Amakhadi evidiyo e-NVIDIA Turing “Super” ahlaziyiweyo ngoku anamaxabiso acetyiswayo

Ngokutsho ngokungekho semthethweni ulwazi, ngomso i-NVIDIA inokubonisa usapho oluhlaziyiweyo lwamakhadi evidiyo kunye ne-Turing architecture, eya kufumana imemori ekhawulezayo, isimamva esithi "Super" ekutyunjweni kwemodeli, kwaye okona kubaluleke kakhulu, ukudibanisa okukhangayo kwexabiso kunye nokusebenza. Njengomthetho, kwi-niche nganye yexabiso, i-GPU kwi-Super series iya kuboleka kwikhadi levidiyo elidala lentsapho yangaphambili, kwaye inani lee-cores ze-CUDA ezisebenzayo liya kwandiswa, lichaphazela ngokuthe ngqo izinga lokusebenza.

ubutyebi I-WCCFTech ngobusuku besigaba sokuqala esilindelekileyo sesibhengezo, wabhengeza amaxabiso ezisombululo zegraphics ezintathu zomgca omtsha, oya kunikezelwa ngokuhambelana ne "wave yokuqala" Amakhadi evidiyo eTuring. IGeForce RTX 2080 Super iya kuxabisa i-$799, nto leyo eya kunyanzela iGeForce RTX 2080 “eqhelekileyo” ukuba ilahlekelwe lixabiso lokuhlalisana ngoxolo. I-GeForce RTX 2070 Super iya kufumana ithegi yexabiso elifanayo kwixabiso leGeForce RTX 2070 ngexesha lokubhengezwa - i-$ 599. Okokugqibela, iGeForce RTX 2060 Super ayizukulandela le algorithm yexabiso; ikhadi levidiyo lixabiswe nge-429 yeedola, ngelixa iGeForce RTX 2060 "eqhelekileyo" ekuqaleni ixabisa i-349 yeedola. Nangona kunjalo, kwimeko yokugqibela, ukunyuka kwexabiso kuhlawulwa kungekuphela nje ngokubonakala kwe-2176 CUDA cores endaweni ye-1920 yangaphambili, kodwa kunye nokunyuka kwememori ye-GDDR6 ukusuka kwi-6 ukuya kwi-8 GB.

• I-GeForce RTX 2080 Ti: 4352 CUDA cores, TU102-300 GPU kunye ne-11 GB GDDR6 imemori @ 14 GHz;
• GeForce RTX 2080 Super: 3072 cores CUDA, GPU I-TU104-450 kunye ne-8 GB yememori I-GDDR6 ene-frequency ye-16 GHz;
• I-GeForce RTX 2080: i-2944 CUDA cores, i-TU104-410 GPU kunye ne-8 GB yememori ye-GDDR6 kunye ne-frequency ye-14 GHz;
• GeForce RTX 2070 Super: 2560 cores CUDA, GPU I-TU104-410 kunye ne-8 GB yememori I-GDDR6 ene-frequency ye-14 GHz;
• I-GeForce RTX 2070: i-2304 CUDA cores, i-TU106-410 GPU kunye ne-8 GB yememori ye-GDDR6 kunye ne-frequency ye-14 GHz;
• GeForce RTX 2060 Super: 2176 cores CUDA, GPU I-TU106-410 kunye ne-8 GB yememori I-GDDR6 ene-frequency ye-14 GHz;
• I-GeForce RTX 2060: 1920 CUDA cores, TU106-200 GPU kunye ne-6GB GDDR6 imemori @ 14GHz.

Uluhlu olungentla lubonisa indlela uluhlu lwamakhadi evidiyo e-NVIDIA kunye ne-Turing architecture iya kutshintsha emva kokukhutshwa kwezisombululo zegraphics ezihlaziyiweyo. Amaxabiso emveliso esele ikhona kolu sapho aya kuthotywa. Amalungu amatsha osapho aya kuthengiswa kwisiqingatha sesibini sikaJulayi. I-Flegship GeForce RTX 2080 Ti ayiyi kuchaphazeleka ngohlaziyo; "idada ngaphezu kwesiphithiphithi kwi-echelon eyahlukileyo," kwaye ukukhululwa kwamakhadi evidiyo kwi-AMD Radeon RX 5700 yosapho akusongeli impilo-ntle yayo. Ekuphela kwento enokukhankanywa kulo mongo kukuba iGeForce RTX 2080 Ti iya kubelana ngeprosesa yemizobo kunye neGeForce RTX 2080 Super, eya kutyunjwa "TU104-450" ngeenjongo zokufihla.

umthombo: 3dnews.ru