Why are certain companies having so much trouble with the PS3 Hardware?

[Ssliasil]

priteshmodi said:
http://www.eurogamer.net/videos/digital-foundry-killzone-2-480p-vs-576p-vs-720p-performance-analysis

It runs at 30 FPS. This was posted less than a week ago. Whether or not you see Eurogamer as a reliable source is up to you. I'm thinking you won't though by the sound of your responses so far.

@ OP: Your question has already been answered. The games that you listed are first party. They benefit from getting help internally from people who designed the machine as well as other internal studios. They also only have to worry about the restrictions presented by one set of hardware.

When developing an engine for multiple platforms you have to take the limitations of each platform into consideration. Unfortunately this gen the limitations of each platform don't make it easy for multiplat development. For example:

The PS3 has 256 MB of memory dedicated to the CPU and 256 MB to the GPU. The speed and architecture of these memory modules isn't generic either making it that much more complicated. In order to bypass memory restrictions the first party studios who don't need to worry about another platform can use this to their advantage. They can split the visual processing work between the SPU's and the GPU and split memory usage in that manner.

On the 360 the memory is shared between the GPU and CPU and is one big lump of 512 MB. This way the amount used by the GPU or CPU can be greater than 256 MB at any given time as long as the other part isn't utilizing much memory.

If a developer is developing for both of these platforms, these two architectures don't make the job easy.

If the PS3 is used as the lead platform, the 360 will have problems scaling to how the code is utilizing the hardware. Special attention will have to be put into this area so that the visual processing work is mainly done by the GPU for the 360 and not offloaded to the processor (which wouldn't be a good use of the 360's general processing CPU).

On the flipside, using the 360 as a lead platform will cause problems (or extra work) on the PS3 as it doesn't have the memory flexibility between the CPU and GPU that the 360 does. A game could be using 300 MB of memory for the GPU on the 360 but the PS3 will only have 256 MB readily available for its GPU. The devs will need to put extra work into offloading some of that work to the SPU's.

The problem with this extra work and research is that it costs money. This isn't the only difference between the two architectures. I just think it's the easiest one to explain to answer your question. The devs aren't lazy, they simply don't have the resources or time given the publisher's demands. That's another restriction that the first party studio's might not have (or have less of).

While thats a very good answer to the original question, how about the Second question?

 

What are they going to do Next Gen when PS3 style Tech is the medium?