Braňo1977 píše:Herná Vega bude výkonom-spotrebou podobne ako gtx 1080Ti (oc)?
Pokud se nestane zazrak, tak tezko.
//
Jen tak mimochodem. Vynecham přestřelku o povesti:
Co se tyce samotneho problemu - tile-based rendering uz z principu vypnuty byt nemuze. To by byl takovy bottleneck, ze by performance nebyla na urovni GTX1080, ale na urovni GT 730. Kdokoliv to rika, tak nevi, o cem mluvi. Pokud to zacne tvrdit AMD, tak je to jen vymluva. Bud mate jeden klasicky rasterized s urcitym throughputem, nebo mate mnoho tile-based rasterizeru s malym throughputem. Vypnout tile-based rendering by znamenalo pouzivat jen ten jeden maly rasterizer s malym throughputem. Pokud je to opacne a jedna se o hybridni implementaci, tak zas nemuzete vubec nic vyhodnocovat z toho, jestli se ctverecky vykresluji sekvencne nebo paralelne, protoze takova synchronizace muze byt vlastnost architektury, stejne jako treba pipelines v CPU. Ty jsou take paralelni a zaroven sekvencni. A delaji operace paralelne, ale jedna stage je vzdy synchronni.
Tak nejak jsem se dostaval přes Google k nejakym podobnym zaverum, (to zvyrezneni tucne) v EN.
Dále ke konspiraci o driveru ze je z ledna a je jen nove zabalen:
„The driver“ being which components specifically?
The file quakechamps.ico is dated 4th of june, the VulkanRT installer was signed on april 6th. Some of the OpenCL files were signed on june 25th. So, I think "the driver is from january" is a bold statement.
Co říká o DSBR samotny AMD
The draw-stream binning rasterizer won't always be the rasterization approach that a Vega GPU will use. Instead, it's meant to complement the existing approaches possible on today's Radeons. AMD says that the DSBR is "highly dynamic and state-based," and that the feature is just another path through the hardware that can be used to improve rendering performance. By using data in a cache-aware fashion and only moving data when it has to, though, AMD thinks that this rasterizer will help performance in situations where the graphics memory (or high-bandwidth cache) becomes a bottleneck, and it'll also save power even when the path to memory isn't saturated.