Interesting that they are aiming to such high-spec systems. I guess they need the performance to run existing games at adequate speed. But trying to go to retail with a $1000 box (my guesstimate for i5-4570/gtx760 setup) will not be easy.
I wonder if they might bundle them with most Valve steam titles, so that people feel that they would get a lot for their money. (on the other hand a lot of the people excited for the SteamMachine probably owns a lot of Valve titles already)
I'm guessing by the time their device get released next iteration of hardware will bring similar performance in the lower price bracket although I've stopped following GPU/CPU price/performance changes a few years back. They are probably just trying to see how the current hardware performs and how it will fit people use cases (that's why they have pretty much every price range in there)
This is a prototype and I think it is reasonable to have max spec. Because this is not a a closed console the consumer version will have many options on hardware and price( maybe a really cheap 200$ version if you only need it for simple games or video ). Or you can buy your own computer and run SteamOS on it.
$1k gaming PC is more reasonable proposition when it's your only (or primary) computer. But spending $1k on a secondary livingroom pc is asking for bit more.
If steamos is what I think it is - ubuntu distro I could use it as a main pc.
And depending on the living room - it may well be worthed. Having multi-TV screen gaming setup with audiophile sound can be very lucrative niche market.
i wonder if you could stream from the Steambox to your office room instead of doing the opposite. id definitely need lots of power for games in the living room but for the office work if it boils down to browsing you could just stream that from your living room box.
Sadly linux is way behind remote desktop from what I have seen :( I have worked from RDP with no problems on low latency broadband, but I am not sure such smooth experience is possible in linux. Lets hope that Wayland can improve it.
Be I could definitely see the fat headless home pc streaming to the devices needing it in the next few years - we could call it home mainframe.
They wouldn't have needed this expensive hardware if they went with the same or similar AMD hardware as in PS4/X1 and take advantage of their Mantle API, which basically makes the system as lean and efficient as a console. Not to mention it would be trivial to port the same highly-optimized console games to the Steam Machines. Carmack thinks Mantle could help Steam Machines with AAA titles a lot, too:
amd hardware in tge xbox one and Ps4 is very low in perfomances. The SteamMachines are about bringing PC power to the living room, and id expect such hardware to be vastly above the console tech if they even hope to make any difference.
Valve will get a wholesale price, and may not even try to make profit off it, much like Google and amazon. They make money off the platform anyway, so charging more than cost on something that gets people on the platform is silly.
wholesale price depends a lot on volumes, though. its very hard to know at this stage what to expect in trms of market sales. launching new hardware without prior experience is very. very tough. But as long as they are willing to learn, its doable.
Valve won't be getting any significant volume discount, because they won't be making any significant amounts of steam machines. The retail units will be made by several OEMs and in several configurations, so it's likely that no single model will be selling in massive volumes and be eligible for massive discounts.
The situation is actually quite similar to the Android world. Google makes the platform and others make the phones, Valve also provides a platform and lets others make the steam machines.
I'd agree. Also cooling Titan+i7 in a tiny box might be a challenge. But on the other hand Anandtech measured whole system consumption for Titan being 430 watts including a 130+ watt CPU. Drop 50 watts off the CPU and it doesn't sound so bad for a 450 watt power-supply.
edit: Tech Report says 323 watts for whole system in their Titan review, so it's certainly feasible.
Would mass-produced closed loop be a viable solution here? I'd imagine the engineering costs incurred ahead of time would add to quite a bit, but I'm unqualified to make any interesting judgment on any final balance of costs.
3 different graphics cards, 3 different processors, 300 total prototypes being shipped. Is around 33 systems at each specification level going to be enough to properly evaluate the different design levels?
Remember that they probably have a number of audiences they are considering targeting. Off the top of my head, there is at least three core groups. The high media consumption family who all game and often game together is currently in vogue for console marketing. The 'hardcore' gamer crowd who play high demand FPSes and MMOs, who used to be the traditional market segment for Steam. Then there's the casual gamer who plays lighter games, but may well play a lot more games than the 'hardcore' gamers. The last one I can think of would be the family which buys consoles for the kids, and the parents have yet to be converted and still need to be properly sold on all the interactive TV features. I doubt that only 8 samples at a given level would give any meaningful data once statistical variation was factored in.
