• 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.

RDNA 3 rumours Q3/4 2022

Status
Not open for further replies.
pc-guy said:
No bench against 4090 disappointed!
Click to expand...

It wouldn't be good marketing to show a competitors card out perform you.

Would be kinda dumb to release a new product and say oh look our product is second best.

If they were to mention nvidia they would likely be better comparing to a 3090ti then saying oh yeah we can't beat their new card but we are 40% cheaper.
 
Last edited:
humbug said:
$999 for what looks like 4090 comparable raster performance, a grand is not little money but its a lot less than $1700.
Click to expand...

Highly doubt it'll be comparable in rasterization. What's important is that it's 3080 performance in RT - so these cards are pretty much worthless and will sell even worse than the 6000 series I imagine.

Nvidia have 80% dedicated GPU market share, this'll probably be closer to 85% a year from now.
 
CAT-THE-FIFTH said:
Nvidia I suspect is paying TSMC more for the privilege - for example Turing used a specialised 16NM variant made for Nvidia,and Ada is using a customised TSMC 5NM called TSMC 4NM. Technically Nvidia is using a better version of TSMC 5NM than AMD is. ATI used to release before Nvidia on new nodes and so did AMD.Since Pascal that has changed.
Click to expand...
I very much doubt TSMC would be giving Nvidia preferential treatment over AMD.

AMD is a bigger customer than Nvidia, AMD use their fab on CPU and GPU...

I think that stuff is Nvidia marketing BS
 
Last edited:
brostradamus said:
an interesting comparison would be a 4090 at 355 w going head to head with amd
also some more details should be available on why amd couldnt increase clocks compared to rdna2
Click to expand...

The Bright Memory Infinite testing results are very similar to the overall chart. Performance barely changes at all, with up to 20% more power available due to clock speeds hitting their limit. Dropping the power limit to 80% decreases the power use by 14% (355W instead of 415W) while only reducing performance by 3%.
Click to expand...

www.tomshardware.com

Improving Nvidia RTX 4090 Efficiency Through Power Limiting

We tested the 4090 Founders Edition at 50% to 120% power limits
www.tomshardware.com www.tomshardware.com

3% performance reduction at 355W... apparently.
 
fs123 said:
How is AMD claiming 704fps at 4K when the 4090 cannot get past 600fps in any scenario with the best cpu's on the market? At 4K both cards should be gpu limited more than cpu limited.
Click to expand...

There's a hardware unboxed video that shows AMD getting 20% more performance when CPU bound as Nvidia get bottlenecked faster than AMD. So it's not unfeasable when you start to get those silly numbers.
 
If we put it this way, their perf is amazing for the price. Not to mention they look a lot cooler than the 4090. :)
https://twitter.com/nerdtechgasm/status/1588282343256907776
Fgq2WD-aYAAL_9D
 
Last edited:
fs123 said:
How is AMD claiming 704fps at 4K when the 4090 cannot get past 600fps in any scenario with the best cpu's on the market? At 4K both cards should be gpu limited more than cpu limited.
Click to expand...
Not at that frame rate... the resolution is not the defining factor there when the frame rate is so insanely high.

Run something like Half Life 2 at 8k and it'll be CPU limited...
 
Funny when you think back a few years, all the hoopla about hbm, it got used twice on amd products then ditched. Back then it was the next big thing, but never really seemed to take off, rocky start with fury x and manufacturing issues.
 
Status
Not open for further replies.
Back
Top Bottom