Forums » General
System Requirements Update
The official System Requirements have been updated recently. (This is a PC-centric discussion, because Mobile requirements haven't really changed much).
As mentioned some time ago, we are going to be dropping support for 32bit Windows. There's a 64bit Windows install in closed testing right now (if you want to help test, ask Luxen to the "game-tester" privilege on Discord), and eventually we'll migrate to this completely-new installer, game layout, update architecture, and binary runtime.
None of this really supersedes what was written earlier this year in the Suggestions Forum RFC on System Requirements, although I'm now trying to aim for an official bottom-target of DX10_0 hardware instead of DX11_0 hardware. The earlier RFC thread is kind of "looking ahead" and the current update doesn't reflect everything we're moving towards, it's more a statement of current expectations.
For clarification: "DX10_0" and "DX11_0" are hardware specifications, they imply specific Shader Model capabilities. Where having the "DX11 API" doesn't mean what kind of hardware you have. I know this is a bit confusing for some people, so read on for a bit more detail..
While I will try to retain compatibility with legacy DX10_0 hardware (circa 2006), my focus is really on DX11_0 or newer (circa 2010). So, generally systems from 2010 or later should be reasonable (with a modern operating system). Otherwise, computers as old as 2006 should still work, but performance might not be great.
Our actual "recommended target", longer term, will be Vulkan 1.3, which usually mandates hardware starting around 2014 or later (although with AMD, this goes back to 2012 on Linux with the RADV driver). This isn't the minimum requirement, it's the recommended target, for the moment. For PC, we do still recommend a discrete GPU, like it says in the System Requirements, but we're doing a lot of development and testing on integrated GPUs like the Radeon 680M.
We are not testing below Windows 10 at this time, that's our official support limit. I suspect we'll continue working on Windows 7 SP1 or newer, for the time being, but no guarantees. I don't think Vista is going to work much longer (if it still does).
What is the point of this change? By raising the minimum we can actually make things perform a lot better on "modern" low-end hardware (like Chromebooks and low-end Windows laptops), because we can take advantage of hardware efficiencies that didn't exist in much older low-end hardware. That's really important to everyone who has a decade-old basic laptop (which is "modern" by our target set), and that's a lot more people than those running PCs from 2004.
This is not about mandating high-end gaming hardware, it's pretty much the opposite of that. We're trying to support the low-end. No one can afford computers anymore, let alone gaming GPUs (last I heard, a 5090 was $9,000?). So we're trying to "meet people where they are" and deliver the best results we can on the hardware available.
It also lets us upgrade the fidelity of the graphics and capabilities of the engine. And it doesn't negatively impact very many people. You'll still be able to run the game on a "potato", it'll just run more efficiently on a modern-potato, and perhaps not on an ancient-potato. At some point, one has to choose potatoes.
I think I'll end it on that mildly weird analogy.
If people have any questions, post here and I'll do my best to answer.
As mentioned some time ago, we are going to be dropping support for 32bit Windows. There's a 64bit Windows install in closed testing right now (if you want to help test, ask Luxen to the "game-tester" privilege on Discord), and eventually we'll migrate to this completely-new installer, game layout, update architecture, and binary runtime.
None of this really supersedes what was written earlier this year in the Suggestions Forum RFC on System Requirements, although I'm now trying to aim for an official bottom-target of DX10_0 hardware instead of DX11_0 hardware. The earlier RFC thread is kind of "looking ahead" and the current update doesn't reflect everything we're moving towards, it's more a statement of current expectations.
For clarification: "DX10_0" and "DX11_0" are hardware specifications, they imply specific Shader Model capabilities. Where having the "DX11 API" doesn't mean what kind of hardware you have. I know this is a bit confusing for some people, so read on for a bit more detail..
While I will try to retain compatibility with legacy DX10_0 hardware (circa 2006), my focus is really on DX11_0 or newer (circa 2010). So, generally systems from 2010 or later should be reasonable (with a modern operating system). Otherwise, computers as old as 2006 should still work, but performance might not be great.
