• 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.
Panos said:
Yup. Is bonkers, also AMD is very conservative to their announcement of 50% (as usual)
If all we know is accurate, including MS announcement, doing the maths 12tf at 175w means 15.7-16tf at 225w a +70% perf over the 5700XT level card!!!!!
A 275-279W card (same power level with 2080Ti) would be 20tf. Over twice as fast than the 5700XT. :eek:

----------------
Before someone starts blabbing about Tf isn't indicator blah blah, GCN TF, blah blah, Vega blah.
We talking about RDNA 1 and RDNA 2. Worst case scenario the performance between the two is 1-1 as is the same architecture in it's core. There is absolutely no chance RDNA2 terraflop been slower in real gaming performance than RDNA1 terraflop.....
Click to expand...

While I hope you are right, I wouldn't get too far ahead of yourself until we have actual silicon producing this performance.
 
Sargatanas2511 said:
While I hope you are right, I wouldn't get too far ahead of yourself until we have actual silicon producing this performance.
Click to expand...

I am using what we know from MS about Xbox GPU specs (12tf 175W RDNA2 APU) and how the 5700XT performs in relation to the rest of the cards.

Did I used any of the usual "big-Navi" WCCFTech click-bait crap? No
If I did used any "fantasy" specs making me be "far ahead of myself" as you put it, please. Show me the err on my post. :)
 
Sargatanas2511 said:
While I hope you are right, I wouldn't get too far ahead of yourself until we have actual silicon producing this performance.
Click to expand...

Exactly.
We don't know what exactly the graphics power draw of a new Xbox is stated for.
Is it just the die or is it the die+vram+power delivery.?
Because 225W for the 5700XT is for the entire graphics card which means the die itself will draw quite a bit less than that.
Since Xbox is sharing memory with CPU than it is very likely the stated power draw is only for the die.
 
Panos said:
I am using what we know from MS about Xbox GPU specs (12tf 175W RDNA2 APU) and how the 5700XT performs in relation to the rest of the cards.

Did I used any of the usual "big-Navi" WCCFTech click-bait crap? No
If I did used any "fantasy" specs making me be "far ahead of myself" as you put it, please. Show me the err on my post. :)
Click to expand...

Simple maths don't always work with things like GPUs as even if the processing power should be there in theory there might be a bottleneck elsewhere in the GPU preventing it from being used fully.

This thread is another hype train which might leave a lot of people disappointed in the end...
 
z10m said:
Simple maths don't always work with things like GPUs as even if the processing power should be there in theory there might be a bottleneck elsewhere in the GPU preventing it from being used fully.

This thread is another hype train which might leave a lot of people disappointed in the end...
Click to expand...

Completely agree on this. The VRAM bottleneck on the 5700XT cripples the card performance at 4K. But my assumption is that there isn't such bottleneck and scale linear.
 
Panos said:
Completely agree on this. The VRAM bottleneck on the 5700XT cripples the card performance at 4K. But my assumption is that there isn't such bottleneck and scale linear.
Click to expand...

Not sure this is the case, I mean the 2080Super has the same 8GB of VRAM so why would the 5700XT be more bottlenecked by the same amount? I just think it's not quick enough.
 
Sargatanas2511 said:
Not sure this is the case, I mean the 2080Super has the same 8GB of VRAM so why would the 5700XT be more bottlenecked by the same amount? I just think it's not quick enough.
Click to expand...
yeah same here, I think it just isn't built to handle that workload and has been given a clock speed bump out the box to square up with the competition to justify a higher price.
 
BigBANGtheory said:
yeah same here, I think it just isn't built to handle that workload and has been given a clock speed bump out the box to square up with the competition to justify a higher price.
Click to expand...

I actually like the 5700XT as a product, the performance and price are great considering the market we are in. I don't really feel any card can properly do 4k other than the 2080Ti currently though.
 
Panos said:
Yup. Is bonkers, also AMD is very conservative to their announcement of 50% (as usual)
If all we know is accurate, including MS announcement, doing the maths 12tf at 175w means 15.7-16tf at 225w a +70% perf over the 5700XT level card!!!!!
A 275-279W card (same power level with 2080Ti) would be 20tf. Over twice as fast than the 5700XT. :eek:

----------------
Before someone starts blabbing about Tf isn't indicator blah blah, GCN TF, blah blah, Vega blah.
We talking about RDNA 1 and RDNA 2. Worst case scenario the performance between the two is 1-1 as is the same architecture in it's core. There is absolutely no chance RDNA2 terraflop been slower in real gaming performance than RDNA1 terraflop.....
Click to expand...

