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

Status
Not open for further replies.
Weeks ago it was like Durango > Orbis, and now it is like Orbis >>>> Durango... lol
Well yeah, thats corporate strategy. I imagine MS engineers must be pissed of building something like this. Having very tight budgets to spend on graphics, but still expecting them to deliver. That was the reason why Ed Fries left MS (although he was disappointed with lack of 360 power, but thats kinds bogus).
 
Well yeah, thats corporate strategy. I imagine MS engineers must be pissed of building something like this. Having very tight budgets to spend on graphics, but still expecting them to deliver. That was the reason why Ed Fries left MS (although he was disappointed with lack of 360 power, but thats kinds bogus).

Budgets != BOM.

I assume they were given a blank check to deliver on high performance chips that has relatively low BOM to OG 360s.
 
Budgets != BOM.

I assume they were given a blank check to deliver on high performance chips that has relatively low BOM to OG 360s.
Look, if MS said "Deliver gaming box with 200-250W cap" they would probably do something like Orbis, possibly with more customized GPU.

When the higher ups tell you "We need 8GB of RAM and we are packing Kinect with it. Console should be 300$ and not a penny more". There is not much they can do. They know that they have to go with slow RAM. They know it will be insufficient for gaming so they have to go with embedded fast memory, but thats also expensive and takes quite a bit of space on die so there is not much room to go on GPU side. Obviously, there are hard limits and only reason for Durango's low compute performances is in MS higher up strategy. I have no reason to doubt these engineers at MS wouldn't do amazing job if it wasn't for that.

*They will probably do amazing job even the way its now, still, there is alot to be desired from Durango unfortunately.

That right there shows you don't really understand what is going on or will be put out.

The majority of games are likely to be 720p/30 fps. It's likely that some games that try to push graphically advanced techniques may even fall below 720p on PS4 (as well as durango). 1080p/30 fps is going to be rare. It is highly unlikely that any games will be released at 1080p/60 fps for either console unless it feature relatively simplistic graphics.

Most of the 1080p/30 fps games are likely to show up in the first 1-2 years of a console while 720p/30 fps will likely dominate the later stages of each consoles lifetime.

Regards,
SB
Actually I think there will be quite a few games that are 1080/60fps on PS4. At least in early year. There is no reason not to really. Majority of most demanding games on PC run at PS4 settings pretty good actually.
 
Please pardon this interruption. We apologize if any interesting posts were nuked reply to noise or garbage. If anyone sees noise or garbage posts, please report them as such and the staff will attempt cleanups as soon as possible.

You may now continue one with your console wars in a sane manner.
 
when i asked about the ram situation of orbis and durango to Timothy Lottes - http://timothylottes.blogspot.in/20...showComment=1359087365844#c721108568894812681

BBgfvf0CIAAdFET.png:large


leaving out the DME - orbis > durango
 
Last edited by a moderator:
Which setup of the two will be more preferable by the 3rd party developers?

leaving out the DME - orbis > durango

This assumption has a little flaw since 3rd party developers tend to (officially) like the system that allows them to make the most profit. If Durango outsells Orbis dramatically because of some fancy Kinect stuff that is packed with each console, then 3rd parties will most likely treat Durango as the lead platform of the next generation. The pure horror for Microsoft would probably be if a (graphically more potent) Orbis outsells Durango and becomes the lead platform for 3rd party game development. But that is very hypothetically speaking.

I'm also wondering how much resources Microsoft is willing to dedicate to Kinect, assuming that it will be an out of the box feature for the new console. This could even increase the gap between Orbis and Durango.
 
Not everybody likes kinect, or see it as an advantage, or something really all that interesting; me for example. Some people are also not going to be physically able to use games that utilize/require kinect, such as those with disabilities or very cramped homes.

An over-reliance on kinect isn't necessarily going to be a benefit for the platform, it could just as easily be a hindrance, not least because motion gaming is kinda difficult to work into many genres and could lead to gimmicky and dumbed-down gameplay.
 
This assumption has a little flaw since 3rd party developers tend to (officially) like the system that allows them to make the most profit. If Durango outsells Orbis dramatically because of some fancy Kinect stuff that is packed with each console, then 3rd parties will most likely treat Durango as the lead platform of the next generation. The pure horror for Microsoft would probably be if a (graphically more potent) Orbis outsells Durango and becomes the lead platform for 3rd party game development. But that is very hypothetically speaking.

