Xbox Cloud compared to PS now

I agree. I was referring to a different quote. Thanks.

In any case, that’s not a strong claim. Providing additional lighting, physics, and motion effects is a far cry from claiming it’s revolutionize graphical prowess, or some such nonsense like you and some of the other ridiculous nay-sayers seem to cling to.



Of course it’s helping my argument. Dramatically so. Pivotally so. Azure is already equipped with GPU’s. Ever heard of Galactic Reign? That ain’t CPU rendering. MS has GPU farms (and a GPU-enabled Azure OS sku), and they’re growing.

So obviously if GPU’s in the cloud can provably provide complex lighting calculations for real-time rendering, it not only helps my argument it *proves* MS’s claim to be true on the lighting front. 100% proven true, in fact. Demonstrably.



I think you're intentionally minimizing the difference.

MMO’s traditionally charge for memberships. Why? Because it’s extremely expensive to host them. With Xbox One’s cloud support, it’s 100% free to developers… so now, persistent worlds don’t just have to live in MMO’s… they can live in games of all genre’s without needing to require dedicated monthly subscriptions. Persistent worlds aren’t new – no – but persistent worlds for *any* game in *any* genre for FREE is, and It’s hugely different. It unlocks all kinds of potential and scenarios for devs… and not just big devs. Indie’s. Smaller studios, etc…

AI – Bots? Bots are *nothing* like what Forza has accomplished… so for you to make any claim like that as some “argument” just exposes your ignorance on the topic. If you believe that argument, you're behind the curve.

As for physics, I’m not sure I disagree with you much there… I’m not sure it’ll make much sense to offload it to the cloud, but it’s possible, and that’s all MS claimed.



1. No one claimed “vastly improved”. That’s just hyperbole from you.
2. Playstation Now will certainly come to fruition, but you and I both agree it won’t live up to what Sony’s promised.
3. I notice you specifically avoid mentioning AI, because that’s provably and non-debatably already been proven advantageous by the Cloud with a LAUNCH game.

Bottom line – the claims you imply MS made are exaggerations at best, and simple falsehoods at worst. MS’s claims about graphics improvements via the cloud are very modest. It was never, “the cloud WILL make games 100x better visually!” – all the claim as ever been is the cloud CAN help in areas like, X, Y, and Z. That’s a very modest claim, and demonstrably true.

The claims which have been much bigger and more prominent are the ones around what cloud compute (CPU usage) can offer… and so far, it’s already delivering… and for a completely new paradigm introduced to gaming – I’d say making real progress at launch is something to be commended
The driveatars in forza5 are amazing. I was really imprressed how well my driveatar mimicked me when I went over a friends house recently and saw first hand, my friend racing against my driveatar. I was impressed. It's little things that I just stated above that truly define next generation IMO.
 
  • Like
Reactions: Mcmasters
I'm quite confident PlayStation Now will not be what Sony promises, but I'm actually not so sure cloud compute won't be what we've promised.

Forza already shows - today - that cloud compute can make a huge difference in AI. I've never raced in a racing game before against a non-human player, and seen so many human-like mistakes, self-corrections, and deviations from typical AI lines and behaviors. It's really an underrated achievement right now, because it's not something which is obvious and immediately recognizable.

So I'm curious, what claims did MS make about cloud compute that you don't think will be lived up to?

Lets be fair. Both companies promise the moon, but never get there with said promises. However Sony is the king for over hyping everything and under delivering big time. Whether it be with Vita, games, or services.
 
Lets be fair. Both companies promise the moon, but never get there with said promises. However Sony is the king for over hyping everything and under delivering big time. Whether it be with Vita, games, or services.

Agreed.

They both fluff statements with "PR speak", but (as I think you and I agree with), there's a scale... And Sony's tends to be much further towards the "fantasy land" extreme than Microsoft's.
 
Maybe Sony can use its cloud service to finally fulfill the 1080p @ 120 fps claim. It'll be 8 years too late and on a totally different console, but better late than never.
 
Agreed.

They both fluff statements with "PR speak", but (as I think you and I agree with), there's a scale... And Sony's tends to be much further towards the "fantasy land" extreme than Microsoft's.

... giant crabs ...
 
Xbox Cloud compared to PS now ?

hmmm

one possibly adds to your gaming experience

one possibly subtracts from your wallet
 
I think most of the "hate" towards MS's Cloud is from BS spewed by the likes of astro who claimed the cloud would magically allow the X1 to run at super high resolution, free AA, and pretty much look like the X1 had quad Titans running in it.

With all due respect, A lot of what Astro says is based on his own opinion which can be quite bias, and a lot of that can we clearly see with each passing day.
 
So I'm curious, what claims did MS make about cloud compute that you don't think will be lived up to?

