I'd love to hear what an optimised load is in this regard as it falls back to loading what is visible. A thing people say already happens on all old xbox one games/engines.
-
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.
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
- Thread starter Bernkastel
- Start date
- Platform
Ascend
Member
Which is a blatant lie. And even IF it was true (which it's not), the advertised 2x-3x effective bandwidth increase was in comparison to the Xbox One X, so... The improvement is still there regardless. But it gets dismissed as 'secret sauce' or or 'magic' or whatever other dismissive term is available.
As for what an optimized load is, it's quite simple. It means it's optimized for what you're actually using. Currently, the games are optimized for an HDD, but not for an SSD. By default it makes it inefficient for the SSD. Any game that is BC and is dropped into the XSX, even though it will work and load a lot faster, is far from the best that the console can achieve.
You're still loading everything from the perspective of HDD technology, which means you're loading as if you have seek times, as if you need to use duplicate data on disc, and as if you're limited to double digit MB/s. You will have unnecessary CPU overhead, unnecessary idle time and you most likely even have duplicate readings that are not necessary, reducing the SSD performance unnecessarily.
And above all that, in the case of loading an untouched Xbox One game on the XSX, you're not taking advantage of the hardware decompression block, the newer decompression techniques, or the lower overhead API.
I hope we can finally lay this to rest.
too much eslim
Member
If someone doesn't understand the concept that none of the games shown were optimized then they are set in their agenda. No point in trying to explain anything else.
I don't why still going on Microsoft is clearly lying they recycled the io from the OG xbox to maintain backwards compatibility and thew the SSD in at the last second phil went to the local best buy and busted out his corporate credit card and bought a few million SSDs. That kid at best buy almost pissed him self having to carry that many to uncle Phil's low rider caddie. He got him self enough reward points to get a switch and ps5 out of the deal!
Last edited:
Wher
Where did you get the confirmation regarding the code optimization?
MoreJRPG
Suffers from extreme PDS
6.5 or 10 seconds is still absurdly fast. Not sure what people are trying to nitpick(well, I mean, I know why they're nitpicking.) After that 10 seconds it never loads again and you forget it ever happened.
Last edited:
Andodalf
Banned
not to mention the visible improvements in pop in
Trueblakjedi
Member
Because you asked nicely
Xbox Series X loading times demo shows State of Decay 2 load in just seven seconds
Xbox Series X changes the game for load times.
www.windowscentral.com
State of Decay X One X vs X Series X loading demo
Last edited:
I think he is asking where it states there was optimisation done for spiderman. Where did you hear such a thing?Because you asked nicely
"State of Decay 2 goes from loading in about 45 seconds on Xbox One X to just seven seconds on Xbox Series X. Even more impressively, this is with no extra work or optimization done, that's just how much State of Decay 2 benefits when installed on an Xbox Series X."Xbox Series X loading times demo shows State of Decay 2 load in just seven seconds
Xbox Series X changes the game for load times.
State of Decay X One X vs X Series X loading demo
JareBear: Remastered
Banned
Ascend
Member
Just for the record;
Both the PS5 and the XSX have hardware decompression that does not rely on the CPU.
Both the XSX and PS5 designed their SSDs to have an as constant as possible throughput without performance falling off a cliff after a short period of time, which is an advantage over the average PC SSD.
Both the XSX and the PS5 reduce/eliminate the driver overhead that PCs have.
The following is contrary to popular belief, but Sony took a brute force approach to the SSD with the PS5, while the XSX has the transfer efficiency approach.
And if you want an answer to your question regarding the spidey demo vs the XSX loading, check these two posts of mine;
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
And this is just ( one of ) the reason why i had right to doubt about Discord. Even members of ordinary Xbox Discord are doing that. Continuously spreading FUD about PS5 is astonishing Again that post AND THIS thread is about XVA, SFS etc...in the XSX. Not about PS5 directly. Only PS5 mentions...
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
The ps4 Spiderman demo load was 8.10 seconds on a PS4 (im sure that that was NOT the whole game) The optimized demo loaded in .8 seconds. The State of Decay demo was a 45 second full game load. That same game with no optimization loaded in 6.5 seconds. Sure an optimized load would be less...
