• 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.
Nope most Ive got is 4.35 and most of the time my single core boost is 4.325 no matter what I try. I tend to lose the silicon lottery

My 2700x in the wifes boosts harder as well. Its almost always around 4.4 on boost. Sucks to lose every time, I have been super lucky with silicon since the q6600 where I got a massively low vid chip.
 
OK for some giggles, Navi could have a part with more than 80cu's, everybody is just thinking 2x5700xt. I don't think 80cu's is the highest end Navi.

oof..... i think at over 500mm2 Big Navi is the largest die AMD have ever made already, i could be wrong?

Nvidia are normally the ones going as big as necessary to win, there is a point in die size that i think AMD are just not willing to go to as yields fall off a cliff and people just aren't willing to pay Nvidia money for AMD GPU's.

Unless i'm wrong about that, at this point i would like to see AMD try, i would like to see them stop ####'### about and just throw caution to the wind and make the biggest fastest GPU they can and try the best they can to take the "performance crown" kudos.

If reviewers use AMD GPU's to test CPU bottlenecks instead of Nvidia that would be a massive and free marketing tool for AMD... this is what Nvidia currently enjoy, they don't have to say anything at all... just let reviewers explain why they use GPU X for CPU performance testing.

AMD have taken on Intel and have in that sense won, the world's fastest CPU is an AMD, and by a country mile... tho reviewers don't use them to test GPU performance 'YET'

If in RDNA2 AMD have the architecture to take Nvidia head on they should throw everything they have at it with the goal of beating Nvidia, if they have a chance now, take it.
 
New video just posted by MLID:


Tl;dr; leaked info stating 80cu, 14GB @ 448 bit for 1008gbps, thinks it could be an internal leak. (edit) but also gave reasons why he thought it was fake, including both companies 'leaking' info to confuse the other.

Average 31.9% over 2080ti and able to use DXR with <10% fps hit. (edit: BF5, Metro, SOTTR, Wolfenstein, COD)

One thing popped in to my head while watching regarding branding;
When AMD launched Ryzen, they poked at Intel with the chipset names (x370 vs z270, x399 vs x299), so they could do the same jab at the RTX brand by going to XTR (to also signify RT capability)

Air and liquid cooled 5700 lol
 
Last edited:
No that reward would go to Fury X at 596mm2. Let's pray that big Navi is better in every way compared to that gpu.

Ah...... thanks :) what's that another 20 CU's to make up 100?

Can't see it but i wished AMD just would, just do it for the Kudos, take a chance.
 
BTW: that's a 2.23Ghz 5700 is a box the size of a large book.

This thing clocks, i don't think as far as 2.76Ghz but why not 2.4Ghz? in desktop form? and its power efficient.

The higher clocks alone make it a pretty good prospect, 2.4Ghz is 30% higher than the reference 5700XT.

Agreed.

Nvidia will need to go even quicker to protect their margins and that's why we are seeing leaks of some massive Nvidia coolers.

I think Nvidia will still be quicker product-for-product, especially ray trace, but at the cost of power/heat and eye watering pricetag(s). Due to the PS5/XBOX/Navi 2020 cabal I can't see Nvidia fielding a product with direct performance parity without big price drops. They won't want to do that and hence, here come the three slot Nvidia air coolers.

Personally, I think the PS5/XBOX performance leaks have jolted Nvidia into a mild sweat and they are now scrambling for another 15% of perf. out of the 3000 series.
 
Agreed.

Nvidia will need to go even quicker to protect their margins and that's why we are seeing leaks of some massive Nvidia coolers.

I think Nvidia will still be quicker product-for-product, especially ray trace, but at the cost of power/heat and eye watering pricetag(s). Due to the PS5/XBOX/Navi 2020 cabal I can't see Nvidia fielding a product with direct performance parity without big price drops. They won't want to do that and hence, here come the three slot Nvidia air coolers.

Personally, I think the PS5/XBOX performance leaks have jolted Nvidia into a mild sweat and they are now scrambling for another 15% of perf. out of the 3000 series.

I think you're right. Nvidia will need to be able to say buy their cards and you will have the best graphical quality money can buy. That's the only way to justify high prices in the face of the console competition. So long as they are faster they can charge as much as the market will stand.
 
oof..... i think at over 500mm2 Big Navi is the largest die AMD have ever made already, i could be wrong?

Nvidia are normally the ones going as big as necessary to win, there is a point in die size that i think AMD are just not willing to go to as yields fall off a cliff and people just aren't willing to pay Nvidia money for AMD GPU's.

Unless i'm wrong about that, at this point i would like to see AMD try, i would like to see them stop ####'### about and just throw caution to the wind and make the biggest fastest GPU they can and try the best they can to take the "performance crown" kudos.

If reviewers use AMD GPU's to test CPU bottlenecks instead of Nvidia that would be a massive and free marketing tool for AMD... this is what Nvidia currently enjoy, they don't have to say anything at all... just let reviewers explain why they use GPU X for CPU performance testing.

AMD have taken on Intel and have in that sense won, the world's fastest CPU is an AMD, and by a country mile... tho reviewers don't use them to test GPU performance 'YET'

If in RDNA2 AMD have the architecture to take Nvidia head on they should throw everything they have at it with the goal of beating Nvidia, if they have a chance now, take it.

We are assuming that two navi's (251mm2) joined together is roughly 2x navi (500mm2)... we are also making assumptions on redundancy in the silicon parts of the gpu that you aren't just doubling up etc also all assumptions are based around what we know of rdna1 not rdna2 and its changes. I seem to read everywhere that 80cu's is 500mm2 based on nothing but pure speculation. Now if we make a different assumption things change somewhat and all of a sudden there is a little more space, I don't think necessarily that 500mm2 = 80 cu is what I am saying! 500mm2 could be 96 CU's for example with some disabled.

Salty speculation for now but 80cu's sounds like a start to me.
 
Last edited:
Status
Not open for further replies.
Back
Top Bottom