Predict: The Next Generation Console Tech

Status
Not open for further replies.
What was the timeline for XDR before PS3? iirc it was not a volume part either, in fact iirc Cell's ring was designed with Rambus/XDR spec in view and demoed at the ISSCC in 2005 and didn't hit production until about the time of the PS3 and that was one of its very few uses? I am foggy on all the details... I should wiki after dinner as I may be remembering wrong.
 
Real money ?

Compare the outlook for Blockbuster to the outlook for iTunes, Zune, NetFlix, Vudu and others.

Oh, and I have a DVD player, one that plays my SACDs too, - another dead physical format.

Cheers

Comparing SACD to Blu-Ray is just stupid, it´s a Sony only format that only lives/lived among total high-end nerds. Compare it to HD-DVD and you have a point.

I too, have a large collection of physical media, around 250 CDs and 200 DVDs. Of all my DVD movies, there are only a dozen or so I wanted to see more than once: Team America, Shaun of the Dead, The Bourne trilogy, LOTR, Blade Runner and Dr. Strangelove. Most movies are just forgettable crap; did I rent Thor ? Hell yes, do I ever want to see it again, hell no.aa

That is a pretty small collection :) and you bought it yourself, so why even begin to complaining about only seeing few DVD´s more than once?

I am sure that the streaming services will "win" at some point, but right now the main competitor for streaming is rental and not physical sales.

Instead of using personal anecdotes why not find something that shows us how HD streaming is cutting in Blu-Ray sales?

v611u1.jpg


I see an excellent point from Microsoft if they want a "complete" Windows 8 and Live experience, with a Console playing Live games, that essentially could play on a PC and Tablet, but imho they will be competing more with Apple than a power hungry PS4..
 
What was the timeline for XDR before PS3? iirc it was not a volume part either, in fact iirc Cell's ring was designed with Rambus/XDR spec in view and demoed at the ISSCC in 2005 and didn't hit production until about the time of the PS3 and that was one of its very few uses? I am foggy on all the details... I should wiki after dinner as I may be remembering wrong.


gddr3/5 were too slow for the cell processor. The current XDR memory runs at full cpu speed @3.2ghz
 
gddr3/5 were too slow for the cell processor.

Rubbish. First of all, GDDR5 is irrelevant because A) it wasn't out at the time (fast forward several years), and B) GDDR5 provides a heck of a lot more bandwidth in practise.

By the end of 2006, GDDR3 already provided comparable bandwidth in practise.

Stop making things up.

The current XDR memory runs at full cpu speed @3.2ghz
