Valve considering making console?

[liolio]

Looking at this, I'm not sure how further Valve can push the specs.
I can't see Valve use custom parts at least from the PC world (read as high end high performances parts) either, neither subsidizing the device.

The device starts @699$ and you got a core I3 and could get an Nv GT 545 for that price.
They could push more solid hardware now AMD has and NV soon will have 28nm parts.
They could also lower the price by using not taking profit on the thing, removing the optical drive, removing 4GB of RAM, using a cheaper HDD, etc. Using AMD CPUs

Still I can see the thing comes out neither impressive or cheap.

The only solution I see for Valve is to go with ARM CPUs and GPU from embedded world. Now (and I mean now Rogue is still not available implemented, neither ARM v8 CPUs, etc.) that would be something like a quad core based on A15 at high speed and high end PoverVR solution.
How that will fare against an AMD quad core APU + a hd7770?
On pure perfs it would be a loss.

Then there is price and power and it's most likely a win for the ARM based platform.
For some form of forward compatibility it's also a win, big money is in the embedded world it evolves fast.

The main conclusion of this and how accurate is the talk above is not that important, is that for Valve to do such a moved it imho implies moving to ARM environment, running windows8 (and how games BC will be handled on ARM version of windows?) and the most important part to me a lock down of the device on pretty much what are now minimal PC requirement, so a lock down on performances (most likely not for long, that's the whole point).

Either way they go for a really expansive system that would make the PS3 price at launch looks like the price of toys.

What do you guys think about it? I feel like no matter how they would be in tough spot, pressed in perfs by PC and price by upcoming consoles.

 

[Mobius1aic]

liolio,

Plenty of cost is absorbed by the X51's need to be upgradeable to a certain extent. Valve could make a non-upgradeable machine for cheaper by putting the GPU on the mobo a la smaller laptops with dedicated graphics. While some people might cry foul at the idea, it would be a relatively cheap, gaming centric, yet fully functional computer that would at least have a good 2 years of decent gaming life.

I would go with a semi X51 approach, eliminate everything unnecessary. The mobo would be mini ITX, two DDR3 slots, a single PCI-E x16 3.0, and no other expansion slots. I/O would be limited to the basics, like HDMI and VGA outputs for the IGP, perhaps 2x USBs 3.0s in the back, 2x USB 3.0s on the front, ethernet, audio and the WiFi antenna. I guess I'd have an optical audio out too. I'd eliminate the DVD drive though. Typical mobos tend to have way too many peripheral devices for basic computer users. If a person really needs more, they can get an external DVD drive or get a USB hub.

 

[RudeCurve]

Something like this?

http://www.amazon.com/Foxconn-NTA35...UTF8&coliid=I17E83UGQXABOW&colid=E85H2R9X4MNB

 

[liolio]

liolio,

Plenty of cost is absorbed by the X51's need to be upgradeable to a certain extent. Valve could make a non-upgradeable machine for cheaper by putting the GPU on the mobo a la smaller laptops with dedicated graphics. While some people might cry foul at the idea, it would be a relatively cheap, gaming centric, yet fully functional computer that would at least have a good 2 years of decent gaming life.

I would go with a semi X51 approach, eliminate everything unnecessary. The mobo would be mini ITX, two DDR3 slots, a single PCI-E x16 3.0, and no other expansion slots. I/O would be limited to the basics, like HDMI and VGA outputs for the IGP, perhaps 2x USBs 3.0s in the back, 2x USB 3.0s on the front, ethernet, audio and the WiFi antenna. I guess I'd have an optical audio out too. I'd eliminate the DVD drive though. Typical mobos tend to have way too many peripheral devices for basic computer users. If a person really needs more, they can get an external DVD drive or get a USB hub.

Just to say the mobo in the X51 is already a mini ATX I don't expect that many pci slots, memory slots. Basically it's what I said you remove the optical drive, use less ram, removing the margins.

