Remix.run Logo
aboringusername 12 hours ago

Games publishers/developers are going to have to wind in their necks a little. Whilst memory is abundant it's also still quite expensive. We should still be aiming for efficiency and the chances are 16gb+ are in the minority here. Fact is, the more VRAM and compute you demand the smaller your customer-base becomes.

I've played many games with 8GB VRAM* and will do so for the forseeable. If that's not enough, I am not a customer. Simple as.

The truth is, there is going to be a massive motivation with the likes of Steam Deck/Machine to actually make titles that are optimised and perform well within their hardware parameters. It's money you won't want to ignore.

*One example was Silent Hill remake on PC, which used the unreal engine. It was optimised beautifully and ran without visual glitches and stutters even with the highest graphic demands on a 8GB RTX

esskay 11 hours ago | parent | next [-]

I think it does also help that a big chunk of Steams userbase are playing smaller indie titles that don't need obscene amounts of vram. The steam deck audience for example has a lot of people playing both a mix of AAA and smaller games. Given this is advertised as 6x as powerful as the deck I'm sure they'll be fine. It's not meant to be a top of the line console thats for sure, and if it was people would be moaning that its too expensive.

lawlessone 10 hours ago | parent [-]

Not using the highest settings obviously but i can run RD2 and CP2077 on the deck fine.

It will be a while before there is ps6 or new xbox.

esskay 7 hours ago | parent [-]

oh 100% I've completed CP, RD2, Fallout 4, and god knows how many other games, it handles it all like a champ. Valves clearly following their own hardware survey results on their hardware plans as the modest specs are better than what most people active on steam are using right now so I think it'll be fine

SchemaLoad 8 hours ago | parent | prev [-]

Memory is also not that abundant anymore. Over the last month PC memory costs have more than doubled due to AI datacenter builds buying out all the manufacturing capacity.