Kuleli sonto, uRaja Koduri, owengamele ukukhiqizwa kwe-GPU e-Intel, uvakashele imboni yakwaSamsung eSouth Korea. Njengoba kunikezwe okwakamuva AbakwaSamsung bamemezele ukuqala kokukhiqizwa kwama-chips angu-5nm kusetshenziswa i-EUV, abanye abahlaziyi bakholelwa ukuthi lokhu kuvakasha kungase kungabi ngengozi. Ochwepheshe baphakamisa ukuthi izinkampani zingangena kwinkontileka lapho abakwaSamsung bazokhiqiza khona ama-GPU amakhadi evidiyo e-Xe discrete azayo.
Uma kucatshangelwa iqiniso lokuthi i-Intel ibibhekene nobunzima obuhambisana nokushoda kwama-chips isikhathi eside, ukuvela kwamahemuhemu anjalo kulindelekile impela. Kungenzeka ukuthi i-Intel ihlela ukuhlinzeka ngomthamo owengeziwe wokukhiqiza ngokusebenzisa izimboni ze-Samsung. Ukwethulwa okuseduze kokuthengiswa kwamakhadi evidiyo e-Intel discrete kungase kube nzima ukushoda kwama-chips kakade ekuqaleni. Ungakugwema lokhu ngokwandisa umkhiqizo wakho noma uqale ukusebenzelana nomphakeli wenkontileka we-GPU ongahlinzeka ngenani elanele lezingxenye.
Ochwepheshe bakholelwa ukuthi ama-GPU amakhadi ehluzo e-Intel discrete esizayo kufanele enziwe kusetshenziswa ubuchwepheshe benqubo ye-nanometer eyi-10 noma engu-7-nanometer. Ngenxa yalokhu, imikhiqizo yale nkampani izokwazi ukuncintisana ne-AMD, okuthiwe kulo nyaka iqale ukukhiqizwa kwamakhadi evidiyo nge-7-nm GPU. Kungenzeka kakhulu, isizukulwane esilandelayo samakhadi evidiyo e-NVIDIA sizophinde sisekelwe kuma-GPU enziwe ngokuhambisana nobuchwepheshe benqubo ye-7nm.
Okwamanje, ukubambisana okungenzeka phakathi kwe-Intel ne-Samsung kusengamahemuhemu angase aqinisekiswe noma aphikwe ngokuzayo.
Source: 3dnews.ru