Ass of Can Whooping
Member
If you're not exclusively counting gaming PCs, then no, they won't. Most won't even have dedicated GPUs.
-
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
If you're not exclusively counting gaming PCs, then no, they won't. Most won't even have dedicated GPUs.
Metroiddarks
Member
Ok last try to be an adult, from 100% of pc only a fraction of them play, from that fraction most them play a very specific kind games (mobas,battle royale, MMOs).
Now from the rest of them which play are not user which bough many games in full price even most of the users doesn't have something more powerful than
gtx 1060 (even most of them use gpu integrated) so they are not market which speed actually much money as many pc gamers believe, the people who
bought a high end gpu are few compare to who just bought a OEM machine with integrated gpu.
Is not a coincident your requirement to play a game since 2013 barely increment, just looks the minimum requisites of CPU, GPU, Ram, olds HDD.
The main market is where the money for AAA and AA games is which is the consoles and for small or simpler games the base are smartphones even now the switch has
a very powerful position here also.
SonGoku
Member
Not trolling btwSosokrates he's been spreading the same FUD in another thread.
Sosokrates
still waiting for your explanation on 10.27TF, false advertising?Im saying AAA publishers prioritize their primary audience, games are designed around consoles spec.
Last edited:
AveragePhoenix
Member
Uh?It's called Master Race not Potato Race for a reason, either keep up, or keep out.You can always join us, your peasant brothers and sisters, we welcome you.
You act as if the target customers are the Master Race, a little hint, they arent, devs will optimize for the largest common denominator
Ascend
Member
The console aren't even out. I can't say how the 2.4GB/s is doing, any more than you can say how good the boost clocks of the PS5 are doing.
Based on what? Games on 4K do not use more than 6GB of VRAM today.
Where did you get that 4.6x from? What is this hard evidence you speak of? Share a link please.
I'll just share a basic analysis here... An HDD under optimal conditions can stream around 150MB/s. The XSX does 2.4GB/s, which is around 16x. Already, your 4.6x doesn't make any sense. In reality, games are programmed closer to 50MB/s or even lower for the HDD, because there are slower HDDs out there, and you have to take the OS and other stuff into account. It's quite a safe number. And the same will apply for the SSD. If you really think that XSX is going to work at a constant 2.4GB/s and the PS5 is going to work at a constant 5.5GB/s specifically for the games only, you're delusional. Most likely, both those numbers are sequential reads, which are at least 4x faster than random reads. Random reads are important for games, and sequential is meaningless.
This is true. But he said the SSD will give more headroom to the CPU/GPU/RAM where needed, and that is simply false. The CPU and GPU won't have to idle as much if at all, and the RAM use can go down since you have to pre-load for much less 'seconds' in the future.
What are you referring to?
Bo_Hazem
Banned
And that's sweet! You know what that means? If those games need SATA 3 to play on Ultra, PS5 can do that and then some. Now you get what I was saying previously?
Metroiddarks
Member
We are going to never user those 8 cores cpus, 16 GB of fast GDDR6, high end gpu, super fast SSD because those poor pc gamers will neverSo what you’re saying is, the PC will be holding these consoles back. Got it.
change their machines.... wait that not sound correct.
James Sawyer Ford
Member
the difference in SSD speeds
I think it will result in better looking games through superior texture streaming
AveragePhoenix
Member
This is a little tricky, last-gen consoles were pretty behind compared to most PCs at the time, but this time the consoles are actually competent compared to most PCs, especially their SSDs and GPUs.Ok last try to be an adult, from 100% of pc only a fraction of them play, from that fraction most them play a very specific kind games (mobas,battle royale, MMOs).
Now from the rest of them which play are not user which bough many games in full price even most of the users doesn't have something more powerful than
gtx 1060 (even most of them use gpu integrated) so they are not market which speed actually much money as many pc gamers believe, the people who
bought a high end gpu are few compare to who just bought a OEM machine with integrated gpu.
