• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Valve really got unlucky with Steam Machine timing, right?

One thing to consider is that it only uses 16GB of DDR5. Which is very expensive, but in a small amount that won't be as bad as with other pre-built.
And because it doesn't have all the bloatware that Windows 11 brings, it might not need all that much system memory.
The GPU uses GDDR6, which is a lot less expensive than GDDR7.

The Steam Machine might be unlucky in it's timing, but it might be the best bang for the buck machine in 2026.
 
It depends on the point of view.

The fact that this generation is being extended due to this chaos could be beneficial from the perspective of the specs they chose; the machine won't become "outdated" quickly if the new generation takes a while to come out.
 
I wish they would hurry up and give us a price point/release date so I can decide to go Deck or Box.
I always go Box. Not sure how you swing, though.

not that there's anything wrong with that seinfeld GIF by myLAB Box
 
if you care about all this nerd shit, the Steam Machine is not for you
what this guy said

but unfortunately, I don't think the Steam Machine is a good buy for anybody, really (if the $999 price is true). desperately needs RDNA4

If you want an easy, console-like experience you'll get much better performance from a PS5 Pro for 30% cheaper

if you want a PC the specs are not good for $1000+. just build your own and put bazzite/steamOS on it for 'best of both worlds'
 
Last edited:
They wanted to make a cheap machine by using available components they didn't have to RD. The main problem is the choice of ram.

But still considering the amount of cash they have, they really lack ambition. You would think they would be confident of selling 10 millions of machines LTD.
 
I never understood the hoopla around the SM. The controller is the only exciting thing to me about it. A new Steam Deck would make my heart pump a lot faster.
 
I never understood the hoopla around the SM. The controller is the only exciting thing to me about it. A new Steam Deck would make my heart pump a lot faster.

I don't personally want it, but I know some that do. I don't know how large the niche is, but there's enough gamers I've come across that want to enjoy PC gaming, but without the hassle of building one or dealing with drivers.

The idea of an entry point that's software managed by Valve, even extra incentive to not deal with Microsoft patching, but still getting the perks of PC gaming (like mods, no online subscription to play the game you already paid for, etc.) is enticing to some.
 
Not really. They just have to sell it cheaper compared to whatever desktop with overpriced RAM avaliable today.
 
It's weird because the Deck was made at the perfect time with the perfect AMD SoC.

Wereas here it looks like old AMD leftover from 2+ years ago. It should've used current zen+RDNA, or they should've waited till that was available.
 
I don't personally want it, but I know some that do. I don't know how large the niche is, but there's enough gamers I've come across that want to enjoy PC gaming, but without the hassle of building one or dealing with drivers.

The idea of an entry point that's software managed by Valve, even extra incentive to not deal with Microsoft patching, but still getting the perks of PC gaming (like mods, no online subscription to play the game you already paid for, etc.) is enticing to some.
41UbU06jYaImV6NJ.jpg


Not dealing with W11 and just being able to legally play a massive amount of PC games without much hassle actually sounds like a niche which hasn't been properly filled. Of course not something I'm personally interested in, but I can see relaxed non-console gamers smoking a joint and starting their SMs and enjoying themselves.
 
As far as I'm concerned, the steam cube is a poor value proposition as a pc and as a console. It's really stuck in no mans land.

It has the following problems:
  • It's very weak as a pc and has specs not built for longevity
  • It has no real upgrade path to improve performance unlike normal pc
  • If we look at it as a console, the rumored prices don't offer the same cost efficiency as a console
  • Very few developers if any will optimize directly for the steam cube losing the optimization efficiencies that enable console like longevity
  • It's weaker than a ps5 that came out 5+ years ago.
  • It has no support for FSR4 and we're currently not sure if it'll support XESS.
Realistically, this machine cannot run any of the major upcoming AAA games with any semblance of smooth performance. GTA 6, 007 First Light, high end UE5 games, etc..

With these specs, I cannot even see the point of such a device. If you're putting out a valve machine, it should be future looking. Instead, valve chose cheap hardware that's already expired before it releases.

It's essentially using the Nintendo hardware philosophy except the problem is that Nintendo has exclusives that causes consumers to purchase the machine. Valve has no such exclusives.
 
Last edited:
No shit. RAM-pocalypse and all that.

