Next Gen - The Verdict
- PaulCunningham
- Baaaaa!

- Posts: 1218
- Joined: Sun Jan 18, 2004 11:47 pm 0
- Location: Walsall, England
- Contact:
Re: Next Gen - The Verdict
PS forgot to mention (and can't edit now). You don't even need to JIT managed code. You can use Ngen.
And then there's AOT (which is how the magic of C# on iOS works).
And then there's AOT (which is how the magic of C# on iOS works).
Re: Next Gen - The Verdict
Well it turned out to be enough to outperform the 360 once devs got used to it.Dio wrote:It's not a very good one.
Re: Next Gen - The Verdict
'tufty wrote:Bullshit. The tools are free. You can build, package, and test in the simulator for free. It's only once you start rolling your app out to devices that you have to pay.Darkflame wrote:And, yes, Apples fee is cheaper then dedicated devkits. But its not exactly jump-in friendly. You need investment before you can even tinker.
Well, of course its free if its in a emulator or what-not.
We are talking about people wanting to tinker with the actual real hardware they have purchased. The cool factor of having your coding running on that little box you purchased.
Or you can dump it anywhere you like and send the link to your friends as any Android device will run it. I assume Ouyo will be set up the same way - they enable 3rd party downloads, then you can send your friends your creations.It should probably be noted that Google's store is not free if you want to distribute paid content ($25 fee applies), and you can *only* put stuff on the store if you happen to be in one of a handful of countries.
.
No ones moaning about fee's if you want to actually publish and make money of stuff here. Its about accessibility, not money.
-
want to read eye-wateringly bad stories? No? Ok.
want to read eye-wateringly bad stories? No? Ok.
Re: Next Gen - The Verdict
That Google Play fee of $25 is a one-off fee. After that you can publish as many apps as you want forever.
Re: Next Gen - The Verdict
Digressing a little, I'm not placing any bets on this generation until Valve officially unveil the Steam Box.
It's greatly dependent on how Valve manage the launch, of course, and what price / power point they're going to chase exactly. It's very easily screwed up, by pricing it too high or fucking up the launch, which is why I'm reserving judgement.
But, in theory, it has serious potential.
Because Valve already have the well-established store. As a PC-in-a-box, it's not going to be the underpowered compromise that the Android-based consoles are - the Steam Box will have the necessary grunt to run "proper" games (how much, we don't know yet but, as modern PCs don't lack in such power, even the most cheap-and-cheerful configuration surely can't be bad).
It's also open hardware, in the sense that other manufacturers besides Valve can make their own Steam Boxes. So even if it does initially lack on the price / power point, internal competition might help drive that price down and the power up. And we already know - from all that media confusion when Xi3's PC-in-a-box was initially mistaken as being the official "Steam Box" - that there are third-party manufacturers keen to get in on the Steam Box action, as Xi3 are even attempting to do it and ride on the wave before the Steam Box is even officially a thing.
And, as Valve are really a software company, they'll likely encourage such an internal market because, for them, it's not the hardware, it's the Steam store. So, from that perspective, if someone else beats them to a more powerful Steam Box at a cheaper price, as long as it's hooked up to the Steam service as a "Steam Box", they still ultimately win.
It's an open platform because they essentially want that kind of thing to happen. It's a longer term strategy, but if an internal market takes off for "Steam Boxes", as once happened with "IBM PC compatibles", they're laughing all the way to the bank.
From that perspective, selling the hardware at cost price - and not giving a shit to make any profits off the hardware itself - is an entirely feasible strategy for them to pursue. Because this is really about expanding the reach of the Steam store in that light.
And, speaking of PCs, an interesting thing that PC-centric TotalBiscuit noted about the E3 line-ups for the XBone and PS4 was that when they said "exclusive", it generally meant "not on our rival's console" but not "this won't be coming to PC" (TB is a PC man, so he explicitly went looking for such confirmation... and generally found it).
Most of the big third-party titles fanfared at E3 will eventually show up on PC anyway (makes sense, really, as both the XBone and PS4 are x86-based consoles and - if it's an easy port, which it likely will be, as that crashed XBone demo machine showed that these were being developed on Windows PC - why would a third-party throw away the PC market, unless Microsoft and Sony start bribing them with sizeable amounts of cash to do otherwise, which they don't appear to be doing?) - and that essentially translates as "available on Steam".