Our actual "recommended target", longer term, will be Vulkan 1.3, which usually mandates hardware starting around 2014 or later (although with AMD, this goes back to 2012 on Linux with the RADV driver). This isn't the minimum requirement, it's the recommended target, for the moment. For PC, we do still recommend a discrete GPU, like it says in the System Requirements, but we're doing a lot of development and testing on integrated GPUs like the Radeon 680M.
We are not testing below Windows 10 at this time, that's our official support limit. I suspect we'll continue working on Windows 7 SP1 or newer, for the time being, but no guarantees. I don't think Vista is going to work much longer (if it still does).
What is the point of this change? By raising the minimum we can actually make things perform a lot better on "modern" low-end hardware (like Chromebooks and low-end Windows laptops), because we can take advantage of hardware efficiencies that didn't exist in much older low-end hardware. That's really important to everyone who has a decade-old basic laptop (which is "modern" by our target set), and that's a lot more people than those running PCs from 2004.
This is not about mandating high-end gaming hardware, it's pretty much the opposite of that. We're trying to support the low-end. No one can afford computers anymore, let alone gaming GPUs (last I heard, a 5090 was $9,000?). So we're trying to "meet people where they are" and deliver the best results we can on the hardware available.
It also lets us upgrade the fidelity of the graphics and capabilities of the engine. And it doesn't negatively impact very many people. You'll still be able to run the game on a "potato", it'll just run more efficiently on a modern-potato, and perhaps not on an ancient-potato. At some point, one has to choose potatoes.
I think I'll end it on that mildly weird analogy.
If people have any questions, post here and I'll do my best to answer.
Oh for heaven's sake.. you should have stopped supporting 15 year old software and OS versions some 10 years ago.. lol
There's a lot of potatoes in the wild - maintaining support for shitboxes isn't inherently a bad thing... :p
That said i'll concede that 32bit machines have become an exception, DX_11 hardware on the other hand is still all over the place and i'd argue the steam hardware survey is a pretty shit metric to go by since it mostly consist of nerds who want to be able to play crysis... :p
That said i'll concede that 32bit machines have become an exception, DX_11 hardware on the other hand is still all over the place and i'd argue the steam hardware survey is a pretty shit metric to go by since it mostly consist of nerds who want to be able to play crysis... :p
Oh for heaven's sake.. you should have stopped supporting 15 year old software and OS versions some 10 years ago.. lol
We make decisions based on empirical data. You would be surprised by what people are using to play VO (and not as an exception, but large numbers of hours).
Ironically, one of the first questions I had after posting this was from a Discord user asking if they'll still be able to continue playing on their 2011 Macbook with an Intel HD3000 GPU (and yes, I think they will, barring some Xcode change by Apple that's out of our control).
There are a lot of people who can't afford new computers, an issue that's only gotten worse recently. I like that our game is as broadly available as possible.
DX_11 hardware on the other hand is still all over the place and i'd argue the steam hardware survey is a pretty shit metric to go by
The problem with the steam hardware survey is that it's optional. So it isn't a picture of Steam's actual audience, it's a picture of the people who wanted to answer the survey. That tends to bias towards people who are proud of the systems they're using (or people who like to participate in these things). It also isn't even offered to everyone, only to a subset of their users, so it's a biased sample of a sample.
Again, we use our own data (we have to, we're cross-platform, far more than Steam). Our data is quite different from Steam's. My plans are all built around our own users, and delivering the best game performance and results that I can for the broadest number of real people who play our game.
We make decisions based on empirical data. You would be surprised by what people are using to play VO (and not as an exception, but large numbers of hours).
Ironically, one of the first questions I had after posting this was from a Discord user asking if they'll still be able to continue playing on their 2011 Macbook with an Intel HD3000 GPU (and yes, I think they will, barring some Xcode change by Apple that's out of our control).
There are a lot of people who can't afford new computers, an issue that's only gotten worse recently. I like that our game is as broadly available as possible.
DX_11 hardware on the other hand is still all over the place and i'd argue the steam hardware survey is a pretty shit metric to go by
The problem with the steam hardware survey is that it's optional. So it isn't a picture of Steam's actual audience, it's a picture of the people who wanted to answer the survey. That tends to bias towards people who are proud of the systems they're using (or people who like to participate in these things). It also isn't even offered to everyone, only to a subset of their users, so it's a biased sample of a sample.