But IMO its not just the AI shitfest. Its the choice of hardware as well. If they aimed just 6 months later, they could be using the new Strix Halo SKU instead of garbage bin laptop parts from 2023.

Strix Halo? $2000 Steam Machine you nuts? Actually no. AMD saw pressures from Panther Lake coming, so they lined up cheaper Strix Halos for 2026: AI Max 388 (8x CPU, 40 CU GPU) and 392 (12x CPU, 40 CU GPU). Of the two, 388 would have been perfect for Steam Macine: Single CPU chiplet plus the I/O-GPU chiplet making it much cheaper than 3 chiplet AI Max+ 395.

Cost wise, think about these factors:

- Die size for 7540U (178mm²) and RX 7600M (204mm²) is bigger combined than AI Max 388 (379mm²).

- Dual chip and dual RAM type motherboard complexity vs single SoC and unified RAM pool (Valve engineers talked extensively about mobo layer complexity to accommodate GDDR6 on Gamers Nexus interview).

- Cooling needs are less costly (140W combined vs 120W). The cooling rig on Steam Machine is crazy with giant heat sink and crazy heat pipes for 2 separate chips. The overall box size could have been smaller to boot.

- PSU could shrink from 300W to 230W (GMKtec AI MAX 395 mini PC).

Even if all those factors even out with newer and more expensive SoC vs goodwill AMD laptop parts, the performance gains you get for similar price point is crazy:

- 7540U gets 2250 single thread and 8000 multi thread on Geekbench 6 while 388 gets 2500 and 14,000.

- RX 7600M gets 11,000 on Passmark GPU test while 8060S in 388 gets 18,000.

- 16GB DDR5 5400 gets you about 80ish GB/s, 8 GB of GDDR6 for 7600M gets you around 256GB/s, while 388 gets you 32GB of 256GB/s LPDDR5X.

- 24GB of RAM for Steam Machine makes Valve juggle 2 SO-DIMM slots and 2 GDDR6 chips soldered on mobo while you only have 4 soldered LPDDR5X 8000 chips on 388 mobo for 32GB.

If I was Valve, I would be kicking myself in the ass for rushing the Steam Machine by mere 6 months...

Your analysis isn't wrong, but from a vibes standpoint, the kind of person the Steam Machine is aimed at really doesn't know or doesn't care about that kind of stuff.
 
It's weird because the Deck was made at the perfect time with the perfect AMD SoC.

Wereas here it looks like old AMD leftover from 2+ years ago. It should've used current zen+RDNA, or they should've waited till that was available.
That would be 2nd half of 2027 for RDNA 5. We are stuck with RDNA 3.5 until then.
 
Your analysis isn't wrong, but from a vibes standpoint, the kind of person the Steam Machine is aimed at really doesn't know or doesn't care about that kind of stuff.
They would certainly know how it barely keeps up with PS5. And they would certainly find out from nerds on YouTube and such how SM quite doesn't hack it for the price.
 
Valve isn't stupid. This is a multi-million, even multi-billion-dollar company. They will have thought this through before bringing a console to market, securing components in time at competitive prices rather than buying overpriced parts five minutes before release.

Awkward Chris GIF by ABC Network
Having money doesn't mean you are smart... just ask Microsoft.
 
They would certainly know how it barely keeps up with PS5. And they would certainly find out from nerds on YouTube and such how SM quite doesn't hack it for the price.

To some extent yes, but the value proposition of a Steam Machine is maintained, despite price increases, because the price increases don't just affect the Steam Machine, but everything else the Steam Machine competes against too.

Of course inflation will price out customers who can't afford any luxury purchase above $900, but for the customer base who can still afford it, it's still an attractive option. Not to mention the potential non-gamer sales if the SM turns out to be a relatively decent general use PC, in a cute package, for casual users.
 
Could also be A BIG chance if they price it right. Many people would potentially grab a Machine rather than a way too expensive, even mid-range, gaming-pc
If that it gets priced right, scalpers will scoop it up in droves, gut it of it's ram and try to make profit on only that.
 
Depends on when they bought the components. If they already had their RAM stocked up before the sudden price increase they will be putting themselves on the back for not waiting 6 months.
I read somewhere that Valve basically bought up "AMD's trash" to make these machines, they got a bargain in the process.

