PCTuning fórum

racecar píše:Co tu trollite? aDDmin to jasne vysvetlil:

aDDmin píše:já bych spíše přirovnal novou RX VEGA k 8800GTX ... ta také na začátku fungovala na půl, nové DX10 která měla mít prvního půl roku vůbec neuměla a nebylo jak otestovat, také přitom hodně žrala atd. ... v podstatě stejná situace.

Srandista píše:Krteq: lenze GPU cip nerobia len HBM cipy. Uz len samotne GPU core je o dost vacsie ako u Nvidie, a potom tu mame interposer, a uplne overkill VRM.

Btw, ako vychadza cenove porovnanie HBM vs. GDDR5?

ArgCZ píše: já se neskutečně bavím

aDDmin píše:já bych spíše přirovnal novou RX VEGA k 8800GTX ...

skaven321 píše:ale pre koncoveho pouzivatela je to plus, ze ma kartu z drahsich suciastok nie? (brutal VRM kaskada, skoro az kusova vyroba HBM2...)

In addition to Rapid Packed Math, the NCU introduces a
variety of new 32-bit integer operations that can improve
performance and efficiency in specific scenarios. These
include a set of eight instructions to accelerate memory
address generation and hashing functions (commonly used
in cryptographic processing and cryptocurrency mining), as
well as new ADD/SUB instructions designed to minimize
register usage.

racecar píše:... v podani overenych zdrojov - wccf a ocuk

Vega Whitepaper píše:An intelligent workload distributor (IWD) continually adjusts pipeline settings based on the characteristics of the draw calls it receives in order to maximize utilization.

One factor that can cause geometry engines to idle is context switching. Context switches occur whenever the engine changes from one render state to another, such as when changing from a draw call for one object to that of a different object with different material properties. The amount of data associated with render states can be quite large, and GPU processing can stall if it runs out of available context storage. The IWD seeks to avoid this performance overhead by avoiding context switches whenever possible.

Some draw calls also include many small instances (i.e.,they render many similar versions of a simple object). If an instance does not include enough primitives to fill a wavefront of 64 threads, then it cannot take full advantage of the GPU’s parallel processing capability, and some proportion of the GPU's capacity goes unused. The IWD can mitigate this effect by packing multiple small instances into a single wavefront, providing a substantial boost to utilization.