Ngesikhathi seComputex 2021, UNvidia umemezele ukusebenzisana neValve ukuhlinzeka ukwesekwa kwe-DLSS (Deep Learning Super Sampling) abakhona emakhadini abo e-RTX.
I-DLSS, noma i-Deep Learning Super Sampling, iyi- ubuchwepheshe obukuvumela ukuthi uthole ukusebenza okuningi ngaphandle kokuyeka ikhwalithi yesithombe eningi kakhulu. Ukwenza lokhu, umdlalo ugijima esinqumweni esingaphansi kwesinqumo sendabuko bese isithombe siguqulwa sibe yisinqumo sendabuko kusetshenziswa ama-algorithms.
Ukumenyezelwa kokwesekwa kweValve ukuxhasa ubuchwepheshe bayo be-DLSS kuyizindaba ezimnandi, njengoba i-DLSS ingathuthukisa kakhulu amazinga wohlaka ngaphandle kokuthinta izinga lehluzo.
“I-DLSS isebenzisa ukunikezwa kwe-AI okuthuthukile ukukhiqiza ikhwalithi yesithombe efana nokuxazululwa kwendabuko futhi kwesinye isikhathi kube ngcono kakhulu, ngenkathi kunikezwa kuphela ukunikezwa okujwayelekile kwengxenyana yamaphikseli. Amasu okuphendula esikhathi esithuthukile aletha imininingwane ebukhali yezithombe nokuqina kozimele kozimele, ”kusho uNVIDIA.
Umthelela we-DLSS ungamangaza emidlalweni esekela lobu buchwepheshe. Kwezinye izimo, ingaphezu kokuphindwe kabili kwamanani wozimele ngaphandle kwe-DLSS, imvamisa engenawo umthelela wokubuka omncane. Intshisekelo yalobu buchwepheshe ilele ekufundeni okujulile.
Inethiwekhi eqeqeshiwe ye-neural ingcono ekuhlonzeni izingxenye zesithombe ezihambisana kakhulu nokubona komuntu kunama-algorithm amadala we-logic algorithm futhi isebenza kangcono kakhulu uma kukhulunywa ngokuqamba kabusha i-raster subsample entweni iso lomuntu elilindele ukuyibona.
Ngeshwa INvidia DLSS iphethwe futhi idinga i-Hardware ekhethekile kumakhadi we-Nvidia amasha (Uchungechunge lwe-RTX 2000 nangaphezulu), ngaphezu kweqiniso lokuthi uNvidia akasinikanga amandla lesi sici kubashayeli babo bendabuko beLinux, nabo abanobunikazi.
Ngokusho kwabanye abahlaziyi, lobu buchwepheshe buzothakazelisa, njengoba iValve kubikwa ukuthi icabanga ukwenza ithuluzi lokudlala ngesandla.
Siphikise ukuthi i-DLSS ingavumela i-next-gen switch ukuthi isebenze kahle ngaphezu kwesisindo sayo, futhi okufanayo kuzokwenzeka nge-laptop ngaphandle kwetoni yamandla wehluzo, okungenzeka kusebenze i-Linux.
Ku-Windows, i-DLSS ingesinye sezici eziningi zeNvidia ezenza kube nzima ukucubungula ukuthuthela ekhadini lemidwebo leRadeon, noma ngabe intengo ilungile nekhadi linamandla. Ku-Linux, izindima zibuyiselwe emuva futhi kunzima kakhulu ukukhetha uNvidia.
I-AMD ivule amashayeli ayo eRadeon weLinux ngo-2015, isebenzisa imodyuli evulekile nevulekile ye-AMDGPU kernel module lapho ithuthukise khona kakhulu ikhwalithi yabashayeli, yenza imidwebo yeRadeon ibe yisinqumo esingcono kakhulu samaGPU asebenza kahle emhlabeni.
Kwabanye, noma ngabe i-DLSS ibisebenzisana nayo yonke imidlalo, "esikhundleni sama-50 noma ama-60 kuphela, kungaba nzima ukukuyeka konke lokho ukuze kukhuphuke izinga lohlaka."
Ubuchwepheshe be-AMD DLSS busemgwaqeninoma. KwaComputex 2021, i-AMD imemezele inguqulo yayo yesampuli eyenziwe ngcono nge-AI, eyibiza ngeFidelityFX Super Resolution (FSR). Kuze kube namuhla, ukusebenza kwe-FSR akwaziwa. Kuyathakazelisa ukuthi i-FSR nayo ingasebenza kuma-Nvidia GPUs, ngisho nalezo ezingasekeli i-DLV yeNvidia.
Ngeshwa, I-FSR iseseyisithembiso nje ngalesi sikhathingoba ngeke ikhishwe kuze kube nguJuni 22nd futhi akucaci ukuthi izotholakala ngokushesha yini kwiLinux ngosuku lokwethulwa.
“Futhi asinawo amasampula wekhwalithi yesithombe amaningi ngaphambi nangemuva njengoba besingathanda. Uma i-FSR ingakwazi ukuncintisana ne-DLSS ngokwezinga lekhwalithi, ngeke kubaluleke kakhulu uma i-FSR ihlangana noma idlula ngisho nezinga layo lokuvuselela okuluhlaza, ”kusho i-AMD.
Yize uNvidia esho ukuthi ukwesekwa kweVulkan kuzofika kule nyanga futhi ukwesekwa kweDirectX kuzofika ekwindla, inkampani ayizange isho umugqa wesikhathi we-DLSS wokuza eProton. Kepha kuhle ukubona ukuthi iyaqhubeka nokududula imidlalo yeLinux ukuze iphile ngokuhlangenwe nakho kweWindows.