ethomaz
Banned
The video actually shows the lack of understanding he has with his job.
Or maybe he close his eyes due the bias.
Last edited:
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
- Status
The video actually shows the lack of understanding he has with his job.
Or maybe he close his eyes due the bias.
IntentionalPun
Ask me about my wife's perfect butthole
Sorry misread your post for sure; but it's a common thing on this forum to claim MS are paying DF.
FranXico
Member
Unfortunate that the way DF presents their results consistently fuels that conspiracy theory.
IntentionalPun
Ask me about my wife's perfect butthole
So what did the thing you know about even say then roughly?
Because flying someone out to do a video would likely come with terms.. as would any scenario where a company provides early access to something, including game reviews. (what can and can't be said) Streamers have to sign shit like that a lot too.
Last edited:
I think that's true for any media, whoever gives them more access and material for content is going to naturally get more coverage and somewhat more positive slant. There was a few moments last year when DF did questionable things, like when Richard gave credence into the github stuff and even after the specs were announced he casted doubt on Cerny's words. Or many instances of Alex going out of his ways to downplay PS5, and contradicting Cerny or any positive thing that developers were saying about it.
Overall though, I do think DF are impartial in doing their tests and showing the data as it is. They do get things wrong and they don't always issue corrections as they should. But I do trust them above most other sources for this kind of analysis.
SlimySnake
Flashless at the Golden Globes
its a valid comparison of the two GPUs. If anything this helps us identify where the bottleneck might be on the Xbox Series X since it seems to be doing fine with nothing happening on screen.
Maybe the CPU is the bottleneck here or the I/O?
Rudius
Member
This test shows that 1080p60 with ray-tracing would be possible, even with some drops here and there. I'm playing it in the 60fps mode, but would prefer to sacrifice resolution rather than ray-tracing. Perhaps during true gameplay the CPU, encumbered by ray-tracing, would bottleneck some action scenes, but in exploration it should run smoothly.
SlimySnake
Flashless at the Golden Globes
MS, Sony, Activision, EA, and even Apple and other consumer product manufacturers fly out and provide hotel accomodations for tech journalists who cannot afford to expense cross country or even international trips for every game. This is normal industry practice.
Sony gave Richard an inside look on the Pro and he saw demos that even we did not see. They also called him up a few nights before the Road to PS5 reveal and gave him an early look at the Road to PS5 presentation. Same goes for the PS5 UI video.
Again, normal industry practice.
LucidFlux
Member
Looks to be a solid stress test of the GPUs and specifically RT as the results align with the theoretical performance difference between the 2 GPUs.
Obviously it's not able to measure overall system performance with the game essentially paused. Things like the I/O, caches, memory, CPU are likely not being taxed much at all. So it's not a full system benchmark between the consoles, but it's still pretty cool that this bug allowed us to finally see those extra XbsX CU's put to work lol.
Anywho, I'm looking forward to finally diving in to this game over the weekend but in the 60fps mode.
Does it though? We've known there is a bottleneck since the very first set of benchmarks. But this doesn't clear anything up definitively. The GPU comparison was never in doubt, it was more about how all the other pieces fit together to produce the final result. And my theory has always been that PS5 was designed to get everything it can out of a smaller chip efficiently, to more or less match the performance of a bigger chip in real productive terms, but not singular benchmarks like this. And we have a lot of evidence which points to this being true.
SlimySnake
Flashless at the Golden Globes
The only problem i can foresee is that the framedrops are way too frequent and unpredictable even on PC. My 2080 was constantly flactuating between 43-54 fps during any kind of physics activity. I didnt even have to be in combat for it to dip to 43 using PS5 settings. Just lift up something and throw it at something else and the framerate goes all over the place. It was a locked 60 fps when standing still which is very close to what we are seeing here on the xbox version.
In the corridor of doom, it goes all the way down to 33 fps. We have seen it dip below 30 fps in DF's gameplay recordings. But who knows maybe a 1080p mode might not have those drops.
Lysandros
Member
Well they are pretty much in the obligation to show data as it is (generally) if not someone else will (as it happens sometimes) and the whole situation will not be very sustainable.
HeisenbergFX4
Gold Member
What you said is correct but what you failed to read is Xbox intentionally set out to make DF "pro xbox"
Its not a point I am going to keep going back and forth on as again not only was I told that directly by some high up Xbox officials but still have the screen caps of some of them bragging about it in game chats
I still watch DF content but with like every "review" site out there they all have their personal favorites so people need to keep that in mind.
IntentionalPun
Ask me about my wife's perfect butthole
And when companies contact you to do a video they promote "why" you should do it quite often. Sometimes it happens because YOU contact the company, but it goes both ways. I have an Instagram w/ enough followers that I get nagged all the time for stuff. It's just how marketing works.