Again, we use our own data (we have to, we're cross-platform, far more than Steam). Our data is quite different from Steam's. My plans are all built around our own users, and delivering the best game performance and results that I can for the broadest number of real people who play our game.
My integrated graphics is from 2013, Intel HD3000 or HD4000, working well in VO.
EDIT: the command execute with user show im using HD2000 :o
But working well anyway
--USER
Graphics:
Device-1: Intel 2nd Generation Core Processor Family Integrated Graphics
vendor: Elite Systems driver: i915 v: kernel arch: Gen-6 bus-ID: 00:02.0
Display: x11 server: X.Org v: 1.21.1.7 with: Xwayland v: 22.1.9 driver: X:
loaded: modesetting unloaded: fbdev,vesa dri: crocus gpu: i915
resolution: 1920x1080~60Hz
API: OpenGL v: 3.3 Mesa 22.3.6 renderer: Mesa Intel HD Graphics 2000 (SNB
GT1) direct-render: Yes
--ROOT
Graphics:
Device-1: Intel 2nd Generation Core Processor Family Integrated Graphics vendor: Elite Systems
driver: i915 v: kernel arch: Gen-6 bus-ID: 00:02.0
Display: x11 server: X.org v: 1.21.1.7 with: Xwayland v: 22.1.9 driver: X: loaded: modesetting
unloaded: fbdev,vesa dri: crocus gpu: i915 tty: 135x43 resolution: 1920x1080
API: OpenGL Message: GL data unavailable in console for root.
*-display
description: VGA compatible controller
product: 2nd Generation Core Processor Family Integrated Graphics Controller
vendor: Intel Corporation
physical id: 2
bus info: pci@0000:00:02.0
logical name: /dev/fb0
version: 09
width: 64 bits
clock: 33MHz
capabilities: msi pm vga_controller bus_master cap_list rom fb
configuration: depth=32 driver=i915 latency=0 resolution=1920,3240
resources: irq:28 memory:fe000000-fe3fffff memory:c0000000-cfffffff ioport:f000(size=64) memory:c0000-dffff
EDIT: the command execute with user show im using HD2000 :o
But working well anyway
--USER
Graphics:
Device-1: Intel 2nd Generation Core Processor Family Integrated Graphics
vendor: Elite Systems driver: i915 v: kernel arch: Gen-6 bus-ID: 00:02.0
Display: x11 server: X.Org v: 1.21.1.7 with: Xwayland v: 22.1.9 driver: X:
loaded: modesetting unloaded: fbdev,vesa dri: crocus gpu: i915
resolution: 1920x1080~60Hz
API: OpenGL v: 3.3 Mesa 22.3.6 renderer: Mesa Intel HD Graphics 2000 (SNB
GT1) direct-render: Yes
--ROOT
Graphics:
Device-1: Intel 2nd Generation Core Processor Family Integrated Graphics vendor: Elite Systems
driver: i915 v: kernel arch: Gen-6 bus-ID: 00:02.0
Display: x11 server: X.org v: 1.21.1.7 with: Xwayland v: 22.1.9 driver: X: loaded: modesetting
unloaded: fbdev,vesa dri: crocus gpu: i915 tty: 135x43 resolution: 1920x1080
API: OpenGL Message: GL data unavailable in console for root.
*-display
description: VGA compatible controller
product: 2nd Generation Core Processor Family Integrated Graphics Controller
vendor: Intel Corporation
physical id: 2
bus info: pci@0000:00:02.0
logical name: /dev/fb0
version: 09
width: 64 bits
clock: 33MHz
capabilities: msi pm vga_controller bus_master cap_list rom fb
configuration: depth=32 driver=i915 latency=0 resolution=1920,3240
resources: irq:28 memory:fe000000-fe3fffff memory:c0000000-cfffffff ioport:f000(size=64) memory:c0000-dffff
I used to play on n even crustier potato - do yourself a favor and put the poor thing in the dirt/give to a n00b! :p
I was so desperate to get more juice out of that blasted chipset I even tried overclocking the poor thing which turned out not being worth the squeeze; marginal performance increase but the instability made me backtrack pretty fast... wouldve been better served by running a more barebones system like arch btw but yeah.