Is not a coincident your requirement to play a game since 2013 barely increment, just looks the minimum requisites of CPU, GPU, Ram, olds HDD.
The main market is where the money for AAA and AA games is which is the consoles and for small or simpler games the base are smartphones even now the switch has
a very powerful position here also.
The reason requirements haven't gone up is because console gaming is a big market devs don't want to miss on, so they optimize their games for that too.
What I expect to happen is, games will be optimized for HDDs until most PCs catch up, then it will be exactly as it is today.
BluRayHiDef
Banned
So, do you suppose that there are redundant Tempest Engine CUs for the sake of having high yield rates? Because if there's literally only one Tempest Engine CU (a CU that's completely unrelated to the 40 CUs out of which 36 are active), then the odds of producing an APU with no functional Tempest Engine CU would exceed the odds of producing an APU with one that is functional.
This is why I think that the Tempest Engine is one of the four left over CUs out of the 40 CUs; the odds of at least 37 CUs out of 40 being active (36 for graphics and one for audio) are far greater than the odds of one CU apart from the 40 designed specifically to be the Tempest Engine being active.
This is why I think that the Tempest Engine is one of the four left over CUs out of the 40 CUs; the odds of at least 37 CUs out of 40 being active (36 for graphics and one for audio) are far greater than the odds of one CU apart from the 40 designed specifically to be the Tempest Engine being active.
AveragePhoenix
Member
That's not what I said at all, I was pretty clear actually. Games are not going to be optimized for SSDs except for the PS5s exclusives, everything else will, as resolution and graphics do scale up and down
AveragePhoenix
Member
You said devs will optimize for PS5, which they won't, games will run just fine on SXS too, so, it all goes full circle, SSD won't be a game-changer, at least you agreed with me at the end
Bo_Hazem
Banned
You really aren't paying enough attention to the thread, but here:
Xbox Series X (11sec) vs Xbox One X (51sec): State of Decay 2. Difference is only 4.6x.
PS5 pre-devkit state (1 year ago, 0.8sec) vs PS4 Pro (8sec): Spider-man. Difference is 10x, with WIRED reporting 0.8 vs 15sec on another test (18x), questioning the 0.8sec being due to other stuff happening inside the system before loading. Plus it's reported to be a slower version. No need to take all that talk as anything, just pay attention to actual videos we can see:
Until we see further proof in the future, we better stick to actual, visible evidence.
+++
And for 6GB VRAM joke:
RE2 Remake, some exeeding 14GB VRAM:
Last edited:
DeepEnigma
Gold Member
SSD won’t be a game changer now. These new accounts made just to troll here, I swear.
Last edited:
AveragePhoenix
Member
This is actually my first time on Neo, I can hand you my twitter if it makes you feel better.SSD won’t be a game changer now. These new accounts made just to troll here, I swear.
And no, as I said, games will be optimized for HDD until SSD becomes the norm on pcs
Bo_Hazem
Banned
Ok, listen carefully:
It works automatically, you don't need to do anything (I'm 100% you didn't even open the video) it automatically optimizes the map, you just throw it there and the system is the easiest to develop for.
Developers optimizing for PS5 meaning other systems will struggle with "Usain Bolt". Make it slower, and PS5 uses that extra room for better graphics and draw distances as the view focus is way more narrower than other system and ultra fast to keep up with rapid turning.
AveragePhoenix
Member
In what part does Cerny say its automatic? He proposed hypothetical capabilities "What if you could just turn around and have everything load in the game as the player turns?" He says you would need 4gb/s to do that, which is actually a hypothetical answer which would heavily vary based on the game. I do believe that the PS5's SSD can do that, but he never says its something "automatic", Mesh shaders have to be integrated into games, it requires development, everything in a game does.
