I-NVIDIA isenokuba ilungiselela iinguqulelo ezintsha zamakhadi ayo evidiyo esekwe kwiTuring GPUs. Ngokwejelo leYouTube iRedGamingTech, inkampani eluhlaza iceba ukuhlaziya ezinye zesantya samva nje sesizukulwana ngenkumbulo ekhawulezayo.
Okwangoku, amakhadi evidiyo eGeForce RTX axhotyiswe ngememori ye-GDDR6 ene-bandwidth ye-14 Gbps ngephini. Ngokomthombo, iinguqulelo zamva nje ziya kusebenzisa imemori ekhawulezayo kunye ne-bandwidth ye-16 Gbit / s ngoqhagamshelwano. Uphuculo olunjalo luya kuphucula kancinci ukusebenza, ngakumbi kwiimeko apho kuxhomekeke kwisantya sememori yekhadi levidiyo. Umzekelo, kwimidlalo ene-high-resolution textures.
Qaphela ukuba ayisosihlandlo sokuqala i-NVIDIA ithatha inyathelo elinjalo. Amanye amakhadi evidiyo esizukulwana sangaphambili esekwe kwiPascal GPUs nawo afumene inkumbulo ekhawulezayo ekuhambeni kwexesha. Ngokomzekelo, i-GeForce GTX 1080 ekuqaleni yasebenzisa imemori ye-GDDR5X kunye ne-bandwidth ye-10 Gbps, kwaye kamva i-NVIDIA yakhupha inguqulelo enememori ye-11 Gbps ekhawulezayo.
Kwangaxeshanye, iinkcukacha eziseleyo zamakhadi evidiyo esizukulwana sePascal zahlala zingatshintshi. Okunokwenzeka, kuya kuba njalo namakhadi evidiyo eTuring. Enyanisweni, umthombo awubandakanyi amathuba okuvela kweemodeli ezintsha. Umzekelo, i-NVIDIA inokukhulula ikhadi levidiyo le-GeForce RTX 2070 Ti, eliya kwahluka kwi-GeForce RTX 2070 kwimemori ekhawulezayo, kwaye inokuba nee-cores ze-CUDA okanye iifrikhwensi eziphezulu.
Ekugqibeleni, i-NVIDIA inokukhulula inguqu engaphezulu ye-flagship yekhadi levidiyo ye-GeForce RTX 2080 Ti, engayi kuba nememori ekhawulezayo kuphela, kodwa umthamo wayo uya kunyuka ukuya kwi-12 GB. Esi sikhawulezisi siya kuba nebhasi yememori ebanzi. Kodwa okwangoku, onke la ngamarhe nje. Mhlawumbi iinkcukacha ezithe vetshe malunga neenguqulelo ezintsha ze-Turing igeneration graphic accelerators ziya kuvela kumboniso ozayo weComputex 2019, oza kuqala nge-28 kaMeyi.
umthombo: 3dnews.ru