All that to say I don't envy you, yeesh!
I was so desperate to get more juice out of that blasted chipset I even tried overclocking the poor thing which turned out not being worth the squeeze; marginal performance increase but the instability made me backtrack pretty fast... wouldve been better served by running a more barebones system like arch btw but yeah.
All that to say I don't envy you, yeesh!
Fett's example is basically sitting right at the bottom edge of what we're moving towards. So, it should still work. But, if we add a lot more asteroids and things, scenes of higher density, it might get slow.
I have a similar era laptop I use mainly for travel, a 2011 Samsung NP900X3A, that uses a dual-core Sandy Bridge CPU, the Intel Core i5-2537M, with the HD3000 GPU. This is an extreme low-power variant, TDP of only 17W, something like half the power of a "normal" Sandy Bridge of the same generation, so it's probably the slowest thing anyone's likely to try using that retains "technical compatibility". That will be my test-case for "minimum operational functionality". But, while I'll try to make the best config I can for it, I don't guarantee it will perform well.
If I were looking to inexpensively upgrade (especially running Linux), I would target a cheaper AMD device (laptop or miniPC). Ideally something with a Radeon 680M or (better) 780M, but even the inferior Vega 8 should be usable for some time.
The Linux drivers for AMD are now basically the best GPU drivers on earth (made by Valve, of all people), and arguably the best Vulkan drivers as well.
As I mentioned earlier, my basic test-case for the new renderer is the AMD Radeon 680M. Of course, we'll also target some graphical cases for big discrete GPUs, but the AMD should look pretty good, and work pretty well.
I have a similar era laptop I use mainly for travel, a 2011 Samsung NP900X3A, that uses a dual-core Sandy Bridge CPU, the Intel Core i5-2537M, with the HD3000 GPU. This is an extreme low-power variant, TDP of only 17W, something like half the power of a "normal" Sandy Bridge of the same generation, so it's probably the slowest thing anyone's likely to try using that retains "technical compatibility". That will be my test-case for "minimum operational functionality". But, while I'll try to make the best config I can for it, I don't guarantee it will perform well.
If I were looking to inexpensively upgrade (especially running Linux), I would target a cheaper AMD device (laptop or miniPC). Ideally something with a Radeon 680M or (better) 780M, but even the inferior Vega 8 should be usable for some time.
The Linux drivers for AMD are now basically the best GPU drivers on earth (made by Valve, of all people), and arguably the best Vulkan drivers as well.
As I mentioned earlier, my basic test-case for the new renderer is the AMD Radeon 680M. Of course, we'll also target some graphical cases for big discrete GPUs, but the AMD should look pretty good, and work pretty well.
"Fett's example is basically sitting right at the bottom edge of what we're moving towards. So, it should still work. But, if we add a lot more asteroids and things, scenes of higher density, it might get slow."
This old hardware (i3 2100 1st. or 2nd generation) and Integrated Graphics is doing things well for now with VO 1.8.775 client at Latos F-12 and in the Unknow/Inferno/System33 and in PvP and in CSR, but ya, i know im reaching limits for VO (my only game) and need an upgrade in the hardware and broadband
"If I were looking to inexpensively upgrade (especially running Linux), I would target a cheaper AMD device (laptop or miniPC). Ideally something with a Radeon 680M or (better) 780M,"
Thanks for recommendations, i was looking for miniPC's some time ago
cy
This old hardware (i3 2100 1st. or 2nd generation) and Integrated Graphics is doing things well for now with VO 1.8.775 client at Latos F-12 and in the Unknow/Inferno/System33 and in PvP and in CSR, but ya, i know im reaching limits for VO (my only game) and need an upgrade in the hardware and broadband
"If I were looking to inexpensively upgrade (especially running Linux), I would target a cheaper AMD device (laptop or miniPC). Ideally something with a Radeon 680M or (better) 780M,"
Thanks for recommendations, i was looking for miniPC's some time ago
cy