Last edited:
DeepEnigma
Gold Member
Ass of Can Whooping
Member
They're not going to wait for PCs to catch up. They won't mind abandoning anyone still using HDDs, just like they didn't mind abandoning the PS3 and X360 after the first year of this gen. and how they won't mind abandoning the current-gen after the 1st year of next-gen.
Technology doesn't wait for consumers to catch up. That's not how it's worked. Ever.
Ascend
Member
Is it going to be 120%? I doubt it. Considering that both have a separate hardware for decompression, no one will be using the decompressed data. It's all going to be compressed. So the difference will be between the 4.8GB/s and the 8-9GB/s, which translates to 67% - 88% difference. Still a big difference. But, we still don't know what they are actually depicting here. It could be a lot bigger or a lot smaller... There are 4 possibilities...
1) Both the XSX and PS5 are sequential read
2) The XSX is sequential and the PS5 is random read
3) The PS5 is sequential and the XSX is random read
4) Both the XSX and PS5 are random read
Considering that even in simple tests for current PCI-E 4.0 SSDs, only sequential reads pass the 4GB/s mark, it is highly likely that the given PS5 spec is sequential read, meaning it will not be representative of actual transfer speeds. The same applies for the XSX. And more importantly, value for sequential reads is not a guarantee for high values in random reads... Use this as a reference;
Sabrent Rocket Q 1 TB M.2 NVMe SSD Review
Priced at $130 for the 1 TB version, the Sabrent Rocket Q SSD is one of the most affordable SSDs on the market. It uses a new Phison E12S controller, which has better thermal performance and can operate with smaller DRAM cache. Write speeds are impressive due to 250 GB of pseudo-SLC cache.
And lastly, remember that we're talking about the I/O. Games will have to work with an additional SSD as well. So technically, the same I/O should be able to handle multiple storage devices, meaning, the games will never be programmed for the max I/O speed, simply because it could technically be cut in half if you have a second SSD that wants to load something at the same time.
Edit: I should add this, since most people don't know how to convert it... But, the fastest result in here, the 73788 IOPS from the ADATA SX8200 Pro, is actually just 288MB/s. Compared to the sequential read, it's literally 10 times slower. This is a worst case scenario, so we should expect at least 500 MB/s from these SSDs for games... But don't delude yourselves into thinking we'll get GB/s in real-time from the consoles.
Last edited:
AveragePhoenix
Member
So this gen didn't have thousands of cross-gen games. Got you.
Edit: Also, I love how all of the videogames nowadays are targeting 2080TIs and ultra-fast SSD; really smart moves.
Last edited:
Ass of Can Whooping
Member
I never said we still didn't get loads of cross gen games.
The big publishers, who spend the most money on development, didn't mind abandoning last-gen consoles after the first year even though it would have brought them a shit ton of money to still support them,.
We're going to see the exact same thing again. We're going to see alot of third party games requiring SSDs by next year.
AveragePhoenix
Member
You mean "Big publishers" like EA... Still laughing games for the PS3 and 360 5 years after it launched... Gotcha
Is it going to be 120%? I doubt it. Considering that both have a separate hardware for decompression, no one will be using the decompressed data. It's all going to be compressed. So the difference will be between the 4.8GB/s and the 8-9GB/s, which translates to 67% - 88% difference. Still a big difference. But, we still don't know what they are actually depicting here. It could be a lot bigger or a lot smaller... There are 4 possibilities...
1) Both the XSX and PS5 are sequential read
2) The XSX is sequential and the PS5 is random read
3) The PS5 is sequential and the XSX is random read
4) Both the XSX and PS5 are random read
Considering that even in simple tests for current PCI-E 4.0 SSDs, only sequential reads pass the 4GB/s mark, it is highly likely that the given PS5 spec is sequential read, meaning it will not be representative of actual transfer speeds. The same applies for the XSX. And more importantly, value for sequential reads is not a guarantee for high values in random reads... Use this as a reference;
Sabrent Rocket Q 1 TB M.2 NVMe SSD Review
Priced at $130 for the 1 TB version, the Sabrent Rocket Q SSD is one of the most affordable SSDs on the market. It uses a new Phison E12S controller, which has better thermal performance and can operate with smaller DRAM cache. Write speeds are impressive due to 250 GB of pseudo-SLC cache.www.techpowerup.com
And lastly, remember that we're talking about the I/O. Games will have to work with an additional SSD as well. So technically, the same I/O should be able to handle multiple storage devices, meaning, the games will never be programmed for the max I/O speed, simply because it could technically be cut in half if you have a second SSD that wants to load something at the same time.
