AMD: Speculation, Rumors, and Discussion (Archive)

Status
Not open for further replies.
Since you are so adamant about the Nvidia 10x perf BS. They made up their own performance metric and used that to get to their magical 10x number... which their "new" architecture doesn't even support across all products.

1.7x from FinFet means that they choose a balanced approach to their design, taking into account frequency, density, and power savings.

And no, it is no longer a 2x perf/w increase from 14FinFet. It is upto a 2.8x perf/w increase from 14FinFet and the Polaris architecture.


And you don't think AMD just made up their own metrics for their own AMD technologies software?

This is what marketing does and always will do.
 
And you don't think AMD just made up their own metrics for their won AMD technologies software?

This is what marketing does and always will do.
You are in way too much denial. We can simply verify this when the footnotes get release, AMD usually just use a game/benchmark for this kind of stuff. Nvidia's 10x number isn't even software you can run.
 
You are in way too much denial. We can simply verify this when the footnotes get release, AMD usually just use a game/benchmark for this kind of stuff. Nvidia's 10x number isn't even software you can run.


What game is part of AMDtechnolgies? I wouldn't be surprised if they are talking about VR.
 
Why are you insisting it's some software or some such?
The "AMD technologies" in that is referring to AMD's technologies in the GPU, like architectural improvements and other new features (it even states that outright in the pillar, "arch+features"


I think its specifically VR. The way he was talking and what he stated at the time of showing that slide and right after as the slides where switching, he stated he is only talking about VR up till that point.

edit:

right around 30 minutes.
 
yeah but that x2.8 is only with AMD technologies software.
We don't know when it reaches 2.8x (maybe it's similar than the VR use case that Nvidia uses to make certain performance claims for the 1080), but it's not unrealistic at all and I don't see why you're bringing software into the mix.

A 2.5x improvement was needed just to match Pascal for some configurations (eg when using Tonga as a starting point), so 2.8x is not a big outlier for some particularly new power friendly workloads.
 
1. Statement of “future-proof” refers to support of current and upcoming technology standards including 14nm FinFET process technology, DirectX®12 and Vulkan™ API support, new display technology, and experiences such as VR. “Future-proof” statement is not meant to serve as a warranty or indicate that users will never have to upgrade their graphics technology again. Support of current and upcoming technology standards described above has the potential to reduce frequency of graphics upgrades for some users.

Its VR.

If this is the same one as the link to the slide.
 
Pretty underwhelming performance, wonder what it'd take AMD to take off with their clockspeeds. If they're still mucking around the 1Ghz mark, it's hard to see them competing with 2Ghz and beyond nvidia chips.
 
Pretty underwhelming performance, wonder what it'd take AMD to take off with their clockspeeds. If they're still mucking around the 1Ghz mark, it's hard to see them competing with 2Ghz and beyond nvidia chips.
We need to see how things pan in future dx 12 titles. The fury x is within 10% of 1080 gtx in some dx12 benchmarks and that's a last gen 28nm card.

The 1080 also does not sustain anywhere near 2Ghz in the founder's edition without tweaking hardware settings. Perhaps in other vendors versions.
 
So for now, the only thing that's really impressive about the RX 480 is the price?

Great for the consumer, of course, but it's only something to brag about as a company if you can achieve that because of some amazing engineering that brought cost down as well: anyone can achieve a low price by cutting into margins.

And right now, I don't see any engineering exploits. Performance between a 970 and a 980 is not exactly overwhelming.
If we assume that a GP106 will be exactly 1/2 of a GP104, with 10 SMs and a 128-bit 8Gbps GDDR5, the 480 should outperform a 1060, but only barely, with very similar power.

So the 480 will have a quite formidable competitor once the 1060 gets to market.

Still, it's way better than having to use the abomination that was Tonga against gm206.
 
That is one tiny card. Granted the cooler shroud makes it look larger. I wonder what the AIBs will end up doing with custom designs.

Hmmm, I guess Samsung is out of the picture? These are all made at Global Foundries. I wonder if that might be why they are having problems getting hardware out despite showing working silicon quite a while back.

It'll be interesting to see how these test out.

Regards,
SB
 
Status
Not open for further replies.
Back
Top