That is, other than first-party titles and the properly exclusive "exclusives", a lot of those games heralded at E3 will eventually end up on PC and, therefore, on Steam and, therefore, possibly on Steam Box (particularly if it does sell like hot cakes from a great launch).
And Valve themselves have major IP too.
I mean, let me just leave you with this final tantalising thought:
What if Valve finally learnt to count to three with the Steam Box launch?
Discuss.
It's greatly dependent on how Valve manage the launch, of course, and what price / power point they're going to chase exactly. It's very easily screwed up, by pricing it too high or fucking up the launch, which is why I'm reserving judgement.
But, in theory, it has serious potential.
Because Valve already have the well-established store. As a PC-in-a-box, it's not going to be the underpowered compromise that the Android-based consoles are - the Steam Box will have the necessary grunt to run "proper" games (how much, we don't know yet but, as modern PCs don't lack in such power, even the most cheap-and-cheerful configuration surely can't be bad).
It's also open hardware, in the sense that other manufacturers besides Valve can make their own Steam Boxes. So even if it does initially lack on the price / power point, internal competition might help drive that price down and the power up. And we already know - from all that media confusion when Xi3's PC-in-a-box was initially mistaken as being the official "Steam Box" - that there are third-party manufacturers keen to get in on the Steam Box action, as Xi3 are even attempting to do it and ride on the wave before the Steam Box is even officially a thing.
And, as Valve are really a software company, they'll likely encourage such an internal market because, for them, it's not the hardware, it's the Steam store. So, from that perspective, if someone else beats them to a more powerful Steam Box at a cheaper price, as long as it's hooked up to the Steam service as a "Steam Box", they still ultimately win.
It's an open platform because they essentially want that kind of thing to happen. It's a longer term strategy, but if an internal market takes off for "Steam Boxes", as once happened with "IBM PC compatibles", they're laughing all the way to the bank.
From that perspective, selling the hardware at cost price - and not giving a shit to make any profits off the hardware itself - is an entirely feasible strategy for them to pursue. Because this is really about expanding the reach of the Steam store in that light.
And, speaking of PCs, an interesting thing that PC-centric TotalBiscuit noted about the E3 line-ups for the XBone and PS4 was that when they said "exclusive", it generally meant "not on our rival's console" but not "this won't be coming to PC" (TB is a PC man, so he explicitly went looking for such confirmation... and generally found it).
Most of the big third-party titles fanfared at E3 will eventually show up on PC anyway (makes sense, really, as both the XBone and PS4 are x86-based consoles and - if it's an easy port, which it likely will be, as that crashed XBone demo machine showed that these were being developed on Windows PC - why would a third-party throw away the PC market, unless Microsoft and Sony start bribing them with sizeable amounts of cash to do otherwise, which they don't appear to be doing?) - and that essentially translates as "available on Steam".
That is, other than first-party titles and the properly exclusive "exclusives", a lot of those games heralded at E3 will eventually end up on PC and, therefore, on Steam and, therefore, possibly on Steam Box (particularly if it does sell like hot cakes from a great launch).
And Valve themselves have major IP too.
I mean, let me just leave you with this final tantalising thought:
What if Valve finally learnt to count to three with the Steam Box launch?
Discuss.
"Words ought to be a little wild - for they are assaults of thought on the unthinking."
- John Maynard Keynes
- John Maynard Keynes
Re: Next Gen - The Verdict
To be fair it is about the computation if that's the bottleneckDio wrote:High performance computing isn't about computation. It's about bandwidth abd IO and caching, which Cell isn't particularly good at, especially if ease of use is part of your criteria.
There is an alternate reality where a next gen Super Cell that has a well integrated graphics part and that stupid fast PS4 memory is combined to make an awesome system that goes like the clappers. The PS3 kind of proved that even if your system can do something amazing eventually game coders will find a way to get it to amaze the end user even if it's ridiculously hard work to do so.
That all said I think Sony have completely made the right decision here. The AMD chip is pretty amazing, as are most modern PC CPUs / GPUs / APUs, they're just not pushed that hard on PC. The programmers seem a bit lazy, performance solutions are sometimes just wait for the next generation of HW to come out. Games are often written for a pretty dumb lowest common denominator and you've got an OS and driver sucking up a shit of load of performance out of the system.
The interesting thing about the coming console generation is if the games on it will give PC coders a bit of a kick up the arse. We'll see
Re: Next Gen - The Verdict
That's the point though. There are only two companies in the world who can provably build low-cost high performance graphics - AMD and nvidia. I think Intel and Imagination could probably do it, but scaling up from their level they probably wouldn't get it quite right first time. Not sure about Qualcomm and ARM yet (although Qualcomm at least is largely staffed by ex AMD people, quite a few of them on a different floor of the same buildingThat all said I think Sony have completely made the right decision here. The AMD chip is pretty amazing, as are most modern PC CPUs / GPUs / APUs, they're just not pushed that hard on PC.
Being deluded into thinking you know graphics better than you do leads to architectural monstrosities like PS2, then dead ends like Larrabee for graphics and Cell APEs.
Which then means:
So they'd still have to buy the graphics IP, and how much extra R&D on Cell to get there? If the $10bn figure on Cell R&D is right, and say they shift 100m PS3s (and the other users of Cell were negligible) - that's a $125+ cost component in every console.gravy wrote:There is an alternate reality where a next gen Super Cell that has a well integrated graphics part and that stupid fast PS4 memory is combined to make an awesome system that goes like the clappers.
Fundamentally Sony learned a big lesson on risk, and the custom APU is a low-risk solution with all the key parameters defined: upfront costs, performance level, programming model, cost reduction over hardware life.
Re: Next Gen - The Verdict
Oh yeah, obviously. It's an economic decision primarily and a good one too.Dio wrote:Fundamentally Sony learned a big lesson on risk, and the custom APU is a low-risk solution with all the key parameters defined: upfront costs, performance level, programming model, cost reduction over hardware life.
I'm making the point that if you had the resources and built something specifically for what the device is intended for you could probably make something awesome and from a CPU pov the CELL isn't a terrible base to start from.
You're missing something here though:
That's a weakness IMHO. Its lead to domination in a shrinking market and minor participation in the biggest sized / fastest growing GPU sector; mobile. Which is a great example of the success you can get building to the customers needsDio wrote:That's the point though. There are only two companies in the world who can provably build low-cost high performance graphics - AMD and nvidia.
So overall, right path for MS / Sony but a bit of a minor shame if you interested in new things / exotic tech.
Re: Next Gen - The Verdict
Nobody's worked out how to do anything new in graphics that actually works
. Our theory is that existing tech pus HSA is the way forward at least for now...
Re: Next Gen - The Verdict
And all of a sudden the only thing I can think of is when I was reading in the 90s that the future of 3D graphics was ray tracing. 
James
James
Re: Next Gen - The Verdict
No - that was the 80s and the 2000s. In the 90s it was voxels. 
- merman
- Sheepie Whore

- Posts: 10943
- Joined: Wed Feb 16, 2005 1:44 pm 0
- XBL Gamertag: merman1974
- Yahoo Messenger: andrewrfisher
- Location: Skegness, England
- Contact:
Re: Next Gen - The Verdict
With PS2 and PS3 we often heard that developers found it hard to deal with the processor, much as the Saturn was a beast to develop for. By bringing the PS4 and Xbone closer to PC tech it makes things easier for developers, and middleware like the Unreal Engine will smooth the transition. Still, it will be interesting to see where things go in a couple of years time when the top developers really understand and can push the hardware close to its limits...
Twitter, Xbox, Wii U and Steam - merman1974
ZZAP is back - https://www.patreon.com/zzapmagazine
ZZAP is back - https://www.patreon.com/zzapmagazine
- gilesgoat
- Sheepie Whore

- Posts: 27259
- Joined: Wed Sep 03, 2003 7:54 am 0
- Location: Monfalcone, Italy
- Contact:
Re: Next Gen - The Verdict
PSSST if you want suggestions about NEW GPUs and stuff we ( me and Jeff alone probably ) would likeDio wrote:Nobody's worked out how to do anything new in graphics that actually works. Our theory is that existing tech pus HSA is the way forward at least for now...
them to do just ask
Probably we'd be the only 2 customers in the world as well but if you want innovation ...
