AMD Vega 10, Vega 11, Vega 12 and Vega 20 Rumors and Discussion

you sortet best not top ;)
I didn't sort anything. I don't use reddit or have an account there, I merely clicked the link to it. Understand what I'm saying yet? It's like linking to a google search and expecting the results to be the same for everyone.
 
PCGamesHardware Germany has now put a message about the Vega features uproar onto their website. Please write Emails to the editorial staff of tech- and games magazines in order to get more public attention for this!

pcgameshardware. de/Vega-Codename-265481/News/AMD-Radeon-RX-Vega-Primitive-Shader-DSBR-Beschwerden-1245870/
reddit. com/r/Amd/comments/7js2wj/lets_make_some_noise_until_amd_will_answer/?sort=new

Sorry for having to post the links this way, but I dont have enough posts yet.
 
People upvoting rants without any minimal knowledge of what a shader is..
Or without knowing how much it will really add to performance. On paper some performance is anticipated, but most expectations have become overblown.

@Alessio1989 ... is this the beginning of *recursion*?
 
Or without knowing how much it will really add to performance. On paper some performance is anticipated, but most expectations have become overblown.

@Alessio1989 ... is this the beginning of *recursion*?
Well what did AMD advertise is would do to performance. And that's why marketers need to be shot! Just because 1 title out of 100 gets 2.6 boost doesn't mean that is the average gain and they shouldn't advertise it any other way. It just causes pain for everyone.

Sent from my SM-G950U using Tapatalk
 
They explicitly mention "new fast paths" in their geometry pipeline :O
Apple computers are more like consoles, maybe it's easier to get this stuff working in a more or less fixed platform than on PCs with its countless permutations of CPUs, northbridges and various software. :p
 
Outside bug-fix and performance issue, the only thing I really miss is the manual control of the output range (0-255 vs 16-235...)

People upvoting rants without any minimal knowledge of what a shader is..

No different than any other Dunning-Kruger effect for this kind of stuff. Years ago there was a lawsuit over Killzone: Shadowfall using early temporal reprojection to hit 1080p/60 in multiplayer, with people claiming that wasn't "native" 1080p and so the box lied to them. Lawsuit was obviously dismissed shortly.
 
I don't see any lawsuit becoming real, it makes no sense. But damn, someone from amd need to speak up. It won't happen, I know... How BTW, back in the day, did ATI explained what happened with R600 and MSAA ?
 
The AMD subreddit is now heavily moderated. Every new thread about "NGG Fast path" and "primitive shaders" gets deleted now. Time to start a petition on change.org and share it on forums....
 
I'm afraid of the adverse effect this petitioning or repetively demanding for answers from AMD may have. They may decide as policy not to offer any detail about any feature until a product is launched with it somewhat working.

So which is worse : less info or more , but potentially misleading?
 
I don't see any lawsuit becoming real, it makes no sense. But damn, someone from amd need to speak up. It won't happen, I know... How BTW, back in the day, did ATI explained what happened with R600 and MSAA ?
FWIW - it was already AMD back then. Just in case you're alluding to the alleged change in corporate culture.
 
No different than any other Dunning-Kruger effect for this kind of stuff. Years ago there was a lawsuit over Killzone: Shadowfall using early temporal reprojection to hit 1080p/60 in multiplayer, with people claiming that wasn't "native" 1080p and so the box lied to them. Lawsuit was obviously dismissed shortly.
At least native/reconstructed 1080p can have some level of material discrepancy in the delivered experience.
Unless you are a developer who bought Vega expressly to write primitive shaders or a specific project assuming the presence of an active DSBR, the idea is that their presence doesn't affect what a user actually sees.
What appreciable harm or loss is there in a product that is "faster than before, but a different kind of faster" or "that culled triangle was culled later and I almost saw it"?

I don't see any lawsuit becoming real, it makes no sense. But damn, someone from amd need to speak up. It won't happen, I know... How BTW, back in the day, did ATI explained what happened with R600 and MSAA ?
It's been a while, but I recall discussion about how RV770's density improved over R600 due to better libraries, which might be an after-the-fact recognition not everything was ready at the time.
I don't recall if AMD ever changed its story that R600 had chosen to not have hardware MSAA resolve, although most chatter about it was "yeah, it was broken".

The AMD subreddit is now heavily moderated. Every new thread about "NGG Fast path" and "primitive shaders" gets deleted now. Time to start a petition on change.org and share it on forums....
Were the original two threads concerning those features kept around? If so, it sounds less like it's heavy-handed moderation than it does spam cleanup.
 
More information has definitely been more entertaining. :)

(Pretty sure you were refering to the drama)

I find learning about dead ends to be interestind, being an engineer myself.

But we should definitely check the ammount of salt we use to take AMD's promises with
 
Back
Top