This is an Xbox Series X thread. Let's try and keep it about that, shall we? If you want to focus on the PS5 tech, there are a gazillion other threads for that, including this one and this one.
thicc_girls_are_teh_best
Member
That method actually might have a potential disadvantage/tradeoff that I've thought of, but I'm gonna sit on it for right now
Last edited:
Fair enough, however I just wanted to know if it was confirmed or not that Spidey demo was actually a different code of the orignal game since it was mentioned right here in this thread..Just for the record;
Both the PS5 and the XSX have hardware decompression that does not rely on the CPU.
Both the XSX and PS5 designed their SSDs to have an as constant as possible throughput without performance falling off a cliff after a short period of time, which is an advantage over the average PC SSD.
Both the XSX and the PS5 reduce/eliminate the driver overhead that PCs have.
The following is contrary to popular belief, but Sony took a brute force approach to the SSD with the PS5, while the XSX has the transfer efficiency approach.
And if you want an answer to your question regarding the spidey demo vs the XSX loading, check these two posts of mine;
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
And this is just ( one of ) the reason why i had right to doubt about Discord. Even members of ordinary Xbox Discord are doing that. Continuously spreading FUD about PS5 is astonishing Again that post AND THIS thread is about XVA, SFS etc...in the XSX. Not about PS5 directly. Only PS5 mentions...www.neogaf.com
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
The ps4 Spiderman demo load was 8.10 seconds on a PS4 (im sure that that was NOT the whole game) The optimized demo loaded in .8 seconds. The State of Decay demo was a 45 second full game load. That same game with no optimization loaded in 6.5 seconds. Sure an optimized load would be less...
This is an Xbox Series X thread. Let's try and keep it about that, shall we? If you want to focus on the PS5 tech, there are a gazillion other threads for that, including this one and this one.
If we also go back some pages we can find several twitts related to PS5 SSD trying to imply that Sony's solution is not abble to stream assets from SSD direct to the GPU...
So, this is the season why I thought it would be relevant...
Last edited:
thicc_girls_are_teh_best
Member
I've personally speculated that and stand by it, because if you watch for the graph in Road to PS5 that shows the I/O complex, GPU/CPU APU, and system memory they highlight the I/O complex having DMA to the RAM but never actually mention anything about SSD asset streaming to the GPU.
Which, I think given the focus Cerny had on the SSD at that presentation and how it was targeted at developers, that would've been the perfect time to mention it. Granted, I don't think PS5 actually needs that feature to make optimal use of its SSD, since it can send data to and from RAM so quickly as-is, and that seems to be the approach Sony has favored. There's a lot of nuance to it though, like anything tech-related.
Bernkastel
Ask me about my fanboy energy!
yup he is.
Louise Kirby is a dude
He basically continued the conversation after that confirming what I said.
Bernkastel
Ask me about my fanboy energy!
Definitely more than 2,3
And all of this before the actual game reveal event on July. They may still reveal something on the June.
JareBear: Remastered
Banned
Bernkastel
Ask me about my fanboy energy!
Other exclusives
Last edited:
oldergamer
Member
I thought it was June 12th?
Bernkastel
Ask me about my fanboy energy!
Date yet to be revealed
June Xbox 20/20 event rumoured to focus on the Xbox Series X itself
Microsoft's June Xbox 20/20 event, part of a monthly reveal of the Xbox Series X and its games, is rumoured to focus on the Xbox Series X itself. Microsoft d
stevivor.com
Bernkastel
Ask me about my fanboy energy!
thats PS5.
jimbojim
Banned
Last edited:
Bernkastel
Ask me about my fanboy energy!
funking giblet
Member
What a whole load of horseshit from a layman...
What do you think they actually explained there?
What do you think they actually explained there?
Last edited:
JareBear: Remastered
Banned
GamingBolt interview with Scorn dev just went up, a couple quotes about SSDs and Velocity architecture
Last edited:
Bernkastel
Ask me about my fanboy energy!
How many interviews is this guy gonna make ?
GamingBolt interview with Scorn dev just went up, a couple quotes about SSDs and Velocity architecture
Dory16
Banned
Was it too complex for you?
funking giblet
Member
No, it was made up nonsense?
Did you not get that? Maybe I can go through it in fine detail!?
