• 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.
Much as i'd love for proper competition again (i've never had a loyalty on either GPU or CPU in 30 years of owning PCs, just whatever looks the best fit for my wallet and needs at the time), something makes me wonder.

1. If the end of October launch is going to be a 'proper' launch, you'd assume the card is finalised and possibly even in production

2. Assuming 1 is correct, they'd have some in-house performance numbers by now and you'd think they'd be leaking those numbers like mad now if it was properly competitive or 'close enough' to encourage people not to drop cash on a 3080 order/pre-order? After all, every 3080/3090 sold is a potential AMD sale lost.

I suppose the other possibility is that its not the 3080/3090 space they're going after anyway, but possibly the 3070 is the competition? OK, that launches ahead of the AMD reveal as well although if supply is as bad as the two nVidia launches to date then no huge loss of time if AMD hit the ground running.

They probably DO have the performance numbers - but it is crazy to release anything right now due to the 3080/3090 frenzy. Let that settle down a little then release the info when the embers settle.
 
They probably DO have the performance numbers - but it is crazy to release anything right now due to the 3080/3090 frenzy. Let that settle down a little then release the info when the embers settle.

When did any performance numbers appear for the 3080 before Jensen's 1st Sep announcement?
 
I hope they can compete with the 3080 and 3090. Would be a little underwhelming if their high end is 3070 level.

Agree. I wouldnt be completely gutted other than the self-hype we all built, if it was a 3070 beater but thats it then it can cost the same as a 3070 or do one.
 
I'm probably tired and misread something... thought you were talking about AMD releasing performance numbers....

I mean, if we take the timeframe of the nvidia release (17th to the announcement on 1st, you can use the leaked performance numbers x days before the 1st) we can work backwards of when useful information is hopefully shared.
 
When did any performance numbers appear for the 3080 before Jensen's 1st Sep announcement?

No real urgency for them to do so though as they had no competition at that point in time. Even the date of the AMD keynote wasn't announced until a week after the 3080 unveil (Sept 9th I think?). Whereas AMD will be launching with the competition already out there (albeit with a stumbling start supply-wise).
 
I hope they can compete with the 3080 and 3090. Would be a little underwhelming if their high end is 3070 level.
But it wouldn't really be any different from this generation. - 30% or so. I'm hoping they get closer to Nvidia's fastest card this time and a 3080 competitor would be doing just that.
 
I dont think you could keep mentioning 'BIG' navi if they thought it was not gonna be something to challenge to be honest.

If you cast back to the MLID last part - he mentions they could have used regular DDR6 and got more bandwidth at a lower power consumption. Then relate that to the AMD lineup which he said was I think 250w.
 
If and when Big Navi is released, and assuming it can compete with 3080 (which I really think it will!)... it'll be the greatest uplift in a gen EVER I think... as where the 5700XT sits right now today, to have made that kind of ground up in ONE gen will be an incredible achievement from AMD... hoping they do I really do... basically it'll be a 100% performance increase to be hitting 3080 speeds... and that's a lot
 
If and when Big Navi is released, and assuming it can compete with 3080 (which I really think it will!)... it'll be the greatest uplift in a gen EVER I think... as where the 5700XT sits right now today, to have made that kind of ground up in ONE gen will be an incredible achievement from AMD... hoping they do I really do... basically it'll be a 100% performance increase to be hitting 3080 speeds... and that's a lot

Im hoping so for all the vega owners too as it means holding off on the 5700XT which wasnt really a decent leap up so agree it would be a huge step. Not sure if the Ryzen guys were drafted in to help but it sounds like they needed all hands on deck otherwise they would be way behind nvidia.
 
They're not allowed to rearrange the pcb? I thought asus had already done that with their strix card.

Not PCB that’s fair game to an extent but I heard that they are restricted in terms of the cooler, they can only do certain designs which may well be true given they are always triple fan etc etc nothing new... Presume they get forced to do something which makes the card different from NVs.
 
Not sure if the Ryzen guys were drafted in to help but it sounds like they needed all hands on deck otherwise they would be way behind nvidia.

supposedly some of the engineers specialising in power efficiency on zen processors worked on RDNA2

http://www.redgamingtech.com/rdna-2...vements-to-take-on-nvidia-analysis-exclusive/

So what does any of this have to do with AMD? Well, according to one of my sources the optimization works of Suzanne Plummer and her team over at Radeon Technology group is now paying dividends and the second generation of RDNA is a drastic improvement over the first because of this.

Back in 2018 Suzanne Plummer provided a few hints to this, “A lot of what we did in Zen was trying to push well beyond what we thought we could do,” says Plummer, “and I think that is something we’re trying to do in the graphics space as well to make a bigger leap forward.

“We’ve pulled in some of the expertise from the microprocessor cores team into the graphics team, kinda helping with our methodology, and improving our frequency and our performance and power. And just taking the best that we have already developed in-house and trying to make sure that we’re using the same improvements across the company.”

more in the link.
 
Im hoping so for all the vega owners too as it means holding off on the 5700XT which wasnt really a decent leap up so agree it would be a huge step. Not sure if the Ryzen guys were drafted in to help but it sounds like they needed all hands on deck otherwise they would be way behind nvidia.


5700xt\radeon vii were more or less sidegrades from a vega 64 at best. 64 is still a decent card though depending on what you play, though starting to show their age in some games.
 
I kind of like the Radeon VII, despite the ridiculous cost. It was a marked improvement over Vega 64 and the die shrink to 7nm resulted in a much smaller die size than Vega 64, and higher clocks too at a similar TDP.

EDIT - Apparently, there are 4 TMUs per CU, so a 80 CU GPU should have 320 TMUs (double the 5700 XT).

This means perhaps the most important thing we don't know is how many ROPs an 80 CU RDNA 2 GPU would have. Ideally, this would be about double the amount the 5700 XT has, which would be 128 ROPS.

The other factor affecting performance is memory bandwidth, I'm not worried about this really, I doubt using GDDR6 would hold back performance much, and they can use a larger memory bus width if needed.
 
Last edited:
Status
Not open for further replies.
Back
Top Bottom