Edit: I should add this, since most people don't know how to convert it... But, the fastest result in here, the 73788 IOPS from the ADATA SX8200 Pro, is actually just 288MB/s. Compared to the sequential read, it's literally 10 times slower. This is a worst case scenario, so we should expect at least 500 MB/s from these SSDs for games... But don't delude yourselves into thinking we'll get GB/s in real-time from the consoles.
Would the highly parallel setup of the PS5's SSD help with random read speed at all?
Ass of Can Whooping
Member
You can nitpick titles as much as you want. It doesn't change the fact that they aren't just going to wait for everyone to transition,.
James Sawyer Ford
Member
Is it going to be 120%? I doubt it. Considering that both have a separate hardware for decompression, no one will be using the decompressed data. It's all going to be compressed. So the difference will be between the 4.8GB/s and the 8-9GB/s, which translates to 67% - 88% difference. Still a big difference. But, we still don't know what they are actually depicting here. It could be a lot bigger or a lot smaller... There are 4 possibilities...
1) Both the XSX and PS5 are sequential read
2) The XSX is sequential and the PS5 is random read
3) The PS5 is sequential and the XSX is random read
4) Both the XSX and PS5 are random read
Considering that even in simple tests for current PCI-E 4.0 SSDs, only sequential reads pass the 4GB/s mark, it is highly likely that the given PS5 spec is sequential read, meaning it will not be representative of actual transfer speeds. The same applies for the XSX. And more importantly, value for sequential reads is not a guarantee for high values in random reads... Use this as a reference;
Sabrent Rocket Q 1 TB M.2 NVMe SSD Review
Priced at $130 for the 1 TB version, the Sabrent Rocket Q SSD is one of the most affordable SSDs on the market. It uses a new Phison E12S controller, which has better thermal performance and can operate with smaller DRAM cache. Write speeds are impressive due to 250 GB of pseudo-SLC cache.
And lastly, remember that we're talking about the I/O. Games will have to work with an additional SSD as well. So technically, the same I/O should be able to handle multiple storage devices, meaning, the games will never be programmed for the max I/O speed, simply because it could technically be cut in half if you have a second SSD that wants to load something at the same time.
Edit: I should add this, since most people don't know how to convert it... But, the fastest result in here, the 73788 IOPS from the ADATA SX8200 Pro, is actually just 288MB/s. Compared to the sequential read, it's literally 10 times slower. This is a worst case scenario, so we should expect at least 500 MB/s from these SSDs for games... But don't delude yourselves into thinking we'll get GB/s in real-time from the consoles.
ps5’s top speed is 22gb/s, won’t be reached commonly but it has a much much higher ceiling than XSX
Games developed from the ground up for these consoles will take advantage of it. PC benchmarks are not a valid comparison.
we’ll see the biggest advantage in first party, but we’ll also see a difference in third party games. Better texture detail, diversity, and less noticeable LOD
Ascend
Member
Yes. I think that's why they can reach the 5.5GB/s sequential read speeds. I don't think they'd advertise it if they can't reach it. So it's already factored in, so to speak. We can theoretically expect a 70%-80% advantage on the PS5 for the SSD over the XSX. But I suspect it will be reduced somewhat (hard to say by how much), since random reads are unpredictable. It's not called 'random' for nothing.