Dory16
Banned
Go ahead. What did he say that was made up?
funking giblet
Member
Let's see
They are conflating using Bank Switching, which he says was directly addressable by memory in a CPU (not what it was used for, and no, it's not addressable, hence SWITCHING) Used in the old days when CPUs had limited address space (They can address Terabytes of memory now directly), for something which now supersedes it.Virtual Memory (which exists in any OS these days and has done for years).
By passing VRAM to read directly into the GPU, that's just not a thing, sorry. Made up as it were. CPUs use read-through caching (the through being, RAM). Caches are too small and variable to actually pre-warm them effectively, and you certainly wouldn't need an SSD for 96kb of L1 cache or 4mb of L2 cache (numbers might be off, but close).
XVA is the following.
Hardware compression, SFS + support for DIrectStorage (and a nice SSD).
They most likely reserve either a partition or dedicated Page file (100gb) so a game can wipe, and use it.
SFS is a hardware feedback buffer which replaces software ones used for a number of years now. Hardware filtering already exists for PRT (ask me why hardware works for PRT better, and why you need it).
I can even explain how feedback buffers work and why SFS won't actually help if your game already uses a buffer to figure out your current residency, as well as how effective they are in software.
Or anything else?
Last edited:
CurtBizzy
Member
So the SSD is basically ram ?
Ascend
Member
What about the hardware texture filters?
Dory16
Banned
No doubt that you know your subject.
In sum, I gathered that the 100GB are a portion of the SSd that the XSX is capable of making the CPU "treat" as VRAM, just very slow VRAM. And the CPU has therefore almost no overhead in accessing that data.
What it can do with it seemed more blurry, he said either discarding it (big deal then) or writing it to RAM (that bus is 560GB/s, so we are 20 times beyond the best SSD speeds and I see why that would be interesting).
You are saying based on your expertise that it is impossible to happen as such and therefore made up or that you just don't believe it because of some other information you have about the Velocity architecture?
Last edited:
funking giblet
Member
Well let's talk about what it brings to the table.
When you are traditionally filtering a texture, you grab the full texture, create a sampling kernel function of some size and algorithm, and run the kernel over the image to create filtered image.
A high quality / filtered texture spread over a flat plane would look very weird at distance, so you may move to anisotropic filtering, and include sampling the various mipmaps to improve rendering at a distance.
With PRT, your textures are split across pages, and of course, mips. Just sampling them will mean you get weird seams when you join two pages with two parts of the same texture together, so you want those other pages too when you sample, and also their mips!
This is a lot of data to track in software, and is a pain in the ass speed wise. Hardware which stores these locations, and improves data locality as well as doing the filtering makes this a better experience for the dev.
The issue with this feature is it exists in hardware for the last 11 years.
The feature you describe is just plain ol' Virtual Memory which the phone in your pocket can do. There can be optimizations around granting a dedicated partition or file for this purpose, but again, nothing I can't do right now on my PC (I mean, your bog standard web dev might tinker with this!). The SSD itself does make Virtual Memory more useful, so let them advertise it as so, but it's a rote feature.
All of this together is very very good for consoles, but I hesitate to round them all up, and say "XVA is the killer feature" when it's just a lot of features coming together with a fast disk too.
Last edited:
Dory16
Banned
Ok. So they're right but it's not new.
It could be knew in the console world. Would be strange of them to advertise that as part of their next gen glossary if it's a feature that every electronic device in the world already has, like you said.
The proof will be in the pudding.
funking giblet
Member
No I disagree fundamentally with how they stumbled around the topic and came up with just wrong answers. The issue I have is they are saying XVA somehow bypasses VRAM into the CPU because the CPU can reference the data on the SSD via something similar to Bank Switching.
You know when The Colonel says a bunch of gibberish at the end of MGS2? It's kinda like that.
You throw enough crap at a wall...
Last edited:
Ascend
Member
Then why are they calling it custom for the Xbox?
Seems like nothing the Xbox does is (allowed to be) new
.
Last edited:
funking giblet
Member
Nothing new under the sun. What is new is having a hardware decomp and a fast enough SSD to make Virtual Memory usage viable. Page swapping is slow. If SRS works automatically it maybe also will encourage use of PRT because the headaches of managing feedback is gone. The actual full tilt speed of the SSD really ain’t so important when playing a game, it’s about swapping data into RAM as fast as possible before the next frame. You may only need a little data but you want it to be ready to use ASAP.
Last edited:
sendit
Member
Just for the record;
Both the PS5 and the XSX have hardware decompression that does not rely on the CPU.
