Caporegime
- Joined
- 18 Oct 2002
- Posts
- 29,852
Tbf the nVidia oem cards are now pretty well clocked (lot better than they used to be).
-
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
- Thread starter Gregster
- Start date
- Status
Tbf the nVidia oem cards are now pretty well clocked (lot better than they used to be).
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
I know same as AMD but I like to take the best boosting clocked card from the previous gen. Because usually, they are the 1st to go when a new card comes out. As they always skew results do to not being considered.
Furthermore, Lise Su her self told investors that they were competing in the higher end GPU market. So, AMD has a lot on the line. If they foul this up their stock price will reflect it.
Also HBM2e is now in high production. One has to wonder who the customers for it are.
Furthermore, Lise Su her self told investors that they were competing in the higher end GPU market. So, AMD has a lot on the line. If they foul this up their stock price will reflect it.
Also HBM2e is now in high production. One has to wonder who the customers for it are.
Last edited:
Caporegime
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
^
Here's a chart...
I still think we might see a 96CU variant.
The way I see it Big Navi, not some random new rdna 2 gpu, would have to be about 95% better then a 5700xt in order for the optics of it to views as next gen gpu.
Here's a chart...
I still think we might see a 96CU variant.
The way I see it Big Navi, not some random new rdna 2 gpu, would have to be about 95% better then a 5700xt in order for the optics of it to views as next gen gpu.
Associate
- Joined
- 21 Apr 2007
- Posts
- 2,487
Truth is we just don't know, I really hope Navi 2X is 80 CU or more runs at 2.2GHz+ and isn't a stoopid price. We are seeing so much made up nonsense atm nothing feels real anymore.
Fundamentally we need Nvidia to reveal their opening hand and AMD to respond, we might get some leaks from AIB partners but again the background noise is so high its near on impossible to filter fact from fiction.
Fundamentally we need Nvidia to reveal their opening hand and AMD to respond, we might get some leaks from AIB partners but again the background noise is so high its near on impossible to filter fact from fiction.
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
Whenever these cards come out the Kingpin 2080ti will be the litmus test not the OEM.
As that card is about 30% better then a 2080 ti OEM in timespy.
Last edited:
Associate
Well, if the FE is 350W like the leaks have said, it doesn't bode well for OC headroom.
What sites and what info?
Last I checked this is the first "info" on Navi performance. Everything prior to that was peoples speculation and labeled as such.
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
I think he's referring to those sites that shape 'news' for the clicks.
It's not even info, it's just more speculation that will likely change again next week. We've went from a 3090 competitor to 50% faster than 2080ti, to 15% faster than 2080ti. Throw enough **** at the wall and something will stick, and that's all that's occurring at present which is standard practice in the run up to gpu releases.
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
hmmm
Soldato
Why? Is Nvidia going to use a 512bit bus in its GA102 based GPUs?
No, but they also don't need to. 16 GB would've been a nice advantage for them, instead we go back to Navi = gimped geforce and no real reason to buy it over Nvidia (that we know of so far).
Soldato
What if AMD made a 24GB SKU then?? If anything,if AMD can get better use of memory bandwidth,ie,don't need a 512 bit memory bus that would be a good thing. GDDR memory controllers and VRAM take up a surprising amount of the power budget.
Also,looking at the AdoredTV twitter,Jim still says the card is 16GB,and Nvidia will launching Ampere before big Navi. I think until we start seeing PCB pictures,we have to take any of these leaks with a pinch of salt.
Soldato
- Joined
- 8 Jun 2018
- Posts
- 2,827
There is more then one sku of rdna 2 that will launch. And this leaker wasn’t able to ID what card he was referencing.
Caporegime
AMD might be delaying the largest SKU.
Nvidia always release the ##80 first with the ##80TI later, AMD might be doing the same, not necessarily through choice, they might need more design tweaking, require a more matured node for yields and power limitations.
Like More's Law is dead i find it hard to believe AMD can't do better than +20% of what they are putting in the XBox Series X.
So RTX 3080 competitor first, RTX 3080TI competitor later....
Nvidia always release the ##80 first with the ##80TI later, AMD might be doing the same, not necessarily through choice, they might need more design tweaking, require a more matured node for yields and power limitations.
Like More's Law is dead i find it hard to believe AMD can't do better than +20% of what they are putting in the XBox Series X.
So RTX 3080 competitor first, RTX 3080TI competitor later....
Soldato
Yeah I would love the Ti or supers to be released but they normally follow quite a lot later and I wouldn't mind stepping off the vega to something fresh as I very rarely get GPUs on release year as they are overpriced.
Soldato
I fully expect Nvidia to release an all-7nm Super line this time next year once TSMC has free 7nm capacity.
The 2080Ti was released only a week later. It might be the new norm.
The 3090 might be held back to counter Biggest Navi. Who knows
- Status