NVIDIA GeForce 700 Series (Kepler Refresh)

[pixell]

Zašto ne dual GTX770 i 8GB??? Ovo nema smisla nikako.

 

[B-ZiGER]

I ja sam u zbunu. Kapiram da je ovo početak i da će izbaciti više MARS modela.

 

[DAS]

Ni ja ne vidim poentu ove karte. 2 x osakaceni GK104? A imali smo pre skoro godinu i po dana 2 x full GK104 u vidu GTX 690...

 

[La.Prestige]

Mozda je vidimo po ceni od 700e. Ostale su bile preko 1000. 🙂

 

[ABET]

Nema šanse da bude brža od 780ti, tako da nema smisla ni za tih 700 evra. Za 400 evra bi bila dobra kupovina.

Plus: potencijalno dobre performanse i Mars brend
Minus: sve ostalo... potrošnja, manjak vRAMa, bučna...

 

[Johnny_z11]

ASUS ROG MARS 760 pictured

From what I’ve heard, ROG MARS 760 is faster than all single-GPU cards. How will it perform in comparison to GTX 690 and HD 7990? You need to wait till Tuesday to know.

ASUS ROG MARS 760 pictured

 

[PCBeast]

Koja je svrha te karte uopste mi nije jasno ?!:S:
Glupost!

 

[PCBeast]

NVIDIA launches Tesla K40 packed with 12GB memory !

http://www.nvidia.com/object/tesla-servers.html

http://videocardz.com/48026/nvidia-launches-tesla-k40-packed-12gb-memory

 

[Rapture]

http://www.kitguru.net/components/graphic-cards/zardon/asus-gtx-760-mars-review-1600p-ultra-hd-4k/

 

[rAMA]

Karta nema ama baš nikakvog smisla...

 

[tomasbg]

za mala kucista je extra

 

[Rayko]

ASUS ROG GeForce GTX 760 MARS review

http://www.guru3d.com/articles_pages/asus_geforce_gtx_760_mars_review,1.html

 

[Rapture]

Ovo ko tetris, sivi i crni :d...

http://videocardz.com/48149/nvidia-geforce-gtx-titan-black-edition-pictured

 

[goldi]

Titan pocrnio, poceo da truhne 😀 znalo se to cim je rodjen Titan killer 😀 😀 😀

 

[Borat]

http://www.guru3d.com/news_story/nvidia_geforce_gtx_titan_black_edition.html

 

[DAS]

NVIDIA Launches Tesla K80 - GK210 GPU

http://www.anandtech.com/show/8729/nvidia-launches-tesla-k80-gk210-gpu

Tesla K80 se sastoji od dva GK210. U pitanju je novi Kepler GPU koji se od GK110 razlikuje po vecim memorijskim registrima unutar SMX-ova. Osim toga u K80 dva GK210 imaju 13 od 15 SMX-ova ukljucenih sto kupuno daje 2 x 2496 = 4992 shejdera. Ukupno ima 24 GB memorije (12 po GPU-u), deklarisan je na 300W, klokovi su 562/875 MHz (core/boost), pasivno je hladjen (stavlja se u kucista sa eksternim ventilatorima).

To that end, while NVIDIA hasn’t made any sweeping changes such as adjusting the number of CUDA cores or their organization (this is still a GK110 derivative, after all) NVIDIA has adjusted the memory subsystem in each SMX. Whereas a GK110(B) SMX has a 256KB register file and 64KB of shared memory, GK210 doubles that to a 512KB register file and 128KB of shared memory. Though small, this change improves the data throughput within an SMX, serving to improve efficiency and keep the CUDA cores working more often. NVIDIA has never made a change mid-stream like this to a GPU before, so this marks the first time we’ve seen a GPU altered in a later revision in this fashion. That said, this also reflects on the state of the GPU market, and how Kepler will still be with us for some time to come.

Overall I suspect that along with the memory change, NVIDIA has used this latest revision to once again tighten up the design of their HPC GPU to correct errata and reduce power consumption (thereby improving performance), which is part of the reason that NVIDIA is able to get two of these GPUs in a 300W card. Meanwhile GK210 will be in an odd place as it will likely be the first NVIDIA GPU not to end up in a consumer card; prior to this generation every GPU has pulled double duty as both a compute powerhouse and a graphics king. But with GM204 clearly ahead of GK110/GK210 in graphics, GK210 seems destined to Tesla cards and at most a Titan card for the budget compute market.

The final piece of the puzzle for Tesla K80 is GPU Boost. In the Tesla space NVIDIA introduced this on Tesla K40 in a far more limited implementation than on their consumer GPUs. Tesla K40 had to obey its TDP, but operators could select which of 3 clockspeeds they wanted, picking the one that comes closest to (but not exceeding) TDP for the best performance. However with Tesla K80 NVIDIA has now implemented a full and dynamic GPU boost implementation; just as in their consumer cards, the card will clock itself as high as the TDP will allow.