Cloud gaming - the end of the desktop performance pc cpu market?

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Cerb

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
kelco said:
Is it failing because it never worked or for other reasons?
Click to expand...
The reason is, and will remain, that fat clients respond faster, that the servers needed to keep up with demand are nearly as power-hungry as a desktop gaming computer, for the same workload, and that people like having limited play option when connectivity isn't perfect.

It's a pipe dream. It has been a pipe dream for longer than I have even been alive.

Look at phone and tablet tech now. The only thing they need to do reach ~2004 performance levels is to make a business decision to spend more money on pads/balls, traces, and chips, to have wider memory interfaces. That's it. And that's not considering that we likely will keep seeing improvements at a more rapid pace than we see on the PC.

podspi said:
I think they would prefer selling to businesses virtualizing GPUs than consumers. Firms will be willing to pay more per card, knowing they can use the same card to service multiple customers.
Click to expand...
OTOH, NVidia's compute devices are cheap, and used locally, they can take advantage of local disk and network speeds. For regular GPU use, there is no need to go off-site to do the job.

Idontcare said:
But don't fall victim to this "technology as a snapshot in time" mentality. Bandwidth is improving moving at ridiculous speeds. Hell, the very fact that the concept of bandwidth itself is prevalent enough that you and I can even have this conversation about the very topic is absurd proof of the pudding type stuff.
Click to expand...
But when will bandwidth actually improve significantly enough? Local performance has consistently outpaced remote bandwidth improvements. Unless somebody starts replacing everything with fiber, at least to approximately a block or so, I don't see it. And, who is doing that, outside of a handful of major cities?

Then, once it is done, what will allow the servers to be efficient enough to be worth paying the power bill for, during peak hours? Even assuming the latency can be dealt with, it's not something on the horizon.

The problem is fundamentally that Netflix can buffer minutes ahead, just like Youtube can. Not only can it, but I've ridden out intermittent internet connections with no drops in playback. Most games are soft real-time systems, and they just don't work that way. With low bandwidth (100% client rendering), and game play that is not dependent on low and relatively constant input lag, it can and does work. But, most games are going to need relatively low constant upper bounds on lag, on top of any other concerns. Without making special deals to distribute servers with heavy locality, which would cost way too much, I don't see how it will work with any tech known to be on the way. The bottlenecks are pervasive.

Never say never, time will make a fool of you every chance it can get ;)
Click to expand...
I won't say never, but I will say it will take a disruptive change, not from a major telecom (or, a disruptive change forced on to major telecoms), to fix the most basic networking problems.

Idontcare said:
If the ping is that long then yes, no question, the model itself won't take off for those games which are dependent on latency (FPS). But there is a large group of people who don't need that - they play farmville already.
Click to expand...
But, Farmville already requires very little CPU, and practically no GPU performance per client. That will leave a user with the weak hardware to just do that unable to play anything more demanding.

Client/server FPSes are nothing new. They've been around since at least Quake. But, they don't render the game on the server. They arbitrate what's going on, and what your client should be able to know about. Then, your client does the rest. Doing that requires fat clients. At that point, the technical benefits disappear.

Even running an old computer, today (C2D and GTX 460), my local computing power has increased at a far greater rate than my internet speed has, over the years. Not coincidentally, there is no real competition between the ISPs, either: one is faster, one is cheaper, and both want you get their TV/net/phone bundle contracts. There was a burst of activity when symmetrical broadband was coming in, and then it's been small incremental improvements since then.

MisterMac said:
The monoply of a certain company is the only thing keeping us back ;).
Click to expand...
Verizon? AT&T? Charter? Cox? There are quite a few to go around.

I find it funny you use an argument like a single CARD cannot service multiple users.
NVidia is movingly heavily towards more threaded architecture on a hardware level - then it's just the same problem on the cpu-side.
Uber Fast "long distance" interconnects.
Click to expand...
It's not the threading. That's 100% a single-user performance issue. Sufficient virtualization technology already exists (sufficient != easy to get and keep running...that will take time). The problem is that you're effectively taking a resource and dividing it up, when it is a resource that a single user can nearly max out, most of the time. It is also a resource with high thermal density. Also, the thermal density has remained one of the bottlenecks to performance, so as more perf/W comes about, greater detail will be demanded and given, and greater detail will available locally to each fat client user at a lower price. On top of that, it's got to go a long distance.

kelco said:
Wanting something and foreseeing its inevitability are two different things. I've worked a gameroom, I've been an It administrator and I'm a gamer. Depending on what hat I'm wearing will determine what I'm looking for from a given technology. But its obvious for many reasons why you'd want virtualization in some scenarios. I'll just get back to talking to myself about such things, pretty much a waste of time here.
Click to expand...
You want it almost entirely for control over the players, and auditing of their actions. Technically, even that can be better done, usually, by having a fat client and server, rendering entirely on the client.
 
Last edited:
I

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
kelco said:
:whiste: What a constructive addition to the thread. Is this what I have to look forward to in this subforum, because I'll gladly leave if this is what anandtech's reader base is. Jeebus.
Click to expand...

I'm sorry you feel that way. My post was constructive criticism. I am not going to repeat what Gryz said because it sums up my opinion as well.
 
C

ChippyUK

Member
Jan 13, 2010
99
1
71
KlokWyze said:
/end thread

online gaming is shoddy by my standards most of the time, unless it's some turn based [stuff] or something.
Click to expand...

Dunno could be like playing original quake before Carmak and co introduced prediction ala Quakeworld :) It would be quite fun to try that again.

Oh to go back to '96/'97 for decent clan and community games.... I have very fond memories of those times. Now games feel so disconnected and quiet, almost soulless as no one chats, there is no personality to the player, just ability.

Okay okay end of rant ;)
 
Last edited by a moderator:
B

Ben90

Platinum Member
Jun 14, 2009
2,866
3
0
ChippyUK said:
Oh to go back to '96/'97 for decent clan and community games.... I have very fond memories of those times. Now games feel so disconnected and quiet, almost soulless as no one chats, there is no personality to the player, just ability.

Okay okay end of rant ;)
Click to expand...
I have noticed this as well. I'm not sure if its just nostalgia, or games have lost their "soul".
 
NTMBK

NTMBK

Lifer
Nov 14, 2011
10,526
6,051
136
blastingcap said:
Gaikai sold for $380 million though, to be fair.
Click to expand...

Newscorp bought Myspace for US$580 million, and sold it for around US$35 million. Overhyped bubbles have happened before, are happening now, and will happen again.
 
blastingcap

blastingcap

Diamond Member
Sep 16, 2010
6,654
5
76
NTMBK said:
Newscorp bought Myspace for US$580 million, and sold it for around US$35 million. Overhyped bubbles have happened before, are happening now, and will happen again.
Click to expand...

See my first post in this thread: http://forums.anandtech.com/showpost.php?p=34044849&postcount=25

I am not a fan of cloud gaming. I was just trying to be fair in terms of what cloud companies may be worth. As much as 380MM may be overstating things, the pittance OnLive sold for may be understating things as well.
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom