Digital Foundry Article Technical Discussion Archive [2016 - 2017]

Status
Not open for further replies.

The worst framerate performance on a Digital Foundry tests since they began to do it in 2008: average 1080p 15 fps with Vsynch on PS4 and 1080p and average 15 fps on Xbox One with Vsynch off..

And 2 to 3 minutes by loading screen... Problem of luminosity on Xbox One...

http://www.eurogamer.net/articles/d...mage-is-the-worst-performing-game-weve-tested

ULTRA BUMP 2000

Does anyone know if this shitshow ever got anywhere near fixed? I assume what happened here is similar to Sword of the Stars 2: Release or die(SOTS2 later got patched to actually work and is apparently a pretty decent game. But, the initial release was a shitshow).
 
ULTRA BUMP 2000

Does anyone know if this shitshow ever got anywhere near fixed? I assume what happened here is similar to Sword of the Stars 2: Release or die(SOTS2 later got patched to actually work and is apparently a pretty decent game. But, the initial release was a shitshow).

It actually got fixed, almost

 
I also wonder how DF staff are going to find out the resolution of any given game when they are over 1080p. I don't think there is going to be steps to count..

There will still be steps to count. You can always magnify a captured image. And even if you didn't it's easy enough to get a large 4k display with large pixel pitch. My 4k monitor has larger pixels than 2.5k monitor it replaced due to it being 49" display versus a 30" display.

Regards,
SB
 
I also wonder how DF staff are going to find out the resolution of any given game when they are over 1080p. I don't think there is going to be steps to count..
Sure there will be. Higher res makes jaggies and other aliases smaller and softer and thus harder to pick out, but there's not a hard cutoff that occurs at native res.

qYYytKu.png
 
I also wonder how DF staff are going to find out the resolution of any given game when they are over 1080p. I don't think there is going to be steps to count..

They have a 4k capture card, so it will probably be business as usual.
 
That's not console discussion but...I already know the answer...yes. There are a few games that already are constrained at certain settings by 4GB, even at 1080P IIRC.
 
Someone needs to tell Digital Foundry to drop RX480 memory clockspeeds (default 224GB/s) 20-25% to achieve 170-180GB/s to simulated memory contention and cpu load on NEO if they want a closer and more likely simulation of NEO gpu's real world capabilities.
If they can't work this out themselves, they don't deserve the clicks. :nope:
 
Someone needs to tell Digital Foundry to drop RX480 memory clockspeeds (default 224GB/s) 20-25% to achieve 170-180GB/s to simulated memory contention and cpu load on NEO if they want a closer and more likely simulation of NEO gpu's real world capabilities.
I disagree. The contention would be roughly the same on the PS4 target machine if they only change the res. Their R7 265 with 1.84tflops has also 178GB/s of bandwidth, and performs very similarly like a real PS4 on some games, like TW3 or RSS. I think the 'coding to the metal' on PS4 is largely enough to overcome the contention parameter.

Where the test is not relevant is when they keep the RX480 at 1080p to test if game can run at 60fps on Neo instead of 30fps.
 
Not a surprise. Nvidia has spent a huge amount of money in mobile GPU and CPU design but hasn't got any big design wins since the Microsoft Surface RT (which flopped). This is great for Nvidia. They get some returns for years of development effort. I am sure Nintendo got a good deal, since Nvidia's mobile side must be desperate. Hard to compete against Qualcomm, Samsung, Apple and others. Intel couldn't compete as the mobile chip profit margins are so thin. Like Intel, Nvidia is also used to high profit margins. They make most of their money with high end GPUs, professional workstations and HPC. All really high profit margin segments.
 
Not a surprise. Nvidia has spent a huge amount of money in mobile GPU and CPU design but hasn't got any big design wins since the Microsoft Surface RT (which flopped). This is great for Nvidia. They get some returns for years of development effort. I am sure Nintendo got a good deal, since Nvidia's mobile side must be desperate. Hard to compete against Qualcomm, Samsung, Apple and others. Intel couldn't compete as the mobile chip profit margins are so thin. Like Intel, Nvidia is also used to high profit margins. They make most of their money with high end GPUs, professional workstations and HPC. All really high profit margin segments.
I think Nvidia has used that experience with mobile GPU/Tegra/Jetson to push into other lucrative markets such as automobile (yeah still very early days for all but scale and margins are much bigger than what they could do with gaming HW).

You could also look at it that Nintendo is desperate as well with no way to differentiate themselves from Microsoft/Sony without going after mobile gaming and the challenges that has, and their previous gen HW failing sales expectations so what hardware/platform does that leave for them.
Could be said they both need each other for mobile console side as it is pretty clear Nvidia backed out from that directly by not updating to X1 from previous Kepler Tegra for the Shield Tablet, nor do I think any mobile device was updated to X1; only Shield TV was updated to the X1 Maxwell SOC.
This also raises what effect Nintendo deal has on Nvidia's pet child project in Shield TV (that is designed to be a complete ecosystem and still being refined - longer term this 'entertainment' technology could be used in other markets), so I could see margins still being a consideration albeit lower as it seems Nvidia felt mobile console/tablet was no longer a priority to them and resources better spent elsewhere such as automobile/Shield TV/etc, especially as it is very hard to compete as you say against Apple/Qualcomm/Samsung in mobile gaming.
It is not a viable profit margin area when those companies mobile demograph is so massive, and one I am still not sure how well Nintendo will do in the long term outside of Japan even with their brand name and IP, analysts see this as an unknown as well when considering both the mobile and home console competition.
Hence why I could see Nvidia wanting some kind of margin as they had moved their development to other potential mobile or associated segments that is more of an equal playing field or in their favour, especially as Nintendo would not be influencing cross platform AAA gaming development for Nvidia.

Cheers
 
Last edited:
Status
Not open for further replies.
Back
Top