• Competitor rules

    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.

AMD Navi 23 ‘NVIDIA Killer’ GPU Rumored to Support Hardware Ray Tracing, Coming Next Year

Status
Not open for further replies.
Yeah they are adept in their machine, and those big hitters like cyberpunk will bag them many punters. I also think the noise cancelling thing is really good, I would use that for sure.


if they can keep that promise, sadly they will get my money as it will cover up the weakness of VRAM (which will only show its face in a very few games) because they have DLSS support.

I feel (probably wrong as I'm a dumbass) in the past year, they dropped the ball on a few titles which were released without DLSS 2.0 support. whos fault is it? the developers? nvidias? nvidia need to work with these guys.
 
Do we think AMD will intentionally leak some details before the 17th to try and steal some thunder from Nvidia and encourage people to hold off on buying the 3080?

Nvidia leaks had been steady for months, yet it seems AMD are very quiet.
 
He thinks Big Navi(Navi 21) will be slightly faster than a RTX3080,board power of around 275W,the GPU will probably be under 2GHZ and it will use HBM2. He also goes on about how AMD can't design proper stock coolers and should get their AIB to design it for them,and how AMD can't seem to get performance drivers out for at least 3~6 months after launch,and leaves performance on the table.

I wish I could believe that they would choose HBM again, but I have no hope. I think they'll just use GDDR6 again & try and lower price as much as possible. I think he raises a good point about Ryzen margins, it's probably a big factor, that's why I don't think they're that worried about how many GPUs they move.
 
I wish I could believe that they would choose HBM again, but I have no hope. I think they'll just use GDDR6 again & try and lower price as much as possible. I think he raises a good point about Ryzen margins, it's probably a big factor, that's why I don't think they're that worried about how many GPUs they move.

HBM really doesn't offer much for gaming - though it is strong for certain types of compute and other usage.
 
HBM really doesn't offer much for gaming - though it is strong for certain types of compute and other usage.

I disagree, it offers a lot of bandwidth which is crucial for high res & ray tracing. Nvidia maybe doesn't care because they have GDDR6X now but what options does AMD have with GDDR6? No matter how you slice it their memory bandwidth will be middle of the pack because there's just no way they're doing a bus larger than 384bit. Best case scenario they get 16 Gbps chips and that gets them 12 GB @ 768 GB/s. 8 GB/s more than the 3080, but far below both their previous best with the VII, as well as compared to the 3090 at 936 GB/s.
And maybe that doesn't sound too bad but last I checked AMD was still behind Nvidia wrt memory compression, which means they need more than their rivals in order to keep up. And of course, whether those 16 Gbps chips are plentiful enough or cheap enough, I don't know, and the costs are higher still due to 384 vs 320bit.

Could it still be competitive? Sure, but at what cost?

Personally though I'm just disappointed because at least HBM was a nice differentiating factor compared to Nvidia's GDDR obsession. And differentiating factors are something which have greatly decreased in number for the red team. I hated Navi because to me it looked like they're making crappier discounted Geforce cards now, and that's not ever going to sell me on it. I really hope this won't be a repeat of that situation.
 
I disagree, it offers a lot of bandwidth which is crucial for high res & ray tracing.

Problem is application - when it comes to game type work loads it is hard to exploit the benefits of HBM over GDDR so you end up with more cost and complexity for the same performance - you can't necessarily put those benefits to work in RT application, etc. or even in some cases it isn't as optimal as GDDR. Grossly simplified but when you are working with time sensitivities like your average game frame it presents a different situation and a different set of challenges compared to your average compute task which might have time sensitivities in seconds or 10s of seconds rather than milliseconds.
 
Do we think AMD will intentionally leak some details before the 17th to try and steal some thunder from Nvidia and encourage people to hold off on buying the 3080?

Nvidia leaks had been steady for months, yet it seems AMD are very quiet.


If they have a product to match NVIDIA, they need to release details ASAP. Otherwise they'll lose a lot of potential sales to NVIDIA simply due to timing.
 
If they have a product to match NVIDIA, they need to release details ASAP. Otherwise they'll lose a lot of potential sales to NVIDIA simply due to timing.

