AMD Radeon RDNA2 Navi (RX 6500, 6600, 6700, 6800, 6900 XT)

It's a bit odd that they don't know which code parts to concentrate on. FXAA has been fantastically successful, and has been released in who knows how many games.

A hyper compact machine learning version might be similarly successful. Teach it to interpolate local gradients over some arbitrary neighborhood so upsampling can look better nigh as a plugin.

Wouldn't compete with DLSS but would be universal. They can compete with DLSS in other, somewhat less universal ways. Any TAA should involve history rejection and history rectification (reshading). Since both end up being guesses anyway, and bad to good implementations are often the difference between bad and good TAA, this seems a good target for AMD and a hopefully compact nn. Other than a code example of TAA upsampling, which like with DLSS needs direct developer involvement to implement anyway, that's the most logical way forward I can see for them.
 
Why do people compare full die to a fairly generous chop and go >hurr >durr >bad value.
?????
That's not how it works.
Maybe people don't look at it not from a company's perspective but from what they get for their money. To them big, little, process tech, cut, full die and whatnot do not yield much, they only see the dollars asked for it and the fps they get out of it.
 
It's 4-8% faster in rasterization and much slower in RT and lacks DLSS compared to the competitor's 400$ GPU. I fail to see how the 480$ price is at all justifiable.
And not everyone shares your values for each feature, not to mention the countless other features from both manufacturers you chose not to list.
 
DLSS needs direct developer involvement to implement anyway, that's the most logical way forward I can see for them.
Not entirely true.
Implemented in less than a day with no assistance from NVIDIA, DLSS’s recently-released Unreal Engine 4 plug ‘n’ play plugin delivers performance boosts of up to 1.6x at 4K. Never before has a plugin enabled developers to quickly and easily accelerate the performance of their titles for millions of gamers to such a degree.

“Adding NVIDIA DLSS to The Fabled Woods was easy thanks to the Unreal Engine 4 plugin, and the impact it makes on performance is substantial.” Joe Bauer, Founder CyberPunch Studios. “With the Unreal Engine 4 plugin, adding DLSS to The Fabled Woods was a no-brainer; it really opens DLSS up to a whole new world of developers.”
 
And not everyone shares your values for each feature, not to mention the countless other features from both manufacturers you chose not to list.
I’m not a huge proponent of RT given what it currently offers but it has to be considered. DLSS also has to be considered despite it being in very few titles. What features does AMD offer that Nvidia doesn't?
 
6700XT comparative performance is heavily benchmark dependent.
TPU has it close to 3070 now after they've updated their suite with titles like ACV.
PCGH has it on par with 3060Ti essentially because they test other games besides newest console ports.
All things considered $480 is feeling a bit high for what's on offer. The value discussion is fairly pointless right now anyway though.
 
IMHO, RX 6700 XT is overpriced not really compared to RTX - here, it's more personal prefs for or against various features, policies and/or quirks - but first and foremost compared to RX 6800. At least going by MSRP.

With 6800 you get 50% more of basically everything except clocks. It's marginally more power intensive (real world, we measured a 4-watt difference. 6800 has more memory reserves, a fast USB-C-Port (;)) and it's running more quiet probably thanks to it's third fan.
 
Then they get to patiently wait until the chops launch.
Simple!
Almost as simple as "Why do people compare full die to a fairly generous chop and go >hurr >durr >bad value.", no? Why would consumer not complain about bad value, when comparing two offerings? They don't care about G6X, 7 or 8 nm and yield-binning. They see what they get and what they pay for it - it's really THAT simple.
 
Back
Top