 

[zoost]

Software hardware integration for:

Head mounted display + 3D + head tracking + kinect + halflife 3

 

[MfA]

Nope, that's not what wearable computing is about ... it's more about augmented reality than virtual reality.

 

[xedi]

If it boots into a locked down windows which only runs games and maybe stuff like Chrome apps then it can still have a known good configuration.

I don't think they could do that since Windows is proprietary so they would not be able to make their own locked down Windows version. The alternative would be to build a Steam Linux, maybe based on Ubuntu due to its popularity, which is specifically targeted to run a console like Steam experience. The problem with this approach is that most games on Steam are Windows only and Valve has currently no interest supporting regular Linux Distros so making their own Linux OS for their console is unlikely. On the other hand Linux on a Desktop is getting more and more popular and is already the leading platform on mobile phones (Android), so maybe Valve will change their policy in the future and support open platforms more.

 

[joker454]

On the other hand Linux on a Desktop is getting more and more popular and is already the leading platform on mobile phones (Android), so maybe Valve will change their policy in the future and support open platforms more.

Is linux really getting more popular on desktop? I've been hearing this said for years now, but when I look at the google analytics of my various websites which get thousands of hits per day, linux has pretty consistently been near 1% of my traffic for years, recently falling down to about 0.84%. Windows is about ~80% of my traffic, around 7.5% for Mac (which has slowly falling), and around 6% for iOS devices (has been steadily rising).

 

[MfA]

I don't think they could do that since Windows is proprietary so they would not be able to make their own locked down Windows version.

Idiot proofing is not the same as DRM ... it's just a question of booting into a sandbox UI at the start and blocking all the ways of exiting it, perfectly possible. A determined user could bypass that of course, since when you get down to it it's still windows. That they could say gain administrator privileges by booting into a different OS and editing system files is not a problem though. Most of the people who go that far to get away from the UI sandbox know the consequences are on their heads.

Personally I'd just dual boot and keep the idiot proof install as a backup myself.

 

[xedi]

Is linux really getting more popular on desktop? I've been hearing this said for years now, but when I look at the google analytics of my various websites which get thousands of hits per day, linux has pretty consistently been near 1% of my traffic for years, recently falling down to about 0.84%. Windows is about ~80% of my traffic, around 7.5% for Mac (which has slowly falling), and around 6% for iOS devices (has been steadily rising).

It varies a lot on how you do the statistics and who does them but from what I have seen the pattern is always that there is growth even though slowly and the market share is still very small. I am sorry I do not have any sources out of the top of my head, only ubuntu specific ones but not linux overall so yeah take it with a grain of salt but my memory is that all statistics agree that there is growth. I think most say that there is around 1-2% market share range but there are some who claim even more like 10%.

Edit: Remembered one source, yey http://www.omgubuntu.co.uk/2012/01/is-linux-marketshare-on-the-rise-it-seems-so/

Out of personal experience I found that many have been starting to dual boot linux (mostly Ubuntu) in the last 2-3 years even though the amount of people who actually use it almost exclusively is very rare. Then again I'm studying at a German college so I do not know whether "normal" people have even heard of it. During my study abroad in the US I only met one person who dual boots Ubuntu.

@MfA: Ok, now I have a clearer understanding of what you meant.

 

[xedi]

Thinking about it a bit more, if Valve would use Windows on their console, would be Microsoft happy about selling more Windows licenses or would they fear that their Xbox sales would go down? If so, could they hinder Valve using Windows on their hypothetical console?

 

[Silent_Buddha]

Edit: Remembered one source, yey http://www.omgubuntu.co.uk/2012/01/is-linux-marketshare-on-the-rise-it-seems-so/

Even if we go by the average that's listed there, in 10 years Linux would still have less than 5% marketshare. And that's incredibly optimistic to think that in 10 years Linux will have more than 2-3% marketshare at most.