Nobody has or can buy anything yet. I'd release information just before the 17th for maximum impact and minimum time for Nvidia to respond. Have doubt in consumers minds when they're thinking of hitting that buy button ;)
 
Nobody has or can buy anything yet. I'd release information just before the 17th for maximum impact and minimum time for Nvidia to respond. Have doubt in consumers minds when they're thinking of hitting that buy button ;)
It does raise the question as to why Nvidia held back the opportunity to order these cards.

This doesn't seem to benefit Nvidia in anyway unless they plan on changing something before launch?

I wonder if Nvidia made their move hoping to force AMD to panic and start leaking stuff. Then they can respond just before launch.
 
It does raise the question as to why Nvidia held back the opportunity to order these cards.

This doesn't seem to benefit Nvidia in anyway unless they plan on changing something before launch?

I wonder if Nvidia made their move hoping to force AMD to panic and start leaking stuff. Then they can respond just before launch.

I think that could well be the case. I'm not sure that if a credible leak emerged on pricing, performance etc. that Nvidia would likely shift their pricing at the last minute (although it's possible), but rather I think it's likely to simply accelerate the leaks from AMD such that they can either prepare for a future price change before the AMD launch, or accelerate plans for marketing the Super/ Ti variants.

I really hope we get some decent leaked information soon otherwise if the 3080 proves to be as good as believed, it's a fantastic card for me and for many.
 
You're trolling right? I mean, seriously :rolleyes:
I've seen plenty of decent AMD coolers from Sapphire etc I just don't understand how in 2020 AMD are still showing us blower coolers
that are so unattractive especially for an upcoming high end launch. :rolleyes::rolleyes:. Maybe you think its a smart move, I don't.
 
I've seen plenty of decent AMD coolers from Sapphire etc I just don't understand how in 2020 AMD are still showing us blower coolers
that are so unattractive especially for an upcoming high end launch. :rolleyes::rolleyes:. Maybe you think its a smart move, I don't.

Because it's not a photo of a 2020 AMD release? :p
 
My personal prediction is amd will price aggressively but they don’t match the performance of the 3080.

likely will be stacked with a decent wedge of vram to compensate
 
Personally though I'm just disappointed because at least HBM was a nice differentiating factor compared to Nvidia's GDDR obsession. And differentiating factors are something which have greatly decreased in number for the red team. I hated Navi because to me it looked like they're making crappier discounted Geforce cards now, and that's not ever going to sell me on it. I really hope this won't be a repeat of that situation.

I know but who bought a Vega when it had HBM? There is no prizing away from the mindshare, and the single digit enthusiasts on forums that understand this wont bring much adoption. Vega only got traction when the prices dropped lower. Even if AMD get HBM on the high end stack there will be many users who still buy nvidia yannow just because "its works" and the marketing brainwashing - just look at the hype in the last few days.
 
My personal prediction is amd will price aggressively but they don’t match the performance of the 3080.

If it's just under I wouldn't worry too much, we know how AMD get performance from their cards over time. If it was £100 cheaper and came with 16GB of vRam then I'd def look at it as an option :)
 
I think what we will see is something near as dammit equalling the 3080 performance but with more vram and a slightly lower price with either option being a good choice but AMD not blowing the 3080 out of the water.
 
If it's just under I wouldn't worry too much, we know how AMD get performance from their cards over time. If it was £100 cheaper and came with 16GB of vRam then I'd def look at it as an option :)


I personally need it to match or I’m out.

I already feel in the intensive games which push GPu to the limit, there is a 50 percent chance dlss will fix it.

nvidias venture into dlss has really paid off now imo.
 
If it's just under I wouldn't worry too much, we know how AMD get performance from their cards over time. If it was £100 cheaper and came with 16GB of vRam then I'd def look at it as an option :)


Btw has anyone done an ‘how they aged’ video of Nvidia vs amd for last gen?
I feel like my 2080 got stronger and stronger this generation. Not sure if amd fine wine is still holding enough water .
 
Status
Not open for further replies.
Back
Top Bottom