"We'll fly you out and this video will drive clicks to your YouTube."
Do you have actual evidence of some on-going "deal" between them
HeisenbergFX4
? Because DF gets special access from Sony too.
Maybe it's more than that; but your last statement of not knowing "deals" beyond how it "started" is a bit suss TBH.
edit: I mean.. all companies that give exclusive access to someone are hoping it leads to them being pro-that-product... lol.. that's at the core of that type of marketing; giving special access.. with the "nudge nudge wink wink we hope you continue to like our brand"
"We'll fly you out and this video will drive clicks to your YouTube."
Do you have actual evidence of some on-going "deal" between them
HeisenbergFX4
? Because DF gets special access from Sony too.Maybe it's more than that; but your last statement of not knowing "deals" beyond how it "started" is a bit suss TBH.
edit: I mean.. all companies that give exclusive access to someone are hoping it leads to them being pro-that-product... lol.. that's at the core of that type of marketing; giving special access.. with the "nudge nudge wink wink we hope you continue to like our brand"
Last edited:
HeisenbergFX4
Gold Member
And when companies contact you to do a video they promote "why" you should do it quite often. Sometimes it happens because YOU contact the company, but it goes both ways. I have an Instagram w/ enough followers that I get nagged all the time for stuff. It's just how marketing works.
"We'll fly you out and this video will drive clicks to your YouTube."
Do you have actual evidence of some on-going "deal" between them
Maybe it's more than that; but your last statement of not knowing "deals" beyond how it "started" is a bit suss TBH.
The only evidence I have is like I said screen shots of Xbox officials talking about going out of their way to recruit DF and Richard in particular
rickybooby87
Member
Here's my response to that video:
I think this purely comes down to more compute units to do RT calculations across, as in scenes without large reflective surfaces, performance is fairly equal. Not to mention, photo shoot mode of static/limited frames isn't going to be a good measurement of memory bandwidth, heck, I get great FPS when I look down at the floor in CyberPunk2077 on my RTX 2060. There is no denying that the XSX is the more powerful box, but they are going to have a hard time showing it, especially if everything is VYSNC'd at 30/60FPS or if using VRR. There simply isn't enough of a hardware gap to make a vast visual difference without eating into those 30/60 target frame times.
IntentionalPun
Ask me about my wife's perfect butthole
See my edit..
I mean.. dude.. that's literally the main reason companies do these promotions. Find a big "social media" person or streamer/youtuber.. give them special access.. hope that it influences them in the future.
You think Sony gave special access to those Japanese YouTubers w/o the hope that it would make them more pro-Sony in the future?
It's a carrot on a stick.. it's exactly what I was suggesting in my post you quoted. That DF's bias could just be the result of that special access, and hoping that if they appear pro-Xbox they'll get more special access in the future. Sony started giving them access too.. and they are likely hoping to get more of that.
Vague quid-pro-quo is the name of the game for "influencer" marketing. Because if you spell it out as a deal, they have to say it's a paid advertisement.
It's why people should just judge products how they want to judge them, and try to find data-driven results, not "opinion pieces" in hobbiest industries. There is no ethics in hobby journalism and there never has been, it's all marketing.
Last edited:
kyliethicc
Member
But saying "others are biased too" doesn't make DF neutral.
IntentionalPun
Ask me about my wife's perfect butthole
I never said they are neutral. In fact I suggested they aren't.
But OF COURSE MS reached out to a YouTube channel with a massive amount of subs hoping to influence them... that's, just how that works. It's a psychology driven industry (marketing, advertising) through and through. Nobody is "neutral" either way.. everyone is influenced by various biases for various reasons.
Last edited:
They would have to do more optimisation than just dropping the res down to 1080p to make it work, and I think the team doing this port is small so it was probably never an option for them to go down that road. I personaly think playing games at 1080p on these console in this day and age is a non-starter. At worst I want a dynamic res.
BGs
Industry Professional
SlimySnake
Flashless at the Golden Globes
Right. Insomniac had to reduce the crowd, the reflection resolution, the traffic density and even some lighting effects to hit 1440p 60 fps. And even then they had to do dynamic res in case that wasnt enough to hold a steady 60 fps with the lowest count being 1080p.
So yes i suspect, Remedy would likely also have to make more changes than just a resolution drop. the problem is that they are already on low settings on pretty much all the other graphical settings lol
kyliethicc
Member
Coincidentally, DF views were on a several year decline. Then after their 2017 reveal of the Xbox One X specs, their channel blew up and grew nonstop.
I'm sure Richard is very appreciative of that coincidental help Philsoft gave him.