If I remember correctly there was speculation that Valve was ramping up production on the machines about 8 months ago. That would lead me to believe there is a warehouse full of these machines somewhere ready to go. They might be able to take a hit on the first wave of machines, but it won't be pretty for the second round most likely. If the price is right get one first wave, although with how they made people wait for the deck in waves, (I believe some pre-orders were 6 months out) this might be impossible.

I wonder if they limit the machines to first wave. Wait a while and open up pre-orders for a second round at a higher price point.

I have a feeling if I wanted one I will probably be shut out of first wave pre-order immediately if they strike the right price point. Assuming of course this does not flop altogether. Which history has proven can even happen to Valve.
 
Last edited:
If that it gets priced right, scalpers will scoop it up in droves, gut it of it's ram and try to make profit on only that.

I think that one of the overlooked aspect of this crazy inflation in prices is that regardless of the value proposition of the Steam Machine as a product, or any future all-in-one device, the mere components of the thing itself will have unusually high inherent and resale value.

Given a certain breakpoint of component inflation, even a piece of shit product is still worth something due to what it's made out of.
 
I dont care about the steam machine just gimme the goddamn controller. I hope atleast thats coming in Q1
Unlikely they will launch them in different time frames. The entire marketing hinges of the eco-system of the Frame, Controller, Steam Deck. But yes, if nothing else I would gladly grab 3 Steam Controllers off the bat.
 
Extremely unlucky with the timing because 8Gb of VRAM was already pathetic a few years ago let alone now in 2026 which is a lot worse and a big insult to consumers.

The fact that it's better than a lot of Steam users current pc also won't work because all those folks would want to get a actual decent upgrade over their old pc that especially isn't limited by 8gb of vram. I was one of them a few months ago.
 
I feel like maybe the biggest corporations in the world are willing to burn the forest down before letting the fatman bring cheap open-source devices to the masses.
 
I don't know how people come to such braindead conclusions. Valve has the most user hardware data of any company on this planet when it comes to what hardware their users have. If you think these specs were chosen out of the blue then you are either a) ignorant, b) shitposting, c) have no concept of hardware development, or d) all of the above.

People better jump on this deal as soon as they are made available for pre-order because the simple fact is the VAST majority of Valve users cannot hope to afford to upgrade their low to mid tier gaming systems. If you think ram is bad now wait until we hit Q3 of this year. 200% increase over Q3/Q4 25 and Q1 26 is nothing compared to reality that is heading our way this holiday season.

People are already grumbling about "deaths" in the pc hardware provider space before Q4 of this year.
 
I don't know how people come to such braindead conclusions. Valve has the most user hardware data of any company on this planet when it comes to what hardware their users have. If you think these specs were chosen out of the blue then you are either a) ignorant, b) shitposting, c) have no concept of hardware development, or d) all of the above.

People better jump on this deal as soon as they are made available for pre-order because the simple fact is the VAST majority of Valve users cannot hope to afford to upgrade their low to mid tier gaming systems. If you think ram is bad now wait until we hit Q3 of this year. 200% increase over Q3/Q4 25 and Q1 26 is nothing compared to reality that is heading our way this holiday season.

People are already grumbling about "deaths" in the pc hardware provider space before Q4 of this year.
Valve failed hard in 2015. Granted they did not make the machines and most vendors dropped out. Valve can make mistakes and overestimate need/want just like any other company.
 
Last edited:
As far as I'm concerned, the steam cube is a poor value proposition as a pc and as a console. It's really stuck in no mans land.

It has the following problems:
  • It's very weak as a pc and has specs not built for longevity
  • It has no real upgrade path to improve performance unlike normal pc
  • If we look at it as a console, the rumored prices don't offer the same cost efficiency as a console
  • Very few developers if any will optimize directly for the steam cube losing the optimization efficiencies that enable console like longevity
  • It's weaker than a ps5 that came out 5+ years ago.
  • It has no support for FSR4 and we're currently not sure if it'll support XESS.
Realistically, this machine cannot run any of the major upcoming AAA games with any semblance of smooth performance. GTA 6, 007 First Light, high end UE5 games, etc..

With these specs, I cannot even see the point of such a device. If you're putting out a valve machine, it should be future looking. Instead, valve chose cheap hardware that's already expired before it releases.

It's essentially using the Nintendo hardware philosophy except the problem is that Nintendo has exclusives that causes consumers to purchase the machine. Valve has no such exclusives.
This post summarizes my thoughts as well. I have absolutely no idea who this product is even for.

