By using this site, you agree to our Privacy Policy and our Terms of Use. Close
SvennoJ said:

@Permalite Procedural generation can be used to save memory, but more so on disk since you still need all that procedurally generated stuff in memory. The simplest way for CPU to help RAM is to keep all data compressed in memory. For textures that's done by the GPU nowadays, yet world data can also be kept compressed and decoded on the fly. That's what we did when I still worked in GPS navigation. All the map data was compressed to the max with hash tables for random access, loading in chunks at a time to decode while searching or rendering the map. All with efficient memory use, storing everything in RAM with as few bits as necessary. More work for the CPU to decode and encode all the time, yet RAM was the bottleneck. For games, think of Minecraft data all being compressed in memory to allow for bigger worlds.

And the other way around RAM can help the CPU by keeping more stuff around and pre-loading / preparing stuff when time left over. So for procedural generation, more RAM allows you to work ahead, buffering basically. Thus when the user decides to sprint ahead, you already have it available. Culling is great for rendering, but as FS2020 showed, actually throwing away the data from memory is problematic when you look around. It's fine if you stare straight ahead all the time, yet for example in VR you can't purge the world behind the user as you can look around at any time. (Well you can in most games)

And yep SSD and direct to video ram loading helps. I still found using a ram disk for cache faster than SSD, but just leaving it in RAM (turning off the aggressive culling in FS2020, which they mistakingly named pre-chaching, works the best obviously. It's more post-caching if you leave the stuff behind in ram until it's out of range. Anyway just means do do double work in the end)

More RAM also makes development easier and faster. Optimization is hard work and easy to introduce hard to trace bugs. Pointer arithmetic for example is great way to optimize code and reduce memory use, but so easy to introduce hard to find fatal errors.

No "r" in my name. So I don't get alerted.

And you are right, compression and loading in chunks is an age-old game design philosophy to work around memory constraints.


And yes, Ram disk will be faster than an SSD, you are running with orders of magnitude more performance, an SSD is just more efficient at streaming than mechanical hard drives, so it doesn't need to pre-cache assets as earlier to keep things running smoothly.

Kyuu said:
Pemalite said:

I understand and agree fully, but Consoles tend to be supported for the entire generation.

I always want the graphics bar to be raised more, I would have liked to have seen the Series X and Playstation 5 to offer more hardware, but the current climate didn't allow for it, hence the mid range hardware.

The Series S has been a success for Microsoft, so it is not going away.

Consoles get "full" (or close to full) 3rd party support throughout a generation when they're adequately powerful and popular, Series S is neither of these. Some developers will not want to be stuck for 10-13 years to a system that is both weak and unpopular. So far it looks like PS5's specs/price is perfect for the current economics, so it (and comparable consoles/PC's) should be the new standard. Mandating Series S would lead to plenty of ambitious PS5/PC games skipping Xbox. MS should take a PC like approach (system requirements) at this point and officially make Xbox a full blown hybrid between console and PC. It would piss off some early adopters, but the pros outweigh the cons... I think.

As far as I'm concerned Xbox's current relative success has nothing to do with the Series S. But again, by the end of 2023, a lot of opinions including some of my own will change. This year should be a lot more indicative than previous years as to where the market is heading.

Series S hasn't been out for a generation yet, so it's hard to make that call on how supported it will be.

Xbox One was weak and unpopular, but still garnered 3rd party support all generation long.

EpicRandy said:
Pemalite said:

Check your math on that. I am talking physical Ram not usable Ram.

Think my math is correct on this. Series S is 10GB and X 16GB 10/16 = 0.625 meaning a loss of 37.5%. 16GB*0.375 = 6GB (the diff between X and S)

6 is 60% of 10.

The increase from 10 to 16 is 60%.



Comes down to how you frame your percentages/question.

SKMBlake said:
EpicRandy said:

My point was it's very odd to consider Series X and PS5 as running well here at max 1440p with framerate issues. Even more odd when you would only consider 1440P 30 fps as the threshold for the S to be also running well. I mean if series S achieved such target I wouldn't expect anything less than 4k@30 maybe even 4k@60 for both PS5 and X. 

I know. And my point was: it's okay to have sub-4k@60fps games, it's still better than the 1080p@30fps we got from most of the last gen games. 

But next-gen at 792p@30fps, nah.

You are missing the point of the Series S. It's not about high graphics, high framerates, it's about being a cheap entry point to console gaming... Because lets face it, many nations are struggling financially, inflation is running rampant, costs everywhere are increasing.

The Series S is the counter to that trend so people can keep playing new games, cheaply.



--::{PC Gaming Master Race}::--