"A rule of thumb we like to use is that [for] every Xbox One available in your living room we’ll have three of those devices in the cloud available," he said.
link

It's looks like he's saying plug your Xbox One into ethernet, and it's 3 times as powerfull. :-/

I know it's probably only for physics, cloth-simulation, fluid-dynamics, lighting, etc, the things he said were ideally for devs to do in the clouds, since you didn't need those on all frames.
But even that's pretty far-fetched, to get a 3 time power-increase in the cloud, to all users.:-/

It still has it's weaknesses if you do those calculations serverside instead of locally, wich he dosn't mention - just like your Forza comparison with 'Driveatars' instead of serverside stored ghosts - the AI will have physics, but it won't be the actual people.
The AI dosn't really look at how you drive either, it just measure a group of settings and learns from that.
If you finnish your 20 first races by doing a 180-turn with handbrake and backing your car in reverse over the finnish-line - and one of your friends loads your driveatar, he won't see your drivatar do the same and reverse over the finnish-line - just ahead of him, because the drivatar don't check for that. If it had been a server-side ghost, and you had ended your race that way, he'd seen that.
So everything has it's pluses and minus, just because it's in the cloud dosn't mean it's better.

Even if Microsoft manages to put out the tools and resources to offer all that he suggest as prime examples of the cloud-computing for developers to do - they will still face the biggest problem, alot of the same problems as Sony had with bringing content to Home - the content providers don't want to put their resources into using the MS-cloud - because Xbox is only a small part of the overall gaming market - and it will also have to also be made for users who isn't connected in the clouds - and MS will most likely have to oversee most of the process. Much better to use their own server-tech, wich they can integrate into all platforms and control themselves. :-/
 
Driveatars are a asynchronously calculated. They could just as easily be calculated during a loading screen inside of the game, and uploaded to any server and then downloaded to friends.. Hell could use an idle core during menus and other non-intensive game portions.

Titanfall appears to actually use the cloud for nothing much different than Unreal Tournament bots on a dedicated server did over a decade ago.

The cool thing at this point about "the cloud" are the API's and potential for infinite instances of whatever server anyone wants to create, dynamically partitioned and allocated.

Cool stuff, but it's a bit tiring to continue to read people talk about "cloud" features that can be done without.

Using distributed computing to do more and more is certainly going to continue to happen.. But MS doesn't have an exclusive on using servers. And most companies are unlikely to build such complex code only for one platform.

Anything that relies on server computing will either become buggy, broken, laggy, Etc.. or be able to have an alternate way to do things.... Or really do both, because there is no magic code that can suddenly handle a physics calculation being done on a server only to have to switch to client compute and gracefully adjust for the sudden non-offloaded task.

I'm just not buying this idea that anyone wants or wants to create games like that.

We'll see an awesome amount of dedicated servers... And if we do see more things like driveatars well see them done cross-platform.
 
It's looks like he's saying plug your Xbox One into ethernet, and it's 3 times as powerfull. :-/

He says it's a rule of thumb, and I don't disagree that it's not exactly the best phrasing of what it means to utilize the power of the cloud, but it's not inherently false either. In some cases it'll be more than that (connecting to a massive persistent world, for example), and in other times it'll be less than that (using Forza's drivatar system)... but it's a rule of thumb, not a promise across the board.

It still has it's weaknesses if you do those calculations serverside instead of locally, wich he dosn't mention

What are you talking about? Of course he mentions weaknesses and limitations:

The X1 cloud architecture will help support "latency-insensitive computations"... latency-sensitive things like animations in a shooter, reactions to hits and shots in a racing game, reactions to collisions... those things you need to have happen immediately and on frame and in sync with your controller

He goes on to mention limitations around ISP instability, how even if lighting were to be offloaded, it would be somewhat asynchronous to the users actions ("So when you walk into a room, it might be that for the first second or two the fidelity of the lighting is done by the console, but then, as the cloud catches up") - the interview is FILLED with caveats, weaknesses, and obstacles to overcome...

Did you read it, or are you just picking out a single line to complain about without actually reading the full context?

- just like your Forza comparison with 'Driveatars' instead of serverside stored ghosts - the AI will have physics, but it won't be the actual people.

The AI dosn't really look at how you drive either, it just measure a group of settings and learns from that.
If you finnish your 20 first races by doing a 180-turn with handbrake and backing your car in reverse over the finnish-line - and one of your friends loads your driveatar, he won't see your drivatar do the same and reverse over the finnish-line - just ahead of him, because the drivatar don't check for that. If it had been a server-side ghost, and you had ended your race that way, he'd seen that.
So everything has it's pluses and minus, just because it's in the cloud dosn't mean it's better.

That's irrelevant to the argument. Regardless of how good or bad the drivatar system is (which I'd argue is easily a huge step up in traditional AI which is actually far worse, far more simplistic, and far more predictable than most people know) - this fact remains - Forza's AI when connected to the cloud is completely 'out-sourced' to the cloud. How much computational power it offloads from the X1 itself is not something I have the specific data for, but the point is it proves the claim of offloading the AI, and it is a 'self-teaching' AI (like Bing, Google, or other big-data driven online portals) - so regardless of its limitations, your preferences, etc. - it supports/proves Microsoft's claim about AI offloading.

