All Channels
Popular
C++ & ASM dev. Lover of cache hits. Loather of pipeline stalls. Data-oriented design loon!

Anonagrog

Contributor
CRank: 5Score: 34520

I find it intriguing that this developer would consider ram size the biggest bottleneck, especially when you consider how much smaller the working-set of every game will be than the amount of available ram in either console. A lot of the ram will be used for caching assets and keeping a hold of persistent game state data.

Now, if they were referring to memory bandwidth, latencies, or maybe stalls relating to memory access contention, I'd go with that wholeheartedly.

4549d ago 0 agree0 disagreeView comment

Impressive looking piece of middleware. I look forward to finding out more on the innards of their pipeline.

4559d ago 0 agree1 disagreeView comment

Recognizing the PS3 disc doesn't mean there's the capability to run code compiled for a completely different ISA, not to mention the interface with external systems like the console OS, along with code design targeting hardware specifics of the PS3's unique architecture (e.g. The Cell PPE's assembly isn't the same as Cell SPU assembly).

There's a world of difference here.

4583d ago 0 agree0 disagreeView comment

Using an engine like Frostbite 3 only means the designers and artists have an easier time bringing their vision to reality through all the supporting FB3 tools. Whether their "vision" of what Battlefront should be is going to be any good or not to gamers is another matter.

4585d ago 1 agree0 disagreeView comment

Bioware didn't use Unreal for Dragon Age 1 and 2, though. They used their own in-house engine.

4594d ago 0 agree1 disagreeView comment

I've always been impressed with Umbra's occlusion tech. Their latest version of the Umbra api looks fantastic. I'm glad to see their future is still looking bright. Good stuff.

4600d ago 0 agree0 disagreeView comment

Do you mean 'Project Titan'?

If so, they've pretty much gone back to the drawing-board with that one. It was mentioned a while back.

4600d ago 0 agree0 disagreeView comment

Edit: Sorry, had a submission error.

4608d ago 0 agree0 disagreeView comment

Edit: Sorry, had a submission error.

4608d ago 0 agree0 disagreeView comment

This title is a little misleading, no?

Sure, they may be observed here, but it's diminishing returns for this particular use case - It's not diminishing returns w.r.t the benefits of multicore processing in general.

The CryEngine threading model; the overall parallelizability of the code-base; gpu frame render times; the amount of general purpose calculations that need to be performed per frame.

Now, all of these things bring about th...

4608d ago 0 agree0 disagreeView comment

Yeah, just two people. Both of them coming from senior positions at Futuremark, so they know about quality graphics and presentation. We just need more game-play stuff on top of this.

4611d ago 1 agree0 disagreeView comment

The overall problem we have here is that of buffer swapping occurring during a screen refresh, rather than in between them. The current method employed treats the graphics device and display device as two disjoint piece of kit that have to work around the minimal co-ordination they have with each other. We need to get them working in harmony.

As nice as it is to be able to dynamically scale down the frame buffer resolution, it's not really a competing alternative to the m...

4617d ago 1 agree0 disagreeView comment

As far as I'm aware Remedy are using their own in-house engine, 'North Light'.

4617d ago 0 agree0 disagreeView comment

It's not just the ICE team involved with that. There's a 'tools & technology' group somewhere else in the U.S., though I don't know much detail on them, and there's Sony's 'Advanced Technology Group' in London looking at a lot of R&D.

Microsoft also have their own 'Xbox Advanced Technology Group' based out of Redmond. They do a similar thing to Sony's teams.

4621d ago 2 agree0 disagreeView comment

The fact that crunch still occurs in the industry as it is today is really nothing to joke about. I've read cases of devs working up to 100 hours a week for several months. A lot of overtime goes unpaid too. It's just ridiculous.

I've seen a small handful of dev's out there that actually advertise jobs with themselves through "x days since last crunch" ads. I think one of those might have actually been the late, great, Blitz Games. :(

4624d ago 1 agree0 disagreeView comment

I've been following the development blog ever since their debut trailer last year. It's well worth a look for those interested, as there's some great info on their rendering pipeline and graphics engine, 'Praxis', to read about:- www.reset-game.net

It's only a 2-person strong team, both coming from senior programming and design positions at Futuremark (the folks behind 3DMark and PCMark, etc).

On the technical side we know it will be i...

4631d ago 0 agree0 disagreeView comment

P0werVR, the console api's will already provide the "huge benefit" you refer to, along with full access to all hardware specifics of either system. Trying to reduce the abstraction penalty is exactly the point here.

4636d ago 0 agree0 disagreeView comment

GTAV wasn't written from scratch. It's built on top of 'RAGE', Rockstar's proprietary engine. Seeing as it's maintained by their own separate technology group and was already made dx11 capable for Max Payne 3's pc release, I'd say expecting dx11 isn't much of a leap at all.

4639d ago 0 agree0 disagreeView comment

Don't let the word "code" confuse here. This is essentially the build log we're looking at. It's nothing to do with the actual game and engine code itself, bar the odd reference to functions by name.

Also, r1sh12, the ps3 has never used C sharp. C sharp is Microsoft's baby for one thing, and secondly it's managed code. The language of choice for all the consoles is native C/C++.

4639d ago 0 agree0 disagreeView comment

We take the nominal visual acuity of a person with nominal eyesight at something like 2 minutes of arc, right? I think that's right...

Anyways, for a screen size of 40" (2160x3840), we're looking at a square pixel size of 0.0090789".

The ideal viewing distance to get a minimum of 2 arcmins between two adjacent pixels in a 40" 4K works out to be roughly around 16 inches.

So, exaggerated viewing figures aside, a 40" ...

4644d ago 1 agree0 disagreeView comment