SonGoku
Member
The I/O wont need to access both SSDs simultaneously why would it limit anything, it will just access whichever drive has the game you are currently playing
4.8GB/s figure is a peak only for textures while 8-9GB/s is a typical figure for all data, don't know about xbox but ps5 has fine grain hw (sram) and sw (id system) customizations to max out reads and minimize
the impact of random reads
xsx wouldn't be inmune to this so why would the difference be reduced?
Last edited:
AveragePhoenix
Member
Most will. The next AC will be cross-gen, the next 2 call of duty will be cross-gen, the next Battlefield will be cross-gen, and the next 5 Fifas will be cross-gen.
Shmunter
Member
Nope. It’s all a guessing game at this point how next gen software distribution and labelling will work.
Bo_Hazem
Banned
I think Xbox team should concentrate on widening the gap between XSX and X1X from 4.6x to a much better number. So far that's an equivalent of 0.46GB/s if compared to 100MB/s max of the X1X or 0.23GB/s compared to 50MB/s minimum.
Last edited:
Rusco Da Vino
Member
Do you have any source to support your claims or are you going to keep spreading the same BS in this thread all night?
Time to cut the meta PC crap here. You are in the wrong thread for that.
Move on.
Last edited:
Metroiddarks
Member
But also most pc gamers use a weaker gpu around 96% of pc according to steam (only compare to PS5 in brute force), cpu also is weaker more than 90% use
less than 8 cores. So according to your logic now the PC gaming will be the ballast of the generation and not all the deltas is fixing just changing the resolution
or some setting in shadows.
https://store.steampowered.com/hwsurvey/videocard/
https://store.steampowered.com/hwsurvey/cpus/
If you like PC gaming please don't degrade it like that. The changes in level design for the use of SSD will not works in pc without a SSD or some another
solution like cache a huge quantity of data in RAM.
Really how old are you ? Is the first change of generation you live of what ? Because this sound to me like a person only who only listen a youtuber which build
pc as business or something similar.
Ass of Can Whooping
Member
This year's Assassin's Creed will obviously be cross-gen. The next one won't. Don't expect the next Battlefield to be cross gen
Funny how your only confidence lies in money milking franchises.
SonGoku
Member
By second year there should be next gen only multiplatforms by 3rd year it'll mostly be next gen only games
Ascend
Member
We technically don't know the ceiling of the XSX. It's quite likely that the PS5's is indeed higher. I don't think it's 10 times higher though. But this is all speculation.
Actually, they are in this case, because they are synthetic. This means that the benchmarks are designed to reach the limit of the SSD. The point was not to show the max speeds of the SSD, but rather the difference between sequential and random reads, and that not all SSDs behave the same. It is impossible to have the same read speed for sequential and random, that much is true. And this applies to all hardware, including the PS5 and XSX.
I don't see how this will be the case compared to the XSX.
And I'm asking again where you got that 4.6x from.
AveragePhoenix
Member
So what should I reference? Indie games? Racing games? AA? I'm listing the bestsellers. The ones that people play the most every year.
Kazekage1981
Member
I think Bo_Hazem is trolling again
Bo_Hazem
Banned
So you didn't open the spoiler in the previous post? I've put it there (I don't want to repeat the same stuff again and again, but let's do it):
Xbox Series X (11sec) vs Xbox One X (51sec): State of Decay 2. Difference is only 4.6x.
PS5 pre-devkit state (1 year ago, 0.8sec) vs PS4 Pro (8sec): Spider-man. Difference is 10x, with WIRED reporting 0.8 vs 15sec on another test (18x), questioning the 0.8sec being due to other stuff happening inside the system before loading. Plus it's reported to be a slower version. No need to take all that talk as anything, just pay attention to actual videos we can see:
Until we see further proof in the future, we better stick to actual, visible evidence.
DeepEnigma
Gold Member
As are others then as well, minus sources of their evidence to their claims.
Last edited:
Ass of Can Whooping
Member
I just told you. The big publishers. Fifa and COD don't make up the majority of those games.
