Oef, this must be difficult to hear for all those "experts" here on NeoGAF
I'm really baffled by this:
"But for the Series X, because the rest of the elements are slower, it will not probably reach its 12 Teraflops most of the time, and only reach 12 Teraflops in highly ideal conditions. "
Holy crap!
So in practise the PS5 is faster than the XBox Series X!
. Jesus wept, this is so embarrassing
I think we got a ban-list full of users proving you wrong and then some more in this thread alone.
I am not saying we know how RDNA2 scales. I am saying that it is at least possible to scale performance by increasing compute units, and that historically this is what GPU manufacturers have done.
Yeah, less work for devs. In no way should be interpreted as is more powerful.
This is blatant click bait.
That must make him immediately an expertHe actually work with CryEngine.
Last year the CryTek junior team which he is part made a game in a week for the Global Game Jam.
As he removed the title Junior he probably received a promotion from last year to now days.
Well the CryEngine game and last year interview:
CRYENGINE | Haven, one of Crytek’s Global Game Jam 2019 entries
Crytek entered four teams into this year’s Global Game Jam. Check out game four, Haven, and find out about the game jam experience in our interview with the team who made it.www.cryengine.com
Wow, you're so personally offended that you're referring to him as a "dimwit." His words must have really struck a nerve.
I disagree with the title, however what I understood from the interview is that gpu power does not equal system performance, other variables are valuable too.
Both 2080 and 2080ti have the same turing architecture. All I wanted to say is larger chip with slower clocks is still faster than smaller with much higher clocks. I want both consoles to be close, but this argument (slower clocks can outweight bigger GPU on the same architecture) cant be proven with PC GPUs comparisons, therefore I cant believe it.Also why Nvidia give better performance for same Tittyflops as GCN ?
You cannot compare different arcitectures, we are on RDNA2 now, so nobody knows how performant RDNA2 is on latest node at TSMC. All we know is below, and NVIDIA 2080 is using older tech than RDNA2..
Lets see what 3080 is doing first...I am sure 3080 will have more shaders, and will be much faster and higer clocked as well.
They really are a special bunch.Xbox discord is definitely trying to downplay this. C c0de is already in the thread at desetera having meltdowns. Doing the promised work to stop any favorable positive PS5 news
Discord
Surprised this person is even brave enough to show up in threads after all of them have been outed as troll fanboys.
My mind is made up because things are crystal clear when using common sense, just like like my mind was made up months before the ps5 unveiling about which console would have the bigger numbers (unlike what insiders were preaching) because things were crystal clear then too, when using common sense. The things this guy says, if he says them because the translation is all over the place, make literally zero sense.
Nope, the special bunch is the Sony mob that rules both of these forums and are trying to either expel everyone, like dictator, that doesn’t agree with their bullshit, out of these forums, or drown everyone else under their bullshit rhetoric. They were humiliated once with their worshiping of people like that kleegamefan guy and osirisblack and they never learn instead doubling down on the efforts to do sony's marketing for free.
Actually there are many «simple» numbers that dictate computing power and then there’s bullshit and outlandish theories trying to conflict with these numbers just because. Sorry I don’t care what some no name engineer,who is foaming at the mouth over everything Sony-related in his social media and whose words we get through a translation from another guy who starts of with «be afraid xbots», thinks.
I want to understand this also. I want to know how cu’s can be detached from clock speeds when it comes to RT. On the surface it sounds like wishful thinking, but I don’t proclaim to understand if there is more to it than meets the eye.
You don’t understand, PS5, always runs max, while Xbox can hardly reach 8 TF.
He never ever says he doesn't have experience with the hardware. He mentions that he doesn't want to talk about the project he's working on, so he only echoes other developers.
Also why Nvidia give better performance for same Tittyflops as GCN ?
You cannot compare different arcitectures, we are on RDNA2 now, so nobody knows how performant RDNA2 is on latest node at TSMC. All we know is below, and NVIDIA 2080 is using older tech than RDNA2..
Lets see what 3080 is doing first...I am sure 3080 will have more shaders, and will be much faster and higer clocked as well.
That makes literally zero sense. Whatever, after the insiders phase we have entered crazy land again. This is my last post about any technical stuff, I will just wait for the DF comparisons.
A lot of it doesn't make much sense. Either the translation is shit or the guy is just bullshiting.
Some things that make absolute nonsense here's his logic.
- U can't look at numbers and see if that results in bigger outputs because there is more limitations going on. True. But those boxes are practical identical so yea u can actaully.
- PS3 had the bigger number but it didn't result in better performance in games like gta. So numbers don't mean much as PS3 had the bigger number and it didn't help them.
2 seconds later, PS3 had the best looking games after all and it did help them???
- PS5 has its own operating system, but microsoft uses its own CUSTOMIZED VERSION of windows. The two are very different
- What Sony has done is much more logical because it decides whether the GPU frequency is higher or the CPU's frequency at certain times, depending on the processing load. For example, on a loading page, only the CPU is needed and the GPU is not used. Or in a close-up scene of the character's face, GPU gets involved and CPU plays a very small role. On the other hand, it's good that the Series X has good cooling and guarantees to keep the frequency constant and it doesn't have throttling, but the practical freedom that Sony has given is really a big deal.
- But if all the parts in the Xbox Series X can work optimally and the GPU works in its own peak, which is not possible in practice, we can achieve 12 TFlops. ( 2 seconds later ) 12 Tflops can be achieved.
This gif the interview.
What I mean is I see repeated throughout the forum that CU count is the sole indicator of RT performance, aka 52 Cu vs 36 being 44% faster in RT operations. Why is this being brought up consistently. This would mean 0 clock speed in effect for RT, I can’t see sense in that, I’m not crazy right?
I'm not an expert, but I think CUs' frequency also play a role.
That makes sense right? So RT should scale with everything else, not related to CU count. E.g half the Cu’s at twice the speed should equal double cu’s at half speed when it comes to RT. Efficiency not withstanding.
This sounds like the bullshit Corrinne Yu would say.
There goes whatever credibility this guy had.