Lysandros
Member
The problem that i have with the 'GPU comparison' is that XSX never had a 18% 'whole GPU' advantage over PS5. XSX has a max theoretical vector ALU advantage over PS5 while being in deficit in others like rasterization, fill rate, cache speed/architecture etc. Thus i don't/cannot consider PS5's GPU 'weaker' as an objective truth.
Last edited:
HeisenbergFX4
Gold Member
I 100% agree and even said in an earlier post every single site out there has to have their own personal favorite of a system and maybe somewhat biased towards said system.
My response was mainly to people saying DF is 100% neutral
I kind of feel you are making my point as well saying this is how it works.
Thirty7ven
Banned
It’s quite amazing really. Listening to them talk and theorize over the results really goes to show that their knowledge is surface level. Richard coming across as a XBOX excited about the results is hmm *chefs kiss*
If anything this “benchmark” just goes to show that there’s more to real time performance than paper specs.
And anyone who still thinks DF doesn’t have a vetted interest in showing Xbox on a more positive light than PlayStation is disconnected with reality. The work speaks for itself, the way data is presented speaks for itself. The titles of their videos and how they play the YouTube algorithm tell the story.
Last edited:
IntentionalPun
Ask me about my wife's perfect butthole
Well you responded to my post suggesting they aren't neutral, and as I said: "could boil down to MS inviting them to their offices." (which I've been saying for a year)
I just think the vague mentions of "knowing something" aren't super helpful, particularly when as you said you only know of MS wanting to curry favor and not an actual deal, when it's clear many on this forum think there's some paid deal behind the scenes... and it's kinda pointless because giving exclusive access is done always by a company with the hopes of currying some favor.
DF whined about Sony not giving them access to things.. if Sony gave more access, they'd probably start being more positive about them. As it stands, they still praise Sony all the time, because they have no real "deal" with MIcrosoft (and if they did, it would be illegal.)
Last edited:
funking giblet
Member
Can we say "synthetic benchmarks"? Of course we can
FrankWza
Member
I’m going with I/O. Not because I know what I’m talking about. It’s more because there’s no precedent for this and nobody really knows. Seems right because nobody has an answer yet.
Dodkrake
Banned
A broken shitbox car must be performing perfect when not being driven, then. You must apply to become a mechanic.
SlimySnake
Flashless at the Golden Globes
your analogy doesnt work for games. You cant run control at 60 fps on a laptop for example even if you just stand there doing nothing. The fact that the xbox can hit 58 fps at 1440p with RT means thats its peak. I can promise you the x1 will not even hit 10 fps when you are standing still with RT on at 1440p.
A better analogy involving cars would be a racing car that runs just fine when being driven in a straight line, and then completely shits the bed as soon as it needs to turn or brake. Thats whats likely happening with the xsx version here. Runs just fine with not much going on but as soon as the physics blow up and the combat begins it starts to falter.
That stood out a mile, I noticed too
Interesting..
Edit: Taken form beyond3d, don't know the main source.
SlimySnake
Flashless at the Golden Globes
can someone post this on era and tag Dictator? lol
Interesting..
Edit: Taken form beyond3d, don't know the main source.
assurdum
Banned
I already noticed in el analist del bit some details is missed on series X raytracing.
Interesting..
Edit: Taken form beyond3d, don't know the main source.
assurdum
Banned
Ironically the same DF.
Interesting..
Edit: Taken form beyond3d, don't know the main source.
rickybooby87
Member
Actually, now that I am tracking back through the video, I am noticing similar things in the side by side comparisons, it's almost like everything is being rendered at a lower resolution for the XSX, or lower res textures?
Interesting..
Edit: Taken form beyond3d, don't know the main source.
SlimySnake
Flashless at the Golden Globes
must be a bug. we must contact Remedy.
kyliethicc
Member
1. xbox one can't do RT
2. the fps would drop if the game was being played. a static image is not gameplay
3. we have no idea how the games would compare when being played with uncapped fps
Rudius
Member
That level of variation only happens with ray-tracing enabled?
Last edited:
Alex is daft, honestly a game that is having more issues running on the XSX than the PS5 but claim XSX victory cos photo mode when it's doing nothing is showing higher FPS! He picks his cherry's alright
This sort of comparisons does not exactly work with PS5, smartshift tries to mitigate race-to-idle when GPU is sitting around doing nothing much (was there much camera panning?). Stuff like this will skew results
Edit: actually don't know how much smartshift can be playing around when things are sitting idle
Last edited:
Dodkrake
Banned
Do you drive your cars by having them run on idle and looking at the RPM meter, or do you you actually drive them around town? Because car 1 is the Xbox, getting 5000rpm on idle, and then not going past 2500rpm when being driven, and car 2 is the PS5, that get's 2500 rpm on idle and 5000rpm while being driven.