Woah!! calm down. We know nothing really. These are just figures on AMD slides.

And, working out performance based on AMD's slides hasn't worked out too well in the past. Remember the RX 480? People were working it out to be faster than the 980Ti based on a slide AMD showed. Remember before the Vega launch? When AMD released the specs of the Frontier Edition, some people started estimating the performance of the Vega 64 to be around 1080Ti levels.

That's the one thing we have learned from the last few AMD launches. Performance is always less than the leaked figures/slides would suggest.
 
melmac said:
Woah!! calm down. We know nothing really. These are just figures on AMD slides.

And, working out performance based on AMD's slides hasn't worked out too well in the past. Remember the RX 480? People were working it out to be faster than the 980Ti based on a slide AMD showed. Remember before the Vega launch? When AMD released the specs of the Frontier Edition, some people started estimating the performance of the Vega 64 to be around 1080Ti levels.

That's the one thing we have learned from the last few AMD launches. Performance is always less than the leaked figures/slides would suggest.
Click to expand...
Yep. I am keeping my expectations low. I would rather be surprised than be disappointed again. I do think they should easily be able to beat Turing, but that is not saying much as that is on 12nm and been out for nearly 2 years!

The only way I would be surprised is if we get 2080Ti + 40-50% performance.
 
I have yet to look at it, just got home.... but this guy is already calling it "The RTX 2080TI destroyer" Click bait title but any truth to it?

 
sigh, what a tool.

yet another self proclaimed expert shouting about things while having sod all info on what the new cards or gpus will even look like.
 
humbug said:
I have yet to look at it, just got home.... but this guy is already calling it "The RTX 2080TI destroyer" Click bait title but any truth to it?

Click to expand...
Big Navi's die is rumoured to be 500MM+ in size, using the latest design and most advanced process node available. If it can't beat a RTX 2080Ti with all those advantages then AMD may as well just give up on GPU's altogether.

GAC said:
sigh, what a tool.

yet another self proclaimed expert shouting about things while having sod all info on what the new cards or gpus will even look like.
Click to expand...

To be fair I don't think he has ever proclaimed himself to be an expert.
 
Freddie1980 said:
Big Navi's die is rumoured to be 500MM+ in size, using the latest design and most advanced process node available. If it can't beat a RTX 2080Ti with all those advantages then AMD may as well just give up on GPU's altogether.
To be fair I don't think he has ever proclaimed himself to be an expert.
Click to expand...

If its that big its two times a 5700XT which would make it 50% faster than an RTX 2080TI.

LoadsaMoney said:
R.I.P Nvidia! :D
Click to expand...

No, the RTX Titan has about 5% more Cuda cores than a 2080TI with a 12nm size of 775mm, which is massive, to make the same size die on 7nm gives you and extra 70% transistors, so assuming everything is grown by 70% that = 7835 Cuda cores, there have been rumours of Ampre being over 7000 cores.

Assuming no IPC or clock speed gains from Nvidia or AMD that would make the best RTX 3000 series about 20% faster than the best Big Navi, which to be fair to AMD that's a lot better than the 45% the 2080TI has over the 5700XT now.
 
Be that as it may ^^^^ it all comes down to price, if AMD want to gain significant market share off Nvidia they have to go for a price to performance level that Nvidia are just not willing to compete with.

AMD's die will be much smaller than Nvidia's, if AMD can get 80% the performance of Nvidia at 60% the price they are on to a winner, Nvidia will want to price thier new <8000 core RTX 3080TI at a lot higher than the 2080TI, its going to be much more expensive to make, the 2080TI was £1200, i'm thinking they are going to want to have the 3080TI at between £1500 and £1600, if AMD do the Big Navi at £999 they will make Nvidia very uncomfortable.
 
greatash said:
you would hope it would destroy the 2080ti, being 2 years later. its competition will be the 3080ti, not the 2080ti
Click to expand...
Yeah. But I get the feeling assuming it comes out before Ampere, it will beat the 2080Ti and ask for £999.

To me 2080Ti performance in 2020 should be no more than £499. I mean it sounds silly but we are talking about less than 30-35% more performance over a 2070S here. That is not a lot.
 
Status
Not open for further replies.
Back
Top Bottom