It's already failed multiple times. OnLive tried & failed. Nvidia's has been in beta for years. Sony has one.
Everyone in this area has tried this. Nobody has seen what could be described as "success", and the cost models so far have been ludicrous. Turns out renting Xeons and Radeon MI Instincts in a professionally staffed, maintained datacenter is way, way more expensive than a handful of consumer chips in a box in the living room with nobody on-call to monitor it.
The GPU here looks to be basically a slightly cut-down AMD MI25. That'd make a single GPU in this stradia cloud gaming service costs more than 10 xbox one x's. How do you make that price-competitive here?
I'm sure they got bulk/promotional pricing from AMD, plus they're very good at both running hardware with low overhead and packing it efficiently.
You can't really pack the hardware here since it's latency sensitive. It's straight dedicated resources to an array of VMs. Dedicated CPU cache, even, hence the odd 9.5MB L2+L3 number.
Bulk pricing only gets you so far here. You're still talking gear that's categorically way more expensive than similar performance consumer parts. Not to mention all the other costs in play - data center, power, IT staff, etc...
Making this price-competitive is a big problem
The other costs (power, people, etc.) are amortized over Google's array of services.
Last but not least, it would be very dumb of them not to run batch workloads on these machines when the gaming service is idle. I bet $1000 these puppies run under Borg.
Power doesn't really amortize, and neither does heat.
And capacity still had to increase for this. They didn't just find random GPUs under the table they forgot about, and now that they have a massive fleet of GPUs it's not suddenly going to start handling dremel queries.
This all still costs money. A shitload of it. Someone is going to pay that bill. More ads in YouTube won't really fund gaming sessions. So will this be ad breaks in the game? No way that's cost-effective for the resources used. Straight-subscription model? This seems most likely, but how much and how will you get people to pay for it?
I know from experience that Google is very cheap. You tell Urs you saved a million dollars and he'll ask you why you didn't save two. Or five.
If this takes off, the pricing of the service will pay for the hardware (assuming they did a reasonable job there of baking it in). Even if it doesn't, organic growth from other, much larger Google services can make use of the idle hardware.
For the record, I was involved in a couple of projects that required a lot of new hardware. One of them even ended up saving the company a lot of money in a very lucky, definitely unintended way.
This strikes me as rather amusing. Google was having such trouble getting their hands on enough GPUs that they decided to build custom hardware accelerators (TPUs) to fill the gaps.
I'm sure they'll find a use for these.
HBM2 memory is super expensive. Like, rumors are 16GB of HBM2 is $320 expensive. Toss in anything custom here and there's zero chance this is under $600/GPU.
Even in the hotly contested consumer market the 16gb HBM2 Radeon VII is $700. And that doesn't have any high speed interconnects to allow for sharing memory with CPU or multi-gpu.
Could they be using these GPUs for other purposes during idle times (AI model training, cloud GPU instances)?
There's also an old paper by Murray Stokely and co. about the fake market that was created to make the most use of all hardware planetwide.