I've been hearing for the past 10 years now how Linux is growning. Or "this" (any one of the previous 10 years) is going to be the year that Linux starts to rapidly grow. Yet its share of the market has been remarkably stable. Any growth or losses can easily be chalked up to the margin for error when doing statistics for overall computer marketshare.

In other words, for well over 10 years now, Linux growth (in %marketshare and not total install base) has basically been flat. And that really shouldn't come as a surprise to anyone.

A Windows based box would be the only really viable solution. Both Apple based (iOS or OSX) and Android based systems will be out of the question as Apple will wish to control any marketsplace on those devices while Google will want to do the same with regards to Android devices. Only Windows systems are basically open without the platform owner controlling who gets to sell things.

Although I suppose Valve building their own Android device may be possible, there's a whole host of problems with that. Convincing developers to port games to your platform. Convincing consumers that your Android device which likely doesn't double as a phone is more compelling than all the other android devices out there.

One possibility is Valve partnering with a boutique PC maker (Alienware or something) which can make standardised and branded hardware.

Or perhaps some kind of certification system so that an OEM PC can sport an Approved for Gaming by Valve for XXXX year logo or something. Similar to the Centrino logo or Windows 7 ready, or other such PC certifications that we've seen in the past.

But the problem with that is that PC developement is a constantly moving target. So a machine approved for year XXXX may not run games made in year XXXX+2 that well.

Regards,
SB

 

[Rangers]

Hmm, "megaton" just hit GAF:

http://www.theverge.com/2012/3/2/2840932/exclusive-valve-steam-box-gaming-console

According to sources, the company has been working on a hardware spec and associated software which would make up the backbone of a "Steam Box." The actual devices may be made by a variety of partners, and the software would be readily available to any company that wants to get in the game.

Adding fuel to that fire is a rumor that the Alienware X51 may have been designed with an early spec of the system in mind, and will be retroactively upgradable to the software.

Apparently meetings were held during CES to demo a hand-built version of the device to potential partners. We're told that the basic specs of the Steam Box include a Core i7 CPU, 8GB of RAM, and an NVIDIA GPU. The devices will be able to run any standard PC titles, and will also allow for rival gaming services (like EA's Origin) to be loaded up.

Part of the goal of establishing a baseline for hardware, we're told, is that it will give developers a clear lifecycle for their products, with changes possibly coming every three to four years. Additionally, there won't be a required devkit, and there will be no licensing fees to create software for the platform.

We're hearing that a wide variety of USB peripherals will be compatible with the boxes, though it will likely ship with a proprietary controller. It's possible that the controller will even allow for swappable components, meaning that it can be reconfigured depending on the type of game you're playing. Think that sounds odd? Well Valve filed a patent for such a device last year.

The most interesting piece of this puzzle may be related to that statement. According to sources, the Steam Box isn't intended to just clash with current gaming consoles. Rather, Valve wants to take Apple and its forthcoming new Apple TV products head-on. Newell has clear questions about Apple's strategy, telling the The Seattle Times "On the platform side, it's sort of ominous that the world seems to be moving away from open platforms," adding that "They build a shiny sparkling thing that attracts users and then they control people's access to those things."

Problem I see is the X51 is $1,000. Dont see this taking off, though the proprietary controller thing sounds cool.

Other problem I see is simple: piracy.

 

[Acert93]

Also, "The devices will be able to run any standard PC titles" = Windows (?) I am all for someone (should have been MS years ago) setting a basic "standard" "gaming" hardware profile that every 2-3 years gets updated, e.g. "Windows Gaming Rig 2010" "Windows Gaming Rig 2012" etc. It sounds like Valve is pushing that direction which could be a benefit to developers and the industry. That said the X51 is what I see wrong with so many PCs directed at gaming: too much put into the CPU and Memory and not enough into the GPU and the aesthetic concessions (namely size) and branding jack the price up way above the "value." Good idea in general but they really need to be aiming a lot lower on the price threshold to be competitive.