Ascend
Member
Nah... The 4.8GB/s and 8-9GB/s figure is simply the difference between compressed and uncompressed. It applies to everything including textures... Textures are also compressed and need decompressing to be used.
If you look at the results from the SSDs I posted above, the difference between the fastest and the slowest SSD in sequential reads is 2470MB/s. The fastest is literally 495% faster than the slowest.
Take the random read speeds, and that difference is reduced to 39153 IOPS, which is 153MB/s. Now, the fastest drive is only 113% faster.
SonGoku
Member
Theres plenty of GPU hardware without extra units for redundancy (caches, aces, rops, ge etc). I think the odds of the specific tiny TE ASIC to be affected by a cripling defect is low enough to not have an impact on yields
AveragePhoenix
Member
Tell me a game by a major publisher coming out in the next 2 years that's not going to be multiplatform. You can just guess, as we really don't know, but I'm curious
It's like Acend didn't watch Cerny's presentation at all. Nobody seems to have.
The SSD was designed to address those issues and limitations. So you get the exact speed. No bottlenecks and other usual limitations. Stop comparing brute PC hardware to custom built consoles.
For the millionth time. Watch and understand how the SSD was made. The SSD specs will fully translate into games. Don't know if SX will be like this. Looks like they're using a standard SSD with a standard controller.
"Every single bottleneck needed to be addressed. And there are alot of them."
The SSD was designed to address those issues and limitations. So you get the exact speed. No bottlenecks and other usual limitations. Stop comparing brute PC hardware to custom built consoles.
For the millionth time. Watch and understand how the SSD was made. The SSD specs will fully translate into games. Don't know if SX will be like this. Looks like they're using a standard SSD with a standard controller.
"Every single bottleneck needed to be addressed. And there are alot of them."
Last edited:
Ass of Can Whooping
Member
Battlefield aint going to be cross-gen. Guaranteed
Dragon Age if it manages to come out by 2022
The next Assassin's Creed game after this one.
Starfield
Beyond Good and Evil 2
That Harry Potter game
Rocksteadys Game
Last edited:
James Sawyer Ford
Member
I dont see how it won’t
Higher SSD read speeds = better vram utilization, less wasted data
the benchmarks aren’t running games built for the ground up to optimize SSD usage w/r/t detail, diversity, or better LOD
Ascend
Member
You really aren't paying enough attention to the thread, but here:
Xbox Series X (11sec) vs Xbox One X (51sec): State of Decay 2. Difference is only 4.6x.
PS5 pre-devkit state (1 year ago, 0.8sec) vs PS4 Pro (8sec): Spider-man. Difference is 10x, with WIRED reporting 0.8 vs 15sec on another test (18x), questioning the 0.8sec being due to other stuff happening inside the system before loading. Plus it's reported to be a slower version. No need to take all that talk as anything, just pay attention to actual videos we can see:
Until we see further proof in the future, we better stick to actual, visible evidence.
+++
And for 6GB VRAM joke:
RE2 Remake, some exeeding 14GB VRAM:
You can't compare those two loading demos. One is an actual game with no optimizations. The other is a fixed sequence with a fixed camera and fixed path. The latter is easily going to be more efficient since there is less variance. The game on the Xbox is still using the old method of loading everything. The demo of the PS5 is loading a specific scene, and then it is trying to use the new method on the PS4, showing how it freezes but the PS5 does not. Apples and oranges.
As for VRAM usage... The fact that you have to resort to a VRAM report bug to somehow prove that games use more than 6GB of VRAM at 4K says it all... Want proof? Here's a timestamped video by Digital Foundry. Starting at 15:49 to 16:10. Just FYI, they say the game runs without stutter at 4K with a GTX 1060 that has 6GB of VRAM.
I guess you've already proven that you have no idea what you're talking about.
Last edited:
Kazekage1981
Member
Seriously?!
- Status