nVidia Fermi

[starx]

Uh dobre brojke, ma bice jak i srednji segment ako je ovo tacno izgleda da je cip jak, tj arhitektura, sta da kazemo nego da sacekamo..

[DAKINS]

Divna arhktetura mada cu verovatno da preskacim i 3xxx kao ati 5xxx jer cak je i ona pravljena 40nm tehnologiji .
Sve indicije postoje da ce se u roku od god dana preci pre 28nm nego 32nm .
A instikt mi kaze da prve direkt11 karte ce mocu samo pokretati direk11 igre samo u pocetku .
I mada se ovo pre odnosi na ati nego Nvidiju ali ipak ne vidim razlog da zurim.

[starx]

Pa ok ako nemoras da kupis naravno da neces pored tvoje 4870 i tako neke , naravno da je glupo zbog toga..

[starx]

Ovde guru3d govori malo o fermiju...

http://www.guru3d.com/article/nvidia...logy-preview/1

ima oko osam strana.



http://techpowerup.com/reviews/NVIDI...tecture/1.html

http://www.anandtech.com/video/showdoc.aspx?i=3721


http://www.hardwarecanucks.com/forum...oscope-13.html

[Picard]

http://en.expreview.com/2010/01/18/g...thed/6475.html

[RUSuper]

http://www.hardocp.com/images/articl...4LAl_1_4_l.png
Svako ko je gledao onaj test GF100(mislim na onaj link gde su po prvi put prikazali grafu) znace odakle je ovo

http://www.youtube.com/watch?v=wkJTg...layer_embedded



Koliko je istinita ova slika sa specifikacijama mogu samo da nagadjam....

A ovo mislim da je skinuo svako ko se prosetao po Nv sajtu

[... i bidi fin]

KONACNO

http://www.hardwarecanucks.com/forum...oscope-13.html

[vlajo]

Šta "Konačno"?

[starx]

... i bidi fin kaže:

KONACNO

http://www.hardwarecanucks.com/forum...oscope-13.html

Verovatno nisi procitao postovao sam to kao link pre tri dana, hehe mrzelo te da citas sve

[Predrag Stankovic]

Interesantno da će GF100 koristiti 384-bitnu memorijsku magistralu, a na GT200 je bila 512bit, verovatno su i ovako isuviše dobre performanse, sigurno kasnije će moći da izbace i karticu gtx390 sa 512bit magistralom ukoliko im AMD/ATI pripreti sa nekim novim konkurentnim proizvodom, možda im je to kao neki džoker

[Picard]

I should pause to explain the asterisk next to the unexpectedly low estimate for the GF100's double-precision performance. By all rights, in this architecture, double-precision math should happen at half the speed of single-precision, clean and simple. However, Nvidia has made the decision to limit DP performance in the GeForce versions of the GF100 to 64 FMA ops per clock—one fourth of what the chip can do. This is presumably a product positioning decision intended to encourage serious compute customers to purchase a Tesla version of the GPU instead. Double-precision support doesn't appear to be of any use for real-time graphics, and I doubt many serious GPU-computing customers will want the peak DP rates without the ECC memory that the Tesla cards will provide. But a few poor hackers in Eastern Europe are going to be seriously bummed, and this does mean the Radeon HD 5870 will be substantially faster than any GeForce card at double-precision math, at least in terms of peak rates.

http://techreport.com/articles.x/18332/5

Analiza troskova izrade i nagadjanja u vezi cene:

http://www.brightsideofnews.com/news...-analysis.aspx

[Aibo]

Ako su ove price o troskovima bliske realnosti, necemo izgleda jos dugo videti nista iz nVidie u srednjoj klasi. Otprilike da ce u skorijoj buducnosti stici samo hiEnd resenja.

[Predrag Stankovic]

^ Nemoj da si pesimista

[Thaedass]

Zanimljivo

» GF100 gets a name: GeForce GTX 480

NVIDIA has determined a name for their next-generation high-end DirectX 11 GPU. The first graphics cards based on their GF100 architecture will be named the GeForce GTX 480 and GeForce GTX 470. From NVIDIA's facebook page:
NVIDIA Fun Fact of the Week: GeForce GTX 480 and GeForce GTX 470 will be the names of the first two GPUs shipped based on our new GF100 chip!

NVIDIA still hasn't disclosed the specs or configuration for the new GeForce 400 series cards, but it's likely that the 470 will ship at slower clocks and with fewer active shading units.

It's quite surprising to see NVIDIA skip the 300 series designation completely with their latest GPUs. Some will inevitably say this gives them room to rename existing GPUs as higher-spec 300 series cards.

Izvor

Link

[player1]

^
Svaki dodatni komentar je suvisan.

[RUSuper]

Sto samo stavili druge brojke :P

[galebns]

Ma nek se zovu i sherpe samo da rade posao kako valja

[Predrag Stankovic]

Glup i bespotreban rebrand, ali njihova stvar, prvo sam mislio da je fake, al sad vidim ima o tome da piše na višeozbiljnijih sajtova

[player1]

Sad ispada da je cela 300 serija jedan veliki fake, tj. rebrand 200 serije, i ama bas nista vise.

[... i bidi fin]

player1 kaže:

Sad ispada da je cela 300 serija jedan veliki fake, tj. rebrand 200 serije, i ama bas nista vise.

Iz tvojih usta u Qzmine usi !