 

[liolio]

The most interesting part to me is about Apple. Basically with MS having a market place on win 8, console being already closed devices Valve must feel a bit cornered.
Still I can't see in which way they can fight back such giants, not too mention that in my view they are likely to face another ie Google which may use the some of the same policies as they are (multiple hardware vendors, etc.). I can't see Google stand still for long in face of an Apple offensive on that front.

 

[patsu]

Cool ! It would be a very interesting GDC to say the least. ^_^

Port the relevant stacks over. It doesn't seem to conflict with Sony's latest moves and rumors.

 

[MfA]

It would be grand incompetence if they go for a fixed platform for 3 years. It would be infinitely smarter to simply have a costly certification, this would keep the number of configurations down without removing hardware competition. Game developers have enough QA resources to test more platforms than one every 3 years even if near infinity is too much to ask.

Valve has a great service, but trying to force PC gaming into the console mould to this extent would be an evil act in my book. It removes so much impetus towards competition and innovation ... if they succeed at it and take most of the PC gaming pie I might as well get a console then, don't fucking do this Valve.

 

[Silent_Buddha]

Also, "The devices will be able to run any standard PC titles" = Windows (?) I am all for someone (should have been MS years ago) setting a basic "standard" "gaming" hardware profile that every 2-3 years gets updated, e.g. "Windows Gaming Rig 2010" "Windows Gaming Rig 2012" etc. It sounds like Valve is pushing that direction which could be a benefit to developers and the industry. That said the X51 is what I see wrong with so many PCs directed at gaming: too much put into the CPU and Memory and not enough into the GPU and the aesthetic concessions (namely size) and branding jack the price up way above the "value." Good idea in general but they really need to be aiming a lot lower on the price threshold to be competitive.

The Alienware is just an example of a machine that is rumored to have been built with the proposed hardware specs of the "Valve box."

Since we have no idea what the baseline is for the Valve Box certification, we don't know just how cheaply they could be made, or even how OEMs would decide to design the exterior.

It would be grand incompetence if they go for a fixed platform for 3 years. It would be infinitely smarter to simply have a costly certification, this would keep the number of configurations down without removing hardware competition. Game developers have enough QA resources to test more platforms than one every 3 years even if near infinity is too much to ask.

Valve has a great service, but trying to force PC gaming into the console mould to this extent would be an evil act in my book. It removes so much impetus towards competition and innovation ... if they succeed at it and take most of the PC gaming pie I might as well get a console then, don't fucking do this Valve.

I'm seeing this as basically a minimum spec for the "Valve Box." A cheap sort of certification.

In other words, how I view this is that Valve is attempting to set a minimum spec for a "Valve Box" which developers can then target as a minimum spec for their game. So, in theory, instead of targetting everything from the lowest spec'd computer to the highest spec'd computer there will be a baseline spec that is higher than what you can currently find on Steam.

I wouldn't be surprised if each baseline spec aligned with each new DX release. So the first baseline spec might require DX11 hardware. And even integrated Dx11 hardware is going to be a lot more capable and performant than say old Dx9 integrated hardare.

Depending on what demographic they want to chase with this "Valve Box" will likely determine just how low or high the baseline spec will be.

You see it as potentially putting an upper limit on what is implemented. I see it as raising the bar of the minimum spec that is targetted. If the idea can take off and consumer can be convinced to buy it, it can potentially raise the minimum bar of what developers target and hence increase the overall quality of PC titles.

Of course, with the majority of AAA PC developers focusing on Console first, we'll still ultimately be constrained by the console centric decision making.

Standards might help bring some of the developers back to PC first developement, but I doubt it. Until the piracy issue is addressed, I have a feeling there isn't much that will entice former PC first developers who are now on consoles to go back to PC first developement.

Regards,
SB

 

[MfA]

I don't care about upper limits ... this is about kneecapping competition in the GPU field.

Valve is too big to play favourites and not have a huge impact.

 

[patsu]

Might as well wait a few more days for GDC. We will have concrete info to discuss. ^_^

The devil is in the details.