fatslob-:O said: Not much is going to materially change from now to next year so at the end of the day the majority of households even in the western world aren't going to have 4K displays ... |
Chinese manufacturers are bringing the price down, market penetration of 4k is going to keep on growing, that's the reality, that's the facts.
fatslob-:O said: Seeing as how Turing struggles to just even do 2 RT lighting effects in modern games, it's a high probablity that Navi will almost certainly fall short of being able to do full scene RT if it even get's close to that ... In fact, AMD are counting on cloud gaming to be able to do full scene RT ... |
We have no idea what AMD's implementation of full-scene RT on console is going to entail, we don't have any low-level information other than a few tidbits.
One thing we do know is that the hardware in the consoles will be a deviation from what we have on the PC currently for AMD's parts even the Navi PC chips.
fatslob-:O said: Let's just get this out of the way, compression algorithms are NOT the panacea that you're looking for ... Yes, I acknowledge that there is a market for 4K content but there's a distinction between demand for downlinking 4K content and demand for uplinking 4K content. For the former, there might be some value in downloading 4K but that doesn't mean we should conflate it with uploading 4K content ... In reality, upload speeds are going to be far worse for anyone trying to do real-time 4K video uploads. Pursuing interactive 4K content isn't compelling when bottlenecks are going to occur elsewhere such as the display resolution or the uplink speeds ... The western world should pursue having 4K displays and high internet speeds first before console manufacturers commit to making interactive 4K content ... |
I am talking about demand for 4k in general, it doesn't matter if you are uploading or downloading, if you are uploading 4k content, then hopefully someone on the other end is downloading 4k content. - That is generally how these things work you know?
Playing around with semantics by pointing out differences between uploading/downloading 4k is thus ultimately irrelevant.
Not everyone is going to be uploading in 4k, that's fine, these consoles will have some relevancy even 10 years from now, Internet speeds will only continue to increase during that interim.
fatslob-:O said: It matters not if Nvidia does or doesn't target the PC market with Tegra. It still demonstrates my point that an integrated graphics solution can't coincide too early with a discrete graphics solution ... As for Intel, they intentionally decide to keep their GPU design team lockstep with their CPU design team since the former didn't have much of an opportunity to design discrete solutions so it's not as if they're a relevant example ... |
And Intel is the example that contradicts your demonstrated point.
AMD keeps it's APU's a step back from is desktop parts on the technology front... I mean who ultimately gives a shit if AMD's 3000 series is Vega or Navi? It could have been Polaris for all I care... As the GPU would still be more capable than the Intel Variant.
Bbut holding back the chips with terrible power management, 12nm and middling DDR4 2400mhz isn't pushing boundries... And AMD should be called out on it for what it is.
As for Tegra, there are reasons why nVidia takes it's approach that it does, it needs to be price sensitive due to the markets that product is perusing... The Ryzen 2700u/3700u is supposed to be AMD's high-end mobile effort.
fatslob-:O said: Would've it have made you happy if AMD decided to delay the launch of Navi with their APUs (bad move regardless) or delay Matisse (would've been a disaster) altogether to be more like Intel ? |
Bit of a stretch to assume Navi is even remotely required.
fatslob-:O said: Given that Ice Lake looks to have a moderately higher IPC with nearly as aggressive clocks as last generation, I think AMD made the right call to launch products whenever the design teams were ready instead of having to delay everything until there other products were ready ... When AMD or Nvidia have nearly independent CPU and GPU design teams, the launch of an integrated solution is ALWAYS going to be staggered so if you don't see Nvidia delaying the launch of their discrete graphics card with their Tegra products then the same should apply to AMD. It's not at all a rational decision to delay launches just to coincide with other products in the future ... Planning ahead in advance is not good enough when engineers end up having to wait for the finished designs regardless and then integrate them together ... |
Then use an older GPU design. The GPU architecture isn't the intrinsic issue, it's everything else that is holding it back... Vega with 704~ shader pipes and only 38GB/s of bandwidth whilst being thermal throttled isn't doing AMD any favors. That's the reality of the situation.
Things don't improve though if we don't give criticism where criticism is due... And AMD's mobile efforts do tend to leave allot to be desired... And that has held true for years.
Remember how big the mobile market is. There is coinage to be made.
--::{PC Gaming Master Race}::--