The whole argument that "Well it's faster than XX% of machines on Steam" doesn't make sense if you think about it for a second. A LOT of machines on the hardware survey are internet cafes, Steam installed on a PC whose primary purpose isn't gaming, or being used by someone who's playing F2P or multiplayer slop games. Are you telling me those people are going to rush out and pay $700 for something that can't play the latest/greatest games or won't run a lot of the multiplayer games they want to play? Bullshit.
 
Last edited:
Yeah, its a pretty shit situation, I was planning on getting one for my Dad when it came out since he's jumping ship from Xbox, but unless they subsidize it to bring it under 600 bucks I just don't see it happening. I will pick up the controller for myself though.
 
I think their part choices will likely save more than we expect, but RAM will hurt them for sure.

I'm hoping it's under $800 for it to have a chance beyond being a small niche. Some people will pay extra to have a device designed better for the TV than other PCs out of the box, and will get better support given the Steam Deck's track record.

I'm not the target audience though, just want that controller baby.
 


I don't know if K KeplerL2 spoke with a pinch of truth or rumours though
Makes sense. These are low volume machines, they're not going to get any decent prices for the build, and it will be DOA if pricing is wrong. It makes sense to go back to the drawing board and wait out the market. It's not like they invested into a custom SOC or anything.
 
If that it gets priced right, scalpers will scoop it up in droves, gut it of it's ram and try to make profit on only that.
Deck wasn't scalped because Valve did it right. You need an account in a good state and only one unit per account. I could get mine without a problem back then.
 
Valve failed hard in 2015. Granted they did not make the machines and most vendors dropped out. Valve can make mistakes and overestimate need/want just like any other company.
5070ti cards upwards of $900 USD now (if not more at nation wide retailers like best buy). I find it difficult to connect the failures (in your words) from 2015 to what Valve is doing now especially given the catastrophe facing the PC gaming industry currently.

If anything Valve is positioning themselves to be the only game in town when it comes to providing an affordable prebuilt pc rig that people (in this case people does NOT refer to all us neck beards with multiple 5090 cards) can pick up, connect to their living room TV, and start playing their steam library.

I am the first person in line with flaming pitchforks when it comes to Gaben...I think he is a super villain...but compared to the likes of samsung, nvidia, microsoft, and skyhinx...Value might be the only bulwark against a complete industry shift to "cloud first" gaming.
 
Using genetic AMD parts isn't going to work, they need to have their own custom SoC and chipset design. Otherwise it will end up like the original Xbox where the whole thing will be a money sink that cannot be redesigned.
 
They should easily be able to keep their machine affordable from the 30% of everything they rake in.
Thats not how capitalism works, if they launch gabecube they want more profits(not revenue, profits) so they will make sure price of that hardware is enough to cover all ram shortages, its not 2006 sony that was losing 200$ on each 599$ ps3.
Rememeber Gabe just bought his new yacht in november, that stuff isnt cheap :D
 
5070ti cards upwards of $900 USD now (if not more at nation wide retailers like best buy). I find it difficult to connect the failures (in your words) from 2015 to what Valve is doing now especially given the catastrophe facing the PC gaming industry currently.

If anything Valve is positioning themselves to be the only game in town when it comes to providing an affordable prebuilt pc rig that people (in this case people does NOT refer to all us neck beards with multiple 5090 cards) can pick up, connect to their living room TV, and start playing their steam library.

I am the first person in line with flaming pitchforks when it comes to Gaben...I think he is a super villain...but compared to the likes of samsung, nvidia, microsoft, and skyhinx...Value might be the only bulwark against a complete industry shift to "cloud first" gaming.
Not saying they will fail necessarily, just that they can. I find humor in how I love Valve and taking the more negative approach and you have negative views on Gabe and think Valve is positioned to succeed. The irony is making my day!!! Pretty Funny!!!. Have a great Weekend!!!
Also, the fact that Valve had to fail to bring us Proton and this current day Steam Machine.
If Valve had actually succeeded the first time, we would all be playing games on native Linux I suppose.
 