I'm also wondering how much resources Microsoft is willing to dedicate to Kinect, assuming that it will be an out of the box feature for the new console. This could even increase the gap between Orbis and Durango.

He meant the ram setup of choice by the developers for their games ! It doesn't include which will be their lead platform .
 
He meant the ram setup of choice by the developers for their games ! It doesn't include which will be their lead platform .

I know what he meant. ;)

I just wanted to point out that there is a difference between "setup of choice" in terms of best graphics and "setup of choice" in terms of economic efficiency. Game developers don't have an infinite amount of resources and they most likely want to keep the business running. But since Sony is rumored to have a pretty decent SDK, Orbis is certainly going to give Durango a very hard time ...unless Microsoft is showing up with some wizzard jizz! :LOL:
 
He speaks about "Durango refresh" as of console that would replace lackluster performing box that Durango could likely end up being.
The problem with his entire blog article is that it's based upon the commonly speculated specs (he openly states that he doesn't have any concrete information - and uses the current speculation as a basis).

His conclusions do seem interesting and reasonable. But even he doesn't know whether the premises he bases his conclusions on are actually valid.
 
leaving out the DME - orbis > durango
Having now read the blog, I think he's way off regards Durango based on what we've heard. I can't see it being pre-GCN if Sony have a GCN architecture, and we've even had suggestion Durango is GCN2. Certainly, his view of it being a Windows box with all of Windows' issues seems pretty wild speculation. Take the opinions of Orbis's GPU possibilities as insightful, and then expect the same from Durango too, IMO.
 
Lottes' out of the Loop? ;)

:runaway:
This is being covered in two threads now (isn't every subject these days? :rolleyes:). He may have insider knowledge from AMD execs who defected in 2nd half of last year, and be using recent leaks to cover his insider knowledge in discussion. This means Durango is either GCN2, or VLIW, with rumours diametrically opposed.
 
He speaks about "Durango refresh" as of console that would replace lackluster performing box that Durango could likely end up being.

Ooh god i hope this isn't true. This would mean the end of getting ms console i can just as easily pump that money in my desktop. :cry:
 
This is being covered in two threads now (isn't every subject these days? :rolleyes:). He may have insider knowledge from AMD execs who defected in 2nd half of last year, and be using recent leaks to cover his insider knowledge in discussion. This means Durango is either GCN2, or VLIW, with rumours diametrically opposed.

If you believe the rumors of it being VLIW then it also must be IBM PPC. That's how the rumor is indicated from the article published.

So Durango is either IBM PPC and VLIW or it's AMD x86 and GCN2.
 
More from Tim:

If 720 gets say GCN2 and PS4 has GCN1? Then 720 needs some API (or lack of API) which exposes what is possible with GCN1 and what is made possible with GCN2. Keep in mind DX11.1 does not even come close to exposing what AMD's GCN GPUs can do. Given a choice between GCN1 GDDR5 and GCN2 with DDR3, the GDDR5 wins every time for me.

Also if one has almost 2x the flops as the other (and has the bandwidth required), then the one with 2x the flops will have a good advantage in pixel quality which translates into either better AA, higher frame-rate, or higher resolution. GPU feature flexibility mostly translates into the capacity for richer content, for instance more objects with different materials. Raw performance is still needed for pixel quality and triangle density per object.

If we go with some wild speculation like GCN2 is a full tiler GPU like PowerVR (taking the EDRAM path to it's logical extreme conclusion), and they have 32MB of on-chip RAM with massive bandwidth (much larger than 256 GB/s), then things like box-filtered MSAA with high resolutions is possible, but triangle density would take a major hit, because you'd need to store all the binned triangles for the scene in that 32MB.

DX11.1 is that much of a drag?
 
Just think..if 720 has gcn2 with advanced api above and beyond 11.1....with it being microsoft and they were smart enough that could very well happen.
 
This is being covered in two threads now (isn't every subject these days? :rolleyes:).
I think it's due to the fact that the *spawn* threads on here are getting out of hand. Having just one big thread with all the next-gen discussion was way more comfortable (even if the discussion got a little messy).

I love how 1.5x the flops almost invariably becomes "almost 2x the flops" and turns a 50% advantage into a 100% advantage.
Up until very recently, a 50% flop advantage was reduced to "no advantage at all" by a secret sauce that turned normal flops into magic flops ... so we're making some progress here ;)
 
Status
Not open for further replies.
Back
Top