NGGP: NextGen Garbage Pile (aka: No one reads the topics or stays on topic) *spawn*

Status
Not open for further replies.
He doesn't say that at all...

We could see one console have a 50% GPU flop count advantage and yet the rival excel in other areas, or the more flop console be gimped by a hardware bottleneck.

This is the point. Although PS2 was slower in most areas compared to the Xbox, it did however make up for it with a MASSIVE bandwidth advantage (an area, in which the Xbox was bottlenecked). This put both consoles much closer together than one might assume (plus the fact that most development actually took place on PS2 in the first place).

It's similar today. PS3 and 360... most people will tell you, that the 360s GPU is leaps and bounds faster than RSX. But most games don't look much different (especially today). There's the odd case of lower fps (on either side, mind you) or missing stuff (shadows, AO...). But overall. the differences are, imho, pretty negligible (unless they heavily influence gameplay, like latency etc, but that usually is not the case, except for RE5 and others).
 
So by that logic all consoles are essentially equal?
No. How do read "a strength in one area may be accompanied by a weakness in another" as "all consoles are equal"? It's the inequalities that make it so hard to determine a total 'most powerful' metric for comparison. 'Better' requires a clear target for comparison. eg. In aggregate, XB had loads more processing power than PS2. However, PS2 was way better at particles than XB, so in terms of pushing particles, PS2 was the more powerful console.

If we consider the two hypothetical extremes - one console with stupid amounts of processing power and a tiddly amount of bandwidth, and the other with a stupid amount of bandwidth and a tiddly amount of processing power - which is the more powerful? When it comes to putting graphics on screen, both are rubbish. Put a moderate, balanced console in the comparison and with moderate processing power and moderate bandwidth, it'll be the most powerful console.

If PS4 and XB3 have exactly the same CPU and GPU and PS4's GPU is 1.8TF and XB3's is 1.2TF measured the same way, which seems to be the thinking behind this thread, either console could have design (dis)advantages that mean better or worse results on screen.

It is impossible for us to determine 'most powerful console' by any meaningful metric. An attempt to categorise the consoles would need extensive investigation into the many factors involved and an aggregate score that in itself could be subjective depending on whether the investigators decide to weight certain factors as more important than others or not. This metric may give forum goers a number to talk about but it would mean very little to game players and developers and anyone who cares about more than just comparing numbers. I understand comparison is human nature, but there are times when it's really not worth bothering with - especially when some people get so emotionally attached to their numbers of choice and have big barnies over irrelevant information.
 
Maybe the criteria for 'more powerful' in the context of these discussions can be taken to be whichever console will have the better looking/performing multiplat titles and the most technically impressive exclusives.

Or alternatively, while each console may have their own advantages over the other the 'more powerful' console is the one whose strengths and weaknesses of are best suited to the types of games most devs are making for these consoles and which occupy the most market share (eg FPSs, third person action games/RPGs, sports & racing titles etc).

So, by that criterion the Xbox would be 'more powerful' generally than the PS2; since it's advantages in memory, processing power, programmable shaders etc are generally more useful than PS2's advantages with things like particles. As most games will benefit more from the former than the latter (which'd be more important for titles like Super Stardust or Geometry Wars)
 
Last edited by a moderator:
eg. XB360 has a considerable advantage in most games regards grass rendering, leaving PS3 looking fairly barren. It's not enough to make a difference that'll change people's buying choice, but it can be seen in games.

Do you have any examples of this? I have both systems and I can't say that I've seen this. I compare games between the two of them pretty often, so I'd like to see what I could compare that'd show this. :)

Thanks in advance.
 
Do you have any examples of this? I have both systems and I can't say that I've seen this. I compare games between the two of them pretty often, so I'd like to see what I could compare that'd show this. :)

Thanks in advance.
He can't be including Flower. What was it? 100,000 blades of grass individually rendered?
 
I would say difference is bigger than 10% for sure (at least from what I've read). Especially when it comes to vertex processing.
Apparently, it's a lot more when it comes to vertex processing. I think that's why games like Uncharted 2 and 3 did vertex processing on SPUs. SPUs,together (correct me if I'm wrong), were equal or better at vertex processing than Xenos.
 
Apparently, it's a lot more when it comes to vertex processing. I think that's why games like Uncharted 2 and 3 did vertex processing on SPUs. SPUs,together (correct me if I'm wrong), were equal or better at vertex processing than Xenos.
I think you mean vertex culling, yes. SPUs cull unnecessary triangles before RSX puts them on the screen. How much time it takes on SPUs I don't know (someone said ~20-30% on UC2) and not sure how it compares to Xenos. That would be question to ask Joker :LOL:
 
RDR and Mafia 2 comes to mind.

Mafia 2 on ps3 imho isn't exactly an example of a good porting (an anonimous company , 2k even touch it) so, we can blame ps3 hardware that much... about RDR is not that terrible, but R is really grow on ps3, I have in mind max payne 3... I'll bet GTA V could be even 720 this time thanks to FXAA...
 
I think you mean vertex culling, yes. SPUs cull unnecessary triangles before RSX puts them on the screen. How much time it takes on SPUs I don't know (someone said ~20-30% on UC2) and not sure how it compares to Xenos. That would be question to ask Joker :LOL:

No. According to the developer slides, all vertex processing was on the SPUs (5 of them, I believe). From an earlier thread (about RSX abilities), it was said that 2 SPUs would effectively double RSX vertex processing capability.
 
Do you have any examples of this? I have both systems and I can't say that I've seen this. I compare games between the two of them pretty often, so I'd like to see what I could compare that'd show this. :)
RDR was a very obvious one. Skyrim has more foliage on the trees. Dirt 3 has a bit more grass on XB360, looking at some screenshots. It also comes up in particles and smoke at times, but I don't keep a list. ;) This shouldn't be a surprise - XB360 has a massive alpha advantage whereas PS3 lacks BW for lots of alpha drawing.
 
No. How do read "a strength in one area may be accompanied by a weakness in another" as "all consoles are equal"? It's the inequalities that make it so hard to determine a total 'most powerful' metric for comparison. 'Better' requires a clear target for comparison. eg. In aggregate, XB had loads more processing power than PS2. However, PS2 was way better at particles than XB, so in terms of pushing particles, PS2 was the more powerful console.

If we consider the two hypothetical extremes - one console with stupid amounts of processing power and a tiddly amount of bandwidth, and the other with a stupid amount of bandwidth and a tiddly amount of processing power - which is the more powerful? When it comes to putting graphics on screen, both are rubbish. Put a moderate, balanced console in the comparison and with moderate processing power and moderate bandwidth, it'll be the most powerful console.

If PS4 and XB3 have exactly the same CPU and GPU and PS4's GPU is 1.8TF and XB3's is 1.2TF measured the same way, which seems to be the thinking behind this thread, either console could have design (dis)advantages that mean better or worse results on screen.

It is impossible for us to determine 'most powerful console' by any meaningful metric. An attempt to categorise the consoles would need extensive investigation into the many factors involved and an aggregate score that in itself could be subjective depending on whether the investigators decide to weight certain factors as more important than others or not. This metric may give forum goers a number to talk about but it would mean very little to game players and developers and anyone who cares about more than just comparing numbers. I understand comparison is human nature, but there are times when it's really not worth bothering with - especially when some people get so emotionally attached to their numbers of choice and have big barnies over irrelevant information.

Really this should have been the first reply before locking this thread. I didn't even know my prior post was moved to this thread since I hate these types of pointless discussions.

He can't be including Flower. What was it? 100,000 blades of grass individually rendered?

How Flower renders the grass is different than how a game such as RDR renders grass through the use of alpha textures (as damien mentions). Flower can get away with rendering grass that way because that's basically all it's rendering, along with the flowers.

What does a PS3 exclusive have anything to do with his point anyways? That's a rhetorical question btw, I don't care for an answer since I already know it.

Apparently, it's a lot more when it comes to vertex processing. I think that's why games like Uncharted 2 and 3 did vertex processing on SPUs. SPUs,together (correct me if I'm wrong), were equal or better at vertex processing than Xenos.

You're wrong. Forgot who here clarified it before, but no matter what's being done on the Cell, RSX still has to be set up and rendered the triangles, so the system is still bound by the limitations of the GPU.

Mafia 2 on ps3 imho isn't exactly an example of a good porting (an anonimous company , 2k even touch it) so, we can blame ps3 hardware that much... about RDR is not that terrible, but R is really grow on ps3, I have in mind max payne 3... I'll bet GTA V could be even 720 this time thanks to FXAA...

I think you may be missing Shifty's point with this post. Look at the differences between RDR and MP3 and how well the ps3 is equipped to handle both games. We have one game with open landscapes filled with alpha textures (grass), this leaves little oppurtunity to cull geometry and the grass is just the thing to eat up fillrate. While the other game is more linear and isn't filled with foliage. Which do you think would work better on the PS3?

I'm not saying GTA5 won't turn out better than GTA4 on the PS3, but you looking at MP3 as an example of how they handled RDR isn't the most accurate way of looking at things IMO.

If I'm wrong on any of this, would like to hear from someone better informed than myself.
 
Last edited by a moderator:
So in which games is the PS3 worse or different with AI or physics?

Well... as he was saying "PS3 is missing out on *better* Physics, AI", since most multiplatform games target the lowest common denominator here.

Though judging "better AI" is quite hard. Physics on the other hand, can usually be "seen". And there are several PS3 games which showcase the advantages of PS3. Case in point Motorstorm Apocalypse or Pacific Rift. Those two games are a real treat, and they even use Havoc as their physics engine, not some PS3 specific engine. But even then, it's not "worlds" better, it's just very good.
 
kagemaru said:
Curious to see if MS do over compensate here since the CPU also seemed like a weakness in the original xbox
You could say that for every console that generation - but it was least problematic in XBox1, so hardly standout among other issues.
The main complaint was UMA bandwith - and there's a clear case of (over)compensating for it with 360.
 
You could say that for every console that generation - but it was least problematic in XBox1, so hardly standout among other issues.
The main complaint was UMA bandwith - and there's a clear case of (over)compensating for it with 360.

Good point, didn't even think about the bandwidth. :smile:
 
Status
Not open for further replies.
Back
Top