Giles the billy goat
"Baa"
"Baa"
- gilesgoat
- Sheepie Whore

- Posts: 27259
- Joined: Wed Sep 03, 2003 7:54 am 0
- Location: Monfalcone, Italy
- Contact:
Re: Next Gen - The Verdict
Well "exactly" .. we had to buy a Mac ( actually 2 ) specifically for that, it's not "a devkit" in the realDarkflame wrote:No, you need a mac and a yearly fee.: iOS doesn't require a dev kit,
sense yet still something you have to buy ( if you don't have it already ) to develop on.
Likewise even if "kinda totally free" we had to invest money and buy at least a couple of different
Android devices to test stuff "for real" and see how it goes "on a real device" and even that for as
( not really SO ) little as it can be is still money you have to spend to develop.
Devkits of other stuff .. "they are still amazing piece of HW", often having MORE sophisticated features
than the real "consumer device", the "consumer device" does not need to have debugger things or more
complex stuff on it.
There's always some investment/money that goes spent into something to make things happen.
You have also to consider pretty well "the kind of investment you are talking about", because sometime
we all I think "loose the focus a little bit".
Let's say to say "a stupid nonsense number" .. "2000 pounds of investment for a devkit ?" .. YES OF
COURSE it's NOT peanuts ..
But think about this .. "I am going to open a stall to sell fried chips and sausages" .. oh I need a profressional
deep frier "safety compliant" and a couple of bottle of gas and/or some attachment and a license
for food and such ..
Oh .. "10,000 pounds of investment .. 4000 just for the frier .. " and that's "normal" and I am "just"
frying some chips and sausages
Up to a certain level all this "games development" is still a place with "not so big startup costs", there's
much much worse than that for stuff that "commonly" seems/look more trivial.
Giles the billy goat
"Baa"
"Baa"
- gilesgoat
- Sheepie Whore

- Posts: 27259
- Joined: Wed Sep 03, 2003 7:54 am 0
- Location: Monfalcone, Italy
- Contact:
Re: Next Gen - The Verdict
[edit] ok maybe 4000 for a deep frier was a bit far fetched but a quick look at industrial restaurant
friers they are around 1000 .. 1600 pounds depending on size, I remember time ago I was checking
for a tandoor they are about that size.
You can probably get away with 200 ... 400 quids if you look something medium size for really not too
extreme frying.
Keep in mind if you fry "decently quite a bit" ideally you should change oil every 15 days or so so that's
probably between 10 and 20 liters of oil that go.
Also keep in mind "time as a cost/investment", if you spend let's say 3 months of your time doing SW
isn't that "an investment" as well ? Should you not consider those 3 months like "3000 pounds
spent in doing something ?" ( that someway you have to get back or you are starting to get into loss ? ).
There are lot of "hidden costs" anywhere but I think it's quite "safe" to say that "up to a certain
point/level" writing SW is still a pretty low investment application, where probably "biggest costs"
are still/always PEOPLE not machines.
friers they are around 1000 .. 1600 pounds depending on size, I remember time ago I was checking
for a tandoor they are about that size.
You can probably get away with 200 ... 400 quids if you look something medium size for really not too
extreme frying.
Keep in mind if you fry "decently quite a bit" ideally you should change oil every 15 days or so so that's
probably between 10 and 20 liters of oil that go.
Also keep in mind "time as a cost/investment", if you spend let's say 3 months of your time doing SW
isn't that "an investment" as well ? Should you not consider those 3 months like "3000 pounds
spent in doing something ?" ( that someway you have to get back or you are starting to get into loss ? ).
There are lot of "hidden costs" anywhere but I think it's quite "safe" to say that "up to a certain
point/level" writing SW is still a pretty low investment application, where probably "biggest costs"
are still/always PEOPLE not machines.
Giles the billy goat
"Baa"
"Baa"


