tome oni najavljuju top to bottom launch u 3 mjeseca, novu arhitekturu na novom procesu (o kojem se ništa nezna) ... brda i doline.
Kad je ono bio predstavljen Fermi, slajdovi i "daska" s holc-šrafima ... jel to bilo u otprilike isto vrijeme lani?

[McG]

Je, tu negde su počeli slati okolo promo materijal, točnije 30.09. prošle godine. Kartice su stigle tek krajem trećeg, tj. početkom četvrtog mjeseca 2010. godine.
Ne znam, ako su tolko uvjereni da im bude opet prošel monolitan dizajn s ogromnim čipom i forsiranje HPC GPGPU rješenja di god sitgnu, onda stvarno nisu normalni.

[McG]

nVidia Corp. shed some light onto its exascale development project known as Echelon at the SC10 supercomputer-related trade-show last week. Even though today's graphics processing units (GPUs) are more efficient than central processing units (CPUs) in terms of raw performance-per-watt, as a result, graphics processors should evolve radically and rapidly in order to enable exascale systems in 2018 - 2020 timeframe.

The Echelon design incorporates a large number (~1024) of stream cores and a smaller (~8) number of latency-optimized CPU-like cores on a single chip, sharing a common memory system. Just like in current architectures, eight stream cores will form a streaming multiprocessor (SM) and 128 of SMs will forum the large pool of throughput-optimized processing elements. Such a chip could deliver 20 teraFLOPS with double precision and a number of them will form a 2.6 petaFLOPS rack. At present nVidia Fermi (GF110) chip 512 with stream processors operating at 1544MHz can deliver 0.79TFLOPS of DP compute performance. Considerint the 25 times difference in performance, it is highly likely that the Echelon will employ post-Maxwell (~2013 ~ 2014) Nvidia GPU design.

At present the Echelon is only a research project and not a chip from Nvidia's roadmap.
From some point of view, the Echelon is much like Intel's single-chip cloud computer (SCC) which belongs to Tera-Scale research project.

Izvor: X-Bit Labs

[nexusone]

Počelo je sempliranje GK107 čipova prema VR-zoneu.

[sapunko_HR]

Da li se zna kad bi trebao izać Kepler,barem neke glasine?

[cccp]

Glasine su da će kasniti, možda čak i Q2.