Just curious as I have not watched the video as I am at work, but why compare with a thread ripper?
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
Digital Foundry || Minecraft DXR on Xbox Series X: Next-Gen Ray Tracing Analysis!
- Thread starter ILikeFeet
- Start date
- Status
Just curious as I have not watched the video as I am at work, but why compare with a thread ripper?
What on earth are you going on about?
Because it is running higher settings it would invariably drop res more often, that is common sense. Why do I need to count things?
Weird Set of demands you have.
Coalition wanted to Show John and Rich these Benchmarks just to Show how powerful the Maschine was - and they thought the 2080 was an impressive gpu to be similar to. That is the point. That is an awesome gpu.
no, the devs said it runs at 4k 100+ fps. DF SAYS the devs showed something but provided no proof of it, and then didnt even do their jobs on the footage we do have.
so whats your final verdict? is the XSX a capable RTX machine? what can we expect from future XSX titles? can it handle Control with basic RTX features?
what i'm on about is, do an RTX 2080 test and count the number of sub native frames for it to hold the same framerate as the public XSX footage. then do the same counts to see if the xsx version is even sub native at all, as you claimed it was, and if so, if its either lower res than or drops frames more than the 2080 in practice rather than in the canned benchmark you claim they ran. aka do the same pixel counting you do for everything else instead of making a blanket statement and not providing any evidence of it.
which wasn't the argument they made in the original article, they claimed the 1060 was faster, and that it was cheaper, when neither was true even in their own benchmarks. aka they straight up lied, and i poked john and got them to correct it. don't believe me i'd be happy to prove it. To be more precise it was october 26th 2018 when this incident of DF blatantly lying to make nvidia look better had occured. It took a couple days of back and forth to correct part of that lie, but the whole lie was never fixed. So yeah, there's direct history that i was personally involved in correcting that proves that DF have slanted things or lied in favor of nvidia at least once in the past, and thus there is precedent to say they might do so again until proven otherwise.
Um, you can extrapolate that logically by utilising the benchmark on PC. If the bench Drops below 60 with drs off, that same bench with drs on would instead scale a certain percentage of those frames. An Rtx 2080 in the bench is not close to 60 fps average at a locked 4K with no DRS on. The games cutscenes are even heavier than the benchmark at times. I spent a lot of time in my PC Review explainung all this. You can run that test yourself or just read about it online.
im confused now - didnt your team just release a video stating the opposite you are saying here? =/
John in the Video says something like, 'not everything ran to par, the CPU...'
OP
OP
the Gears 5 demo separates GPU and CPU performance, so you can see each individually. the XSX cpu performing worse doesn't mean less overall performance as it means the game itself isn't cpu bound
Right, but my point is you made the claim its sub native on XSX and didn't provide frame by frame analysis to show exactly how often, or how far sub native it is, or do a comparison to a 2080 on PC ultra directly, which favors the 2080 mind you, to see if performance is better, worse, or the same. Furthermore, you make claims of a benchmark done behind closed doors and provide no proof, when the technical director is on record saying over 100FPS currently. AND you fail to take into account that they've been at it 2 weeks, not a full development cycle. So, how would the RTX 2080 run the game 2 weeks into the development of the pc version? better or worse than this? How will the XSX run it after a full development cycle? better or worse than it does now? Or do you assume that it runs as good as it ever possibly can already?
okay! thanks for the quick replies and clarification. :) appreciate it
They said in the video that benchmark run was NOT using dynamic resolution. So...
Static 4K resolution + Ultra PC settings on both:
AMD Ryzen Threadripper 2950X + RTX2080 + 64GB RAM is slightly faster than Xbox Series X.
They also do not claim that this is as good as it can be and reiterate trillion times that this is 2 weeks of work port. Also it seems that CPU is a slight bottleneck in comparison to threadripper.
but they provided no video proof that this benchmark even exists, nor showing the framerate or results of said benchmark. And they have a proven history of lying to make nvidia look better, as mentioned earlier. They claimed the 1060 was faster and cheaper than the RX580 at the end of 2018, both of which were outright lies even using their own benchmark charts. And it took me days to get them to fix it, they didn't indicate on the article they edited it to reflect changes, nor did they publicly admit to having made a mistake, they quietly swept it under the rug and didn't even correct every lie posted that day. So yeah, there's a history of lies slanting things against amd and towards nvidia there.
Let me just say that the devs are reposting our coverage on Gears 5 and they think it aligns with reality.
CPU in both consoles are great. Highish mid tier PC great which is enough when the consoles are more likely to be GPU bound anyways.
A port is a port, the original game was not designed for the new hardware so its not gonna show everything.
Aren't they even saying they are targeting 120fps@4K for the multiplayer?
If this is what we can expect from ENHANCED BC Titles than this is awesome!
way better than I expected. I thought it's only FPS and RES Boost, but having PC Ultra settings and even more is very Good to hear.
If this is what we can expect from ENHANCED BC Titles than this is awesome!
way better than I expected. I thought it's only FPS and RES Boost, but having PC Ultra settings and even more is very Good to hear.
Fast forward 1 year, and you are proven wrong on so many fronts.
2080Ti runs Minecraft DXR 90fps @1440p NATIVE, and Turing is so vastly superior to RDNA2 and consoles in RT.
The Xbox Series X is barely 2060 Super performance in RT just like many here told you so.
Oh and your formula was proven so hopelessly wrong!
Sorry for the old thread bump.
This thread and many others are pretty fun to go back and read. People sure did have some unrealistic expectations.
It just goes to show that next gen systems are always slower than we expect
I still can't believe Microsoft has shown a damn thing on the series x. I mean shit, they did nothing for the One X.
Not gonna say lazy. Just wonder if they even care.
Not gonna say lazy. Just wonder if they even care.
It didn't, but it's hilarious to think someone had a year old grudge about pixels.
I did not initially realise this thread was old and got very confused when "Lockhart" was mentioned.
Lazy platform holders. I though I'd seen everything. Man, what a wild take and what a wild bump.
As for the bump, anyone who was paying attention, i.e. anyone who read about the RT setup of these machines and/or who listened to DF, already knew this. It was only the fanboys who, as usual, overstated the RT capabilities of these machines. Not worth bumping a year-old hardware thread for a "told ya" take, since you're always going to get quite a few hyperbolic takes when it comes to next-gen hardware. While we're at it, let's bump other threads to point out how stupid it was that people were freaking out about the variable clock rate of the PS5's SoC or about the Series consoles' memory setup.
- Status