After using Shadow for a month do you wish....

I have already had an experience with the Ghost and for me it doesn’t really add on much aside from having something with native support for Shadow. I really like the little thing, don’t get me wrong. I would just rather move up in tiers. Even if you’ve got the Ghost, you’ve still got to pay the fees, so instead of wasting that hundred on the device, I’d rather just put it toward a new tier when it releases.

Shadow Boost is more than enough for me and i use it in all my devices from my phone to my PC, having a high tier isn’t that much needed rn, and since i have lot of things that can run Shadow i don’t need the Ghost.

I could of ordered a ghost but running it on a HP Wave to get around the issue with mic over bluetooth - loving the idea of fanless!

What I struggle with is how the Shadow guys can commercially run the service - Titan RTX for £50 a month - its like credit with a negative interest rate - they pay us to take it!

Even visualised Nvidia Grid GPUs don’t work out commercially viable to offer at this level - I thought there was some cross subsidy but it feels like thats not the case.

Anyway, its made a gamer out of me again but loving the idea of going big on the Titan RTX….although with my boost I can play Starcraft 2, COD WWII etc on Ultra 1080p 60fps so thats all good but want to max out other games!

Titan RTX is probably overkill but ironically you could virtualise 2 machines on it and pass them both 2080 Super cards! 

It feels as much like the future as the Tesla X does to me! I own an IT company and its pushing me into virtualisation more and more! 

I’m on Boost, and so far it’s doing everything I need to across many devices, so I’m not in a hurry to upgrade. Where did the “ 0.2ms latency” come from? I didn’t see that anywhere, and lower latency would be worthwhile indeed!

RE: “ What I struggle with is how the Shadow guys can commercially run the service - Titan RTX for £50 a month - its like credit with a negative interest rate - they pay us to take it! “

Technically, you’re not really paying for just that. These are all racked in a datacenter right? So they have a large array of storage, CPU’s/Memory, and likely a separate array of GPU’s. These GPU’s are all interconnected, and can service X number of users at a time. Soooo - they average the number of users at any given time, and they throttle that with how many subscribers they can have active. They can throttle traffic more by limiting usage, such as they have with COVID-19. Then they charge monthly for it and increase capacity as usage increases, or increase users as usage decreases.

Want another array of users? Collect enough pre-orders for initial investment in hardware, purchase, implement, activate. Rinse and repeat. Unfortunately, it seems demand has exceeded capacity, hence the long pre-order time to 2021.

It’s like insurance - you’re paying into a pool that a lot of people contribute to, but not everyone uses 100% of the time. With only 80% utilization on average (or whatever their capacity looks like), they’re making a profit - there’s a LOT of overhead and I’m super-over-simplifying it, but it makes sense.

What I wonder though is why this can’t be done ala AWS/Azure/cloud-infrastructure style with rapid response ramp rate based on demand usage, then just charge accordingly - unlimited bandwidth, unlimited users, with essentially a “service fee” for configuration, storage, optimization, updates, prioritization, capacity, etc. etc. etc. (all of which are significant value adds).