Since cars are to be driven, congrats, Alex just proved that the shitbox runs better while not being driven around.
Lysandros
Member
Soon to be labeled as a peculiar 'error' or bug not hardware related at all by the Great Governor i presume. His judgement is final for me, i will wait for it.
SlimySnake
Flashless at the Golden Globes
Nope. When i ran my tests using ps5 settings, i was getting almost 100 fps in some scenes with nothing going on but as soon the combat started it would be in the 85-95 range with some rare drops into the high 70s.
That's without rt at native 1440p. With dlss and no rt, i was getting 150 fps with drops to 125-135 fps in combat scenarios.
There's a fundamental problem that's being missed in these discussions about why the XSX has been underperforming, electrical power and heat dissipation constraints. XSX's GPU and CPU clocks are always at a fixed frequency, under load it has to draw more power to drive more transistors being switched concurrently and therefore has to get rid of a lot more heat faster. Question is to fully push all RDNA2 52 CU's @ 1.8ghz and all 8 cores @ 3.6ghz is 300watts sufficient? and realistically the whole system should be sustained within 240watts (80% is the PSU's efficiency range). Compare it with an equivalent pc, the fans are spinning faster louder with them mounted directly up onto the heatsinks and much bigger PSU. Of course the same drawbacks can be said about the PS5, but the PS5 has been designed from the ground up to work efficiently with these constraints
The vapour chamber is great but is thermally coupled to the APU by paste, liquid metal is a much better thermal conductor or put it another way paste is way more heat resistive. Heat (just like electricity and everything else in nature) will flow through the path of least resistance, the advantage of vapour chamber isn't being efficiently utilised here because of the paste. Would like to see the temperature difference between APU and how much is being conducted onto its heatsink, I bet there's a much bigger difference than PS5's ie XSX has greater thermal drop (meaning that there's higher thermal resistance between APU and heatsink)
A good analogy is XSX is like a rigid body due to the fixed frequencies, there's no flex/elasticity to give when it gets strained under load. It's constraints to the very tight power budget, the system is unable to be fully pushed with the watts available and cooling solution it has
It's server config will afford it that higher power budget an better cooling, and bonus it with 20gb ram so no more gimped memory bus. It may finally perform as it's intended paper specs but only to run 4 x Xbox1 VM's
Hypothetically even if the XSX was built with a bigger PSU and more effective cooling so it can run at it's full potential, it's performance can edge above the PS5 on cross-platform games. But then again the PS5 does have a lot of architecture innovations that used correctly will still outperform the boosted XSX
The vapour chamber is great but is thermally coupled to the APU by paste, liquid metal is a much better thermal conductor or put it another way paste is way more heat resistive. Heat (just like electricity and everything else in nature) will flow through the path of least resistance, the advantage of vapour chamber isn't being efficiently utilised here because of the paste. Would like to see the temperature difference between APU and how much is being conducted onto its heatsink, I bet there's a much bigger difference than PS5's ie XSX has greater thermal drop (meaning that there's higher thermal resistance between APU and heatsink)
A good analogy is XSX is like a rigid body due to the fixed frequencies, there's no flex/elasticity to give when it gets strained under load. It's constraints to the very tight power budget, the system is unable to be fully pushed with the watts available and cooling solution it has
It's server config will afford it that higher power budget an better cooling, and bonus it with 20gb ram so no more gimped memory bus. It may finally perform as it's intended paper specs but only to run 4 x Xbox1 VM's
Hypothetically even if the XSX was built with a bigger PSU and more effective cooling so it can run at it's full potential, it's performance can edge above the PS5 on cross-platform games. But then again the PS5 does have a lot of architecture innovations that used correctly will still outperform the boosted XSX
Last edited:
assurdum
Banned
Now you said an interesting thing. Could even be smartshift trick some perfomance result? Theorically developers have full control to the frequency for their purpose but I don't know how the setting profile could works in the different game mode
Last edited:
Neo Blaster
Member
Shmunter
Member
Neither did he, someone told him on Twitter. But most knew it to be possible anyway as did I.
But solid result for Xbox, maybe the 52 cu’s are coming in handy for RT.
Yet I can’t help but feel photo shots without action seems to be insufficient. What’s the framerate when other parts of the render are also being stressed, alpha effects, partials, etc.
I like Richards idea of running faramerates below 60 divisible into 120hz. However the mentioned 40 is too low. I’d like consoles to support 50hz, all modern displays support it.
Neo Blaster
Member
Nah, Xbox is winning, they won't bother.
Lysandros
Member
I am sure Slimy is a very nice fellow but his posts can often result in such psychological alterations, i share similar sentiments.Oh, boy...
- Status