Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Thats exactly what used to happen, until Nvidia chopped out the circuitry, why, god only knows but presumably to save a few bucks per card. Crazy that a cheap toaster has better protection than a £2000 premium GPU.
I wouldnt say it was simple interpolation since the GPU has access to all the data such as motion vectors and there is a cost to FG but it looks to me like the input latency is always more than two frames anyway. So the best can do is try to streamline the process for minimal latency addition or brute force it with a higher base rate.
I wouldnt say it was simple interpolation since the GPU has access to all the data such as motion vectors and there is a cost to FG but it looks to me like the input latency is always more than two frames anyway. So the best can do is try to streamline the process for minimal latency addition or brute force it with a higher base rate.
105 FPS in Cyberpunk to me is like 22.2ms latency - add FG and its ~174 at 26.4ms - those are nvidia's overlay number though
These people are NVIDIA themselves... You can go raise it with them that you don't believe them.
You can check this article and all the NVIDIA slides about FG: https://hothardware.com/reviews/nvidia-geforce-rtx-40-architecture-overview For FG to work it needed to use Optical Flow Accelerators, to avoid big artefacts. These need 2 frames to compare, do calculations of how stuff move and then AI can generate (interpolate) frame in between quickly, insert it and then display the second real frame. With new algorithm they got rid of optical flow and instead used pure AI method, but it's still just comparing 2 real frames and then interpolates frames in between. It just does it in a smarter way than TVs do.It's frame interpolation is bit more complex than what TVs are doing, as it's optimised for latency (hence, Reflex is required). TVs don't care about latency at all, hence their own interpolation is useless for gaming (way too high input lag). Also, games aren't responding to input with every single frame with 16.6ms latency at 60FPS. Use PresentMon tool or at least NVIDIA's built in stats to see actual input latency of the game - to get it down to 16.6ms in my own experience, game needs to run natively at about 175FPS, not 60FPS. That also depends on the specific engine. When you look at how long it took for the GPU to render a frame with old school tools, you don't actually measure FG at all, you just measure real frame time usually. Gamer Nexus has few good videos about it and PresentMon, discussing difficulty of measuring real latency introduced by FG and it gets even harder with MFG.
That's not how it works, though.
FG and MFG don't work in VR for a reason. Reprojection used in VR is similar to what Reflex 2 is doing, not FG. Completely different things.
Hi mate, Was wondering if you know where to download the Palit 5080 Gamerock OC bios. Couldn't find it anywhereLooks like a new nvflash 5.867.0 works with series 5. I'll be throwing the 600w OC bios onto my Palit when it arrives. That 25w and 1% gain wouldve really bugged me
Hey, the bios usually get uploaded to techpowerup bios list. There's only 3 or 4 5090 bios available on there atm. The Palit Gamerock oc bios hasn't been uploaded by anyone yet. The ability to extract and flash the 5 series has only been a few days and im not even sure ive seen an OC in the wild yet. So over the coming weeks, i'm sure it will be uploaded by someone.
Welcome to modern era graphics. It has been that way since the 30 series.
According to TPU it's ~12% faster for a card that came out over 3 years later with much less VRAM. After that length of time that's hardly outstanding progress is it?
I agree , it's one of the outlying examples but part of that is just because of how bad value the 3090 was in the first place. ~15% faster than the 3080 for more than double the money.
Yea, I posted that yesterday. Very much a case of 'i'll believe it when I see it'. MSRP 5090s are >6 months off as far as i'm concerned.apparently in a month there is going to be an excess of stock?
RTX 5090 supplies to be 'stupidly high' next month as GB200 wafers get repurposed, asserts leaker
Nvidia's Blackwell data center demand dip could mean scalpers have to take a loss.www.tomshardware.com
apparently in a month there is going to be an excess of stock?
RTX 5090 supplies to be 'stupidly high' next month as GB200 wafers get repurposed, asserts leaker
Nvidia's Blackwell data center demand dip could mean scalpers have to take a loss.
Can I ask what Driver and settings you're using to achieve these please?This is why I like to use modern tools to measure not just frame latency but game engine latency, along with GPU Busy. Ergo, nothing beats Intel's Present Mon currently (it's Open Source too), though even that has some trouble with FG. NVIDIA seems to have their own tool based on Present Mon but they modified it and did not made any changes public, so it's a bit of a black box, hence I avoid it.
Example of what I see in CP2077 just now on my 4090 - exact settings, scene etc. irrelevant (it's with all settings to max, including PT and DLSS Quality), just pure comparison of Reflex on/off and FG:
Reflex off, FG off: 72FPS, 50ms game latency
Reflex on, FG off: 72FPS, 26ms game latency
Reflex on, FG on: 125FPS, 36ms game latency (and frame time itself is about 16ms)
Reflex by itself cuts game latency by half, then FG adds 10ms on top of it (still with Reflex, which matter a lot with FG). Frame pacing isn't ideal with FG, but playable with base 72FPS, even though I can already feel game lagging a bit whilst using my mouse. Adding game latency, mouse and KB latency and monitor latency and we're looking (even with Reflex) into 60ms+ of overall input lag. Humans need on average over 250ms of time to react to things changing on the screen, which this puts in summary at over 300ms of input lag including humans. Younger gamers are quite a bit below 200ms reaction time, though, apparently.