Last edited:
Valve could practically lock in a win by making their Steam console the easiest place to start gaming. Bundle access to a curated, rotating library of titles with the device and make the value immediate: buy in, plug in, log in, and you already have great games waiting. No extra barrier to entry, like Sony and Nintendo locking multiplayer behind a paywall (valving out loud!). The more people who adopt the console, the more leverage Valve gains with developers looking for visibility on the platform.

Fuck shit up Valve!
5Kciaa9PLXd6FYDn.gif
 
Valve could practically lock in a win by making their Steam console the easiest place to start gaming. Bundle access to a curated, rotating library of titles with the device and make the value immediate: buy in, plug in, log in, and you already have great games waiting. No extra barrier to entry, like Sony and Nintendo locking multiplayer behind a paywall (valving out loud!). The more people who adopt the console, the more leverage Valve gains with developers looking for visibility on the platform.

Fuck shit up Valve!
5Kciaa9PLXd6FYDn.gif
The only problem with this is it would lessen the perceived value of the "Steam Sales" (Valve's trump card). The big selling point becomes the free games not the cheap games. I don't think Valve wants that. (I don't believe you are referring to a sub). There are tons of free games on Steam of course, but they are not particularly highlighted, and the majority don't set the world on fire.

Also, it really hasn't worked for Epic. People grab the games, probably never play them. Fortnite keeping them alive though.
 
Last edited:
Valve could practically lock in a win by making their Steam console the easiest place to start gaming. Bundle access to a curated, rotating library of titles with the device and make the value immediate: buy in, plug in, log in, and you already have great games waiting. No extra barrier to entry, like Sony and Nintendo locking multiplayer behind a paywall (valving out loud!). The more people who adopt the console, the more leverage Valve gains with developers looking for visibility on the platform.
Any entry level PC/notebook can do it too without having to subsidize crowd who will be on your platform anyway

A lot of people just have a strong wish that someone else sell them things cheap even when it has zero business sense
 
Any entry level PC/notebook can do it too without having to subsidize crowd who will be on your platform anyway

A lot of people just have a strong wish that someone else sell them things cheap even when it has zero business sense

It wouldn't replace Steam sales, it would feed them. I'm not talking about new releases, just older titles. The bundled library would be a discovery hook: people can try games they missed, wishlist the sequels, buy during sales or grab sequals at laucnh, DLC's for the older titles, and newer entries. yada yada yada.
 
It wouldn't replace Steam sales, it would feed them. I'm not talking about new releases, just older titles. The bundled library would be a discovery hook: people can try games they missed, wishlist the sequels, buy during sales or grab sequals at laucnh, DLC's for the older titles, and newer entries. yada yada yada.
Putting banner of sale on store main page with big font "-80%" is much cheaper
And your idea will just increase price of box from 1000$ to 1100$ as Valve has no reason to pay for you
 
Any entry level PC/notebook can do it too without having to subsidize crowd who will be on your platform anyway

A lot of people just have a strong wish that someone else sell them things cheap even when it has zero business sense

That's assuming the audience is the same. It isn't. A console-style Steam device with a built-in library isn't aimed at existing PC users, it's for people who don't want to build a pricey PC, shop for parts, and then immediately have to buy a game library on top of the hardware. The bundle isn't a subsidy for people already on Steam. It's an onboarding tool to pull in users who otherwise wouldn't enter the ecosystem at all. More users means more wishlists, more engagement, and more sales. It expands the market instead of discounting it.

And even current PC users might bite if the value is there. I'd consider a Steam box when one of my PCs needs upgrading, but right now, the hardware alone doesn't offer me anything I don't already have or couldn't get elsewhere.

You're seem to think this is just me wanting cheap stuff with no business logic. I see it differently, assuming a company should only focus on its current audience instead of expanding into new ones is what doesn't make business sense to me.
 
It wouldn't replace Steam sales, it would feed them. I'm not talking about new releases, just older titles. The bundled library would be a discovery hook: people can try games they missed, wishlist the sequels, buy during sales or grab sequals at laucnh, DLC's for the older titles, and newer entries. yada yada yada.
It's sort of a Gamepass situation. I remember Phil touting that Gamepass subscribers bought more games than non-subscribers. Phil Please!!!
If you give people the low hanging fruit, they aren't going to see a need to spend more money if you place the free stuff in front of them. You are keeping them plenty busy with the free games. Why shell out for more stuff you won't ever get to play most likely?
 
Last edited:
Top Bottom