![amd radeon r9 380x amd radeon r9 380x](https://cdn.prohardver.hu/dl/uad/2021-12/5362715/img_20211127_122414.jpg)
GP106 (chip replacement for GTX 960) is not due out till fall, and should be around Polaris 10 performance. WE MIGHT SEE A 1060 Ti cut of GP104 if Polaris 10 is just too aggressively priced.
Amd radeon r9 380x full#
Polaris 10 should be well above full Tonga, somewhere around 390 performance. And Nvidia will need to actually supply the bus-powered GTX 950 cards they promised us two months ago to compete with Polaris 11 Polaris 11 presentation way back in march was targeted at GTX 950-level performance on around 50w., so GTX 960 and full Tonga live-on. Neither AMD not Nvidia re releasing GPUs in that performance segment until later in the year.
![amd radeon r9 380x amd radeon r9 380x](https://m.media-amazon.com/images/I/61XnO3G8dLL._AC_SL1024_.jpg)
so, what are you guys using? if this is covered somewhere can you point me to it? basically I'm trying to replicate your procedures at home because I know I can trust your methods. turn off error detect and it runs for hours but there has to be errors that are not visible. I can run heaven or valley or any game I play all day with what i'd call my highest stable OC but as soon as I test with OCCT with error detection it spits out errors instantly and constantly. We found these completely disappeared with the core set to 1040MHz." but what program/game shows the diminished results and or artifacts? I'm finding that what is "stable" varies greatly based on what is used to test. Once we passed 1040MHz we began seeing artifacts in game. I see info like: "Next we began adjust the core clock until performance began to diminish. Ive looked and looked and cant seem to find any info on what you guys use to validate your gpu OCs. Informative and to the point as always, nice review!