[For those who don't know, Quo vadis means "Where are you going?"]
I've searched for similar topics, or even a Steam OS Discussion thread, but only found the general threads for Steam. If I've missed something, vulkanize me.
As many others in this forum, I was very interested in SteamOS, but was taken aback by the lack of performance in its current state, which sadly caused a number of OEM manufacturers to freeze their efforts to bring a steam machine to the market. Understandably so: We weren't talking about about 2-3 percent, the drop in performance compared to Windows 10 would in some cases make the difference between "playable" and sub-30fps. For now I'll ignore 60fps-GAF's opinion on this matter.
Now as far as I know, there's nothing new to report on the OpenGL performance front: I've searched the web, but this topic hasn't been picked up again yet, although I have little doubt it will at some point.
There is a reason why I'm making this topic nowish though. No thread was made for this, but I think this is kind of a big deal: SteamOS natively supports Vulkan since February 22nd... So this got me thinking.
First of all, and this is also a question for GAF since I couldn't find a coherent answer on this: Vulkan doesn't change the performance of games that weren't built on it. Is this correct? So for example, Left for Dead 2, which was optimized for OpenGL, won't perform better through Vulkan, even though it's OpenGL's successor.
If this were true (it isn't according to Durante, but let's assume it is), here is why it probably isn't a big deal: Even though the game catalogue unoptimized for Vulkan is huge, this problem posts only a timed disadvantage at best. Most of these unoptimized games (let's estimate all games pre-2006, even though this is not entirely true) will run perfectly fine through raw CPU/GPU power, which in most cases even an Intel/Intel HD combo is able to accomplish. So as CPUs and GPUs get stronger in quick cycles even today, the games that are running in a sub-par fashion nowadays will work eventually if you throw enough TFLOPS at them, so to speak. Which would mean that the only question that matters is (and I guess the entire discussion boils down to this)...
...will Vulkan catch on, and if it does, do you believe SteamOS will gain market percentage, or will it eventually die a slow painful death? Is it already dead, since its market share plummeted below 1% on Steam [citation needed]? Are there other reasons why you believe people (users, not OEMs) are staying away from SteamOS?
Did Valve put too much on its plate by initiating a war with two frontiers, consoles on one side and Windows 10 on the other?
Link to the GAF Vulkan Thread
To clarify, this thread is meant to focus on the SteamOS-factor rather than Vulkan itself - that's what the Vulkan thread is for.
I've searched for similar topics, or even a Steam OS Discussion thread, but only found the general threads for Steam. If I've missed something, vulkanize me.
As many others in this forum, I was very interested in SteamOS, but was taken aback by the lack of performance in its current state, which sadly caused a number of OEM manufacturers to freeze their efforts to bring a steam machine to the market. Understandably so: We weren't talking about about 2-3 percent, the drop in performance compared to Windows 10 would in some cases make the difference between "playable" and sub-30fps. For now I'll ignore 60fps-GAF's opinion on this matter.
Now as far as I know, there's nothing new to report on the OpenGL performance front: I've searched the web, but this topic hasn't been picked up again yet, although I have little doubt it will at some point.
There is a reason why I'm making this topic nowish though. No thread was made for this, but I think this is kind of a big deal: SteamOS natively supports Vulkan since February 22nd... So this got me thinking.
First of all, and this is also a question for GAF since I couldn't find a coherent answer on this: Vulkan doesn't change the performance of games that weren't built on it. Is this correct? So for example, Left for Dead 2, which was optimized for OpenGL, won't perform better through Vulkan, even though it's OpenGL's successor.
This is not correct.
To fully leverage the performance advantages of Vulkan (or any other low level API) you have to target it from the ground up.
However, even a direct port can have significant performance advantages. In the initial Vulkan version of The Talos Principle, described by the developer as a quick port, I measured up to a 50% performance advantage over OpenGL in CPU limited scenarios. (On a beta Vulkan driver no less)
Of course, a direct port still requires some time and effort, it's not a 1:1 mapping.
If this were true (it isn't according to Durante, but let's assume it is), here is why it probably isn't a big deal: Even though the game catalogue unoptimized for Vulkan is huge, this problem posts only a timed disadvantage at best. Most of these unoptimized games (let's estimate all games pre-2006, even though this is not entirely true) will run perfectly fine through raw CPU/GPU power, which in most cases even an Intel/Intel HD combo is able to accomplish. So as CPUs and GPUs get stronger in quick cycles even today, the games that are running in a sub-par fashion nowadays will work eventually if you throw enough TFLOPS at them, so to speak. Which would mean that the only question that matters is (and I guess the entire discussion boils down to this)...
...will Vulkan catch on, and if it does, do you believe SteamOS will gain market percentage, or will it eventually die a slow painful death? Is it already dead, since its market share plummeted below 1% on Steam [citation needed]? Are there other reasons why you believe people (users, not OEMs) are staying away from SteamOS?
Did Valve put too much on its plate by initiating a war with two frontiers, consoles on one side and Windows 10 on the other?
Link to the GAF Vulkan Thread
To clarify, this thread is meant to focus on the SteamOS-factor rather than Vulkan itself - that's what the Vulkan thread is for.