Both the XSX and PS5 designed their SSDs to have an as constant as possible throughput without performance falling off a cliff after a short period of time, which is an advantage over the average PC SSD.
Both the XSX and the PS5 reduce/eliminate the driver overhead that PCs have.
The following is contrary to popular belief, but Sony took a brute force approach to the SSD with the PS5, while the XSX has the transfer efficiency approach.
And if you want an answer to your question regarding the spidey demo vs the XSX loading, check these two posts of mine;
Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
And this is just ( one of ) the reason why i had right to doubt about Discord. Even members of ordinary Xbox Discord are doing that. Continuously spreading FUD about PS5 is astonishing Again that post AND THIS thread is about XVA, SFS etc...in the XSX. Not about PS5 directly. Only PS5 mentions...Xbox Velocity Architecture - 100 GB is instantly accessible by the developer through a custom hardware decompression block
The ps4 Spiderman demo load was 8.10 seconds on a PS4 (im sure that that was NOT the whole game) The optimized demo loaded in .8 seconds. The State of Decay demo was a 45 second full game load. That same game with no optimization loaded in 6.5 seconds. Sure an optimized load would be less...
This is an Xbox Series X thread. Let's try and keep it about that, shall we? If you want to focus on the PS5 tech, there are a gazillion other threads for that, including this one and this one.
Yep, lets call PS5s approach to I/O brute force.
Last edited:
Ascend
Member
Do you have anything of value to add or are you just going to make baseless posts as feeble attempts at ridicule?Yep, lets call PS5s approach to I/O brute force.
sendit
Member
Just laughing at your baseless post. Nothing to see here.
Ascend
Member
I guess it's the latter then. Alright. Thanks for telling me upfront. I can safely put you on ignore without losing anything of value. Goodbye.
Is what is being said impossible, or is what is being said simply different than the traditional way of doing things? I am asking because, sometimes people have trouble seeing new possibilities when they have been in something for too long. Like, it's hard for a tennis player to bat properly in baseball, and vice versa.
sendit
Member
Neo_game
Member
I knew it. A college going chick who happens to code game engine for Xbox, Switch in her free time. Just did not sound right. lol
Trueblakjedi
Member
Texture residency hardware enhancements for graphics processors
WO2018151870A1 - Texture residency hardware enhancements for graphics processors - Google Patents
Systems, methods, apparatuses, and software for graphics processing systems in computing environments are provided herein. In one example, a method of handling tiled resources in graphics processing environments is presented. The method includes establishing, in a graphics processing unit, a...
patents.google.com
[0019] A first enhancement includes a hardware residency map feature comprising a low-resolution residency map that is paired with a much larger PRT, and both are provided to hardware at the same time. The residency map stores the mipmap level of detail resident for each rectangular region of the texture. PRT textures are currently difficult to sample given sparse residency. Software-only residency map solutions typically perform two fetches of two different buffers in the shader. , namely the residency map and the actual texture map. The primary PRT texture sample is dependent on the results of a residency map sample. These solutions are effective, but require considerable implementation changes to shader and application code, especially to perform filtering the residency map in order to mask unsightly transitions between levels of detail, and may have undesirable performance characteristics. The improvements herein can streamline the concept of a residency map and move the residency map into a hardware implementation.
[0020] A second enhancement includes an enhanced type of texture sample operation called a "residency sample." The residency sample operates similarly to a traditional texture sampling, except the part of the texture sample that requests texture data from cache/memory and filters the texture data to provide an output value is removed from the residency sample operation. The purpose of the residency sample is to generate memory addresses that reach the page table hardware in the graphics processor but do not continue on to become full memory requests. Instead, the residency of the PRT at those addresses is checked and missing pages are non-redundantly logged and requested to be filled by the OS or a delegate.
[0078] Example 19: A method of handling partially resident textures in a graphics processing unit (GPU), the method comprising establishing, in a GPU, a hardware residency map with values indicating residency properties of a texture resource, and sampling from the hardware residency map at specified locations to determine corresponding residency map samples each comprising an integer portion and fractional portion, where the integer portion indicates an associated lowest level of detail presently resident for the texture resource, and where the fractional portion represents an associated smoothing component for transitioning in a blending process with a next level of detail concurrently resident for the texture resource.
Thats from the HW texture filter patent.