Even if Microsoft manages to put out the tools and resources to offer all that he suggest as prime examples of the cloud-computing for developers to do - they will still face the biggest problem, alot of the same problems as Sony had with bringing content to Home - the content providers don't want to put their resources into using the MS-cloud - because Xbox is only a small part of the overall gaming market - and it will also have to also be made for users who isn't connected in the clouds - and MS will most likely have to oversee most of the process. Much better to use their own server-tech, wich they can integrate into all platforms and control themselves. :-/

This is an assumption/assertion which really isn't valid. Unless you can point to enough developer feedback to suggest as much, I'd argue this point simply comes from ignorance.

1. Using Microsoft's cloud for X1 is free for devs.
2. Using Microsoft's cloud for other platforms is possible, and affordable (SO much more affordable than Devs building out there own solutions)
3. Using Microsoft's cloud gives you virtually free load balancing, and literally free multi-regional world-wide support... which is something EXTREMELY costly (usually prohibitively so) for any devs - even the ones building "AAA" titles.

So yeah, this last point is really just not true at all, and I don't believe it's a reasonable claim in the least. If you can get some data to show otherwise, let me know.
 
Last edited:
Flynn, you asked 'out of curiosity' for a real example on what sceptic people thought wouldn't be delivered.
I brought up what I tought wouldn't be delivered. Now it's MS job to prove me wrong,and live up to the promises, not the other way around. :)

But since I'm nice, I can explain some more.

I didn't mean he didn't mention any weakness - i just mentioned it had weaknesses wich he didn't mention and likened it with your Drivatar-bragging - and explained a real-world scenario you can test right now today for yourself for such a weakness in the drivatars - wich were a AI with so human-like mistakes as you said. Feel free to try it out for yourself. :-/
The driveatars isn't 'self-teaching' - that's also misleading - its reading a set of parameters, according to Forza-developer interviews, I've read atleast.
As I understand it, - one set of parameters it looks on is your latest speeds going into and leaving particular types of curves on the track, and positioning - and then decides since you've done those types of turns this way for example the last 3 times - your Drivatar will attempt a similar turn, on those turns, and that's how it learns.

A programmer has to go in and program the AI parameters what to look at - i.e. to see if you're actually finnishing your game, with the car in reverse backwards - the Drivatar would not be tought (or self-tought, if you wish -) to finnish the race the way I described - no matter how many times you are doing that example I mentioned - the drivatar wouldn't learn it, until the programmer went in and set that as a parameter to read - But a ghost would do that,. i.e. it all got pluses and minuses.

The last assumption/assertion I made, I can't guarantee - but I figure most gamers will see how I deduce my logic.
If devs don't put resources into porting to smaller platforms such as Vita/Wii U, despite their dollar are probably just as welcome as PS4/XB1 gamers dollars - I think it's fair to assume that this will also be the case for even smaller scale xbox cloud-computing platform - wich only a smaller segment of the gaming-population have.
That's also mighty close to the excuse MS used to put Kinect on every XB1, despite it increased costs for gamers much more initially, and many gamers don't want it, MS needed to make sure that all XB1-gamers had it - in order to get dev-uptake.
- MS can't guarantee that all Xb1-gamers ISP's will be able to connect with XBcloud good enough to guarantee it would be there when they want to play something to use it as a mandatory option., so it will be treated as a less important addon platform, wich might not allways be welcome. :-/
I could give you half a dozen more examples, but you get my point. :-/
 
Not an expert on hardware, but I read a while ago an interview with the man who is in charge of PS4 hardware, & his opinion on cloud computing.

He mentioned, that while he see some potential (on non server stuff) in cloud, he think its better to beef up the spec of the PS4, to give developers a strong & consistent baseline from the get go, instead on relying on something that has a lot of variables.

The more I think about it, the more I agree. Bandwidth differ from place to place, & not all gamers may have a constant internet connection to the console. Also it takes extra effort to integrate the cloud computing into the pipeline. Unless its game changing, or give substantial boost, I cannot see many 3rd party developers putting in extra development time in making good use of cloud computing.

I wished to be proven wrong, but the bottomline is MS have talked a lot, on the decision on their hardware, & how close is the performance compare to PS4 (though they never mention it to be superior), how Cloud & Kinect enhance game experience.

What they need to do is to talk less, show more. Show something that can only be done via kinect, cloud, that blows people away. Show that sub HD resolution is only early in console cycle & will be a distant memory soon. At $100 higher price, its about time (post launch) to convince the rest of the gamers (that is not big XBOX fan or hardcore gamers) to pick it up, over the competition, or even its predecessor, the 360.
 
Remember when always online was a bad thing? What about the troops?? :laugh:
Then they can't stream games to their PS4? I don't understand what's so difficult to understand the differences between a service like this and what MS wanted to do.
 
Remember when always online was a bad thing? What about the troops?? :laugh:

This is optional though? If you want to play PS3/PS2/PS1 games on the PS4, the only way is to stream them and obviously you need to be online to do that...