The only conclusion from that which makes any sense is that Valve are going after a single demographic at the start, most likely heavy gamers. But I am concerned that this will both cause key issues to be overlooked, and that this will make expanding outwards into other segments very hard. Although they will have some big name assistance with getting into a wider market, just look at Apple versus Android. Despite the leaps and bounds android has come on, iPhones and iPads are still viewed as less geeky and easier to use.
On the other hand, however, perhaps Valve are intending just that result. After all, the controller will often need custom mappings to be selected, so perhaps that is a good market for them. If I were a living room electronics manufacturer, however, I'd really want ease of use to be at the core of my products. I wouldn't want to integrate with my premium TVs something that could be called "geeky".
Shameless plug but I was writing about this a few days ago on the very same topic and questions: http://pandoralive.info/?p=2004
I also think they will be going for the heavy gamer at first. Because that is mostly their current market already. As they expand they will have several issues to break into other segments, namely regarding brand recognition (who will know what a SteamMachine is when it's built but 10 different manufacturers) and what the differences in positioning are ? It reminds me a lot of the 3DO situation 20 years ago when you had a single console standard/OS but about 4-5 manufacturers and there was a real lack of strategy as a platform. And pricing was, at that time, a big issue as well for the launch window.
Hardware upgrades on the way are going to make this a real headache down the road, but I'm sure they are already thinking on how to address this.
I feel like Microsoft approached the console market this way, with a big, ugly box that only a hard core gamer could love. There are differences of course but generally, I think it's smart they're focusing on a single demographic. You can't focus on multiple demographics simultaneously since that's not really focus. Hopefully this will allow them to iterate quickly.
28 comments
[ 0.90 ms ] story [ 77.2 ms ] threadA niche, but a decently-sized one.
And depending on the living room - it may well be worthed. Having multi-TV screen gaming setup with audiophile sound can be very lucrative niche market.
Be I could definitely see the fat headless home pc streaming to the devices needing it in the next few years - we could call it home mainframe.
http://www.tomshardware.com/news/amd-mantle-api-xbox-one-pla...
The situation is actually quite similar to the Android world. Google makes the platform and others make the phones, Valve also provides a platform and lets others make the steam machines.
And I think that i3/660 is fairly well balanced entry level futureproof for 2 years 1080p machine.
edit: Tech Report says 323 watts for whole system in their Titan review, so it's certainly feasible.
Remember that they probably have a number of audiences they are considering targeting. Off the top of my head, there is at least three core groups. The high media consumption family who all game and often game together is currently in vogue for console marketing. The 'hardcore' gamer crowd who play high demand FPSes and MMOs, who used to be the traditional market segment for Steam. Then there's the casual gamer who plays lighter games, but may well play a lot more games than the 'hardcore' gamers. The last one I can think of would be the family which buys consoles for the kids, and the parents have yet to be converted and still need to be properly sold on all the interactive TV features. I doubt that only 8 samples at a given level would give any meaningful data once statistical variation was factored in.
The only conclusion from that which makes any sense is that Valve are going after a single demographic at the start, most likely heavy gamers. But I am concerned that this will both cause key issues to be overlooked, and that this will make expanding outwards into other segments very hard. Although they will have some big name assistance with getting into a wider market, just look at Apple versus Android. Despite the leaps and bounds android has come on, iPhones and iPads are still viewed as less geeky and easier to use.
On the other hand, however, perhaps Valve are intending just that result. After all, the controller will often need custom mappings to be selected, so perhaps that is a good market for them. If I were a living room electronics manufacturer, however, I'd really want ease of use to be at the core of my products. I wouldn't want to integrate with my premium TVs something that could be called "geeky".
I also think they will be going for the heavy gamer at first. Because that is mostly their current market already. As they expand they will have several issues to break into other segments, namely regarding brand recognition (who will know what a SteamMachine is when it's built but 10 different manufacturers) and what the differences in positioning are ? It reminds me a lot of the 3DO situation 20 years ago when you had a single console standard/OS but about 4-5 manufacturers and there was a real lack of strategy as a platform. And pricing was, at that time, a big issue as well for the launch window.
Hardware upgrades on the way are going to make this a real headache down the road, but I'm sure they are already thinking on how to address this.
We'll see what their plan are very soon.