The point is if sth is too difficult and take too much effort to code then developers will not or will at least take a long time to adopt the technology.
About 60% of PS2 titles utilize VU0 and around 25% use both VU0 and VU1(this is based on Sony's internal testing)- and you are pretty much forced to handle those in assembly(particularly those titles released in the first several years).
You seem to imply that great titles will have to be MT.
Any big advances in gameplay will require it. Tetris is still a blast to play- but we have evolved a bit.
From my understanding, you were trying to say MT programming is easy from the example you provided.
I'm saying that blanket statements don't work. If you are of the mindset of making your application MT from day one it isn't going to be nearly as difficult to multithread it as some would have you believe.
In the present, a processor's power is basically linearly proportional to its transistor count.
Complete BS and you know it. Compare the P4 3.2GHZ EE with the standard P4 3.2GHZ and look at the transistor counts and then compare the performance. You can try and get a rough estimation via transistors devoted to computational output- but in order to do that you need to remove a sizeable chunk of the counts for both the P4 or the Athlon(front end translation from x86 to uops, BPU, trace cache etc).
It is cheap and lacking power compared to what Sony originally claimed with those TFLOPS numbers and unrealistic statistics that was supposed to be a PC killer.
Which of Sony's claims in terms of TFLOPS rating are you saying are unrealistic? I've seen several numbers thrown around over the last few years, some of them are completely unrealistic, some of them are easily proven accurate(in terms of peak throughput).
Sony has to control cost thus die size.
Sony is more agressive with large die sizes then any consumer electronics company has ever been. Planning on shipping millions of units packing processing units with a die size in excess of 300mm is insane. They don't *have* to control die size in fact they have rather proven that they are far more willing then Intel or AMD to push the limits of any fabrication process they have worked with.
As the article pointed out, cache size and latency issues could plague it too, which will make it harder to code with the PPE units.
Most people who have worked with IOE processors will tell you that you will find yourself a lot less dependant on cache hits then you are with an OoE core.
Regardless you can give me long list of great PS2 racing games and I can give you a long list of great PC racing games as well.
Please give me that list of great racing games- because I have looked all over and can't find one that is remotely close to GT4 let alone Forza(which is easily superior compared to GT4).
My point is PCs can do everything a console can do, but console can't do everything PCs can do in gaming
No, they can't do everything as well as consoles- not even close. I have listed a large number of genres that PCs are utterly horrible at and have waited for a rebuttal- there isn't any. If you want to talk about how with enough customization you could wire up a LAN and use TV outs and then plug them all in to a large screen HDTV in order to get some decent MP going with your buds sitting next to you- you can also make the XBox into a 'PC'- and a lot easier too.
The fact is, once RTS and FPS and MMORPG took off, the demand for other types of games has diminished quite a bit.
The largest three MMORPGs combined have less subscribers then Myst sold copies- they are a very small slice of the gaming pie even on PCs. Myst clones tanked when people realized the consoles dominated the genre now as they always have. The demand hasn't shrunk in the slightest iota for genres outside of RTSs and FPSs- the exploding sales on consoles is testament to that. PCs simply don't handle the genres as well. Its flexibility that makes it so well suited for certain uses eliminates the ability to polish a title to the level that a console game can have.
These days it's simply not possible to program any reasonably complex piece of software in assembler without modern data structures and compilers to support you. There are thousands of studies that show structured high level languages always come out on top. That's not laziness, that's advancement in computer science.
Really?
Many of the upcoming Playstation 2 games like Gran Turismo 2000 are beginning to look far more promising than they did at E3. SCEE's own internal titles such as Formula One 2000 and Wipeout Fusion are looking a better than simply high-res, perspective-corrected Playstation products. Certainly spirits amongst the PS2 development community are higher than they were at the beginning of the year at Sony's European PS2 development conference, where nobody seemed to be getting good results on the machine. Slowly but surely Euro coders used to Assembly coding are getting their heads around the console.
Link.
The team working on the project consisted of several engineers, and everyone worked on all of the research tasks. Our work was focused on studying the materials Sony provided, writing Sony library simulation using OpenGL, dissecting OpenGL and porting to OpenGL. The research was also based on learning assembly, working on general graphics and on Nurbs sky simulation, volume shadow, soft shadow, fluid, noice and turbulence, palatted texture, animation, four-Nurbs methods, and more. The management of the preparation process was excellent, and included distribution of test materials, and attendance of game development conferences in order to keep the team updated.
Balancing EE and VU1 calculations, is a big issue with the Playstation 2. VU1 is known as highly efficient, however its resources are limited. The truth is that the VUs are designed to do specific calculations, that is, vector and matrix parallel calculation.
Another portion of the article-
Our efforts to perform heavy assembly level development and optimization turn out to be a big win. For each development platform game developers always tend to dive into assembly level to achieve maximum performance. It's especially true for Playstation 2 developers because it's necessary to use assembly in some parts of Playstation 2 development process.
Link.
I know what I am talking about here, I follow the console scene extremely closely including development processes. Creating self contained threads and working with assembly on consoles is
normal right now. Not because of the new consoles- it has been that way(at least in terms of dealing with assembly).
Right, so the PC developers that have been developing real engines for years that work in the real world are dishonest but the console developers who neither have finished hardware or products are to be trusted completely?
Console devs who code in assembly regularly tend to have a much better understanding of exactly what they are working with compared to a PC dev. These guys also have shipped numerous titles they built from the ground up for consoles- something Sweeney has done..... never.
It is a performance limitation because it cripples the processor when running code other than at 100% tuning.
That is NORMAL in the console industry.
Which Cell are we talking about now? Are the military using PS3 processors too? If so why don't they ditch their $30,000 systems and buy themselves some consoles instead?
So why not dump data warehouses for $300 emachines..... Your point cuts both ways.
Of course they do and the likes of DirectX, COM and .NET are going completely the opposite direction of your ludicrous "program in assembler or you're lazy" claims.
I'm saying the lazy devs are the ones whining like little b!tches. Most of the comments from the more ambitious people have been more along the lines of 'this is so much easier then the PS2'.
Microsoft have invested millions and millions of dollars into developing tools to make developer jobs easier because that is the way of the future.
MS created the specifications for Xenos on their own- nothing else in IBM's lineup is quite like it. They made the call to do exactly what you are saying they wouldn't.
Again, what exactly does "up and running" mean? How much of that program is multithreaded? And how fast does it run compared to PCs?
Knowing Sweeney- none of it was multi threaded.
And how fast does it run compared to PCs?
If it is all Sweeney I'm sure it's much slower. How fast can your PC decode a dozen HD video streams at once? Mine struggles with one- Cell had no issue pulling it off at the PS3's unveiling. Wonder how fast that would be if it were ported over to the PC?
So far you've demonstrated nothing but extremely vague comments to back your claims.
Now you have some developers on the record talking about it. Want me to pull up more from the PS2 devs? Unfortunately you aren't going to see too much public talk about development on the next gen yet as they are still under NDA.
Oh, and I thought Sweeney wasn't to be trusted? Or is he only to be trusted when he's pro-console?
He showed off the engine running in front of thousands of people on stage and on camera- I have the vid clip on my rig right now. If it wasn't for that I wouldn't trust him in terms of consoles- he is a noob in the console dev scene at best. He is trying hard to make his company a name as he knows where the market is- but in order to do that he is going to have to roll up his sleeves and do some real PITA work to compete with the guys who don't mind going to assembly from the start.
Facebook
Twitter