I have several friends that work at AMD and NVIDIA in the graphics (hw and sw) groups. They’re all in a slow transition to other compute groups. Their comments are that once we got ray tracing there wasn’t much in terms of hardware optimization for them other than increasing memory bandwidth. Everything else comes down to software.
Multi frame generation and dlss has made give improvements. It’s pretty clear that with a little more improvements, gpus will be able to output 5-10x the data with the same hardware. I’m not surprised they are focusing on that right now.
I can expect 5-10% hardware improvements, but the sugar e improvements on the 50 series is amazing. The 50 series is the new 1080ti in terms of the jump.
Multi frame generation was already possible with Lossless. The fact that it’s only officially supported by one generation of cards is absolute bullshit.
Lossless adds lag, mfg uses a whole different gpu pipeline. They are not even close to the same performance. I’ve used both and it’a night and day in terms of total system latency. I would never use lossless for gaming that requires input speed. Now using lossless for visual novels to limit cpu demand, but that is about it
I have several friends that work at AMD and NVIDIA in the graphics (hw and sw) groups. They’re all in a slow transition to other compute groups. Their comments are that once we got ray tracing there wasn’t much in terms of hardware optimization for them other than increasing memory bandwidth. Everything else comes down to software.
Multi frame generation and dlss has made give improvements. It’s pretty clear that with a little more improvements, gpus will be able to output 5-10x the data with the same hardware. I’m not surprised they are focusing on that right now.
Maybe 5-10%?
I can expect 5-10% hardware improvements, but the sugar e improvements on the 50 series is amazing. The 50 series is the new 1080ti in terms of the jump.
Output 5-10x with same HW? I need what you’re smoking because the current HW can barely handle current amounts of data, let alone a magnitude more
Multi frame generation was already possible with Lossless. The fact that it’s only officially supported by one generation of cards is absolute bullshit.
Lossless adds lag, mfg uses a whole different gpu pipeline. They are not even close to the same performance. I’ve used both and it’a night and day in terms of total system latency. I would never use lossless for gaming that requires input speed. Now using lossless for visual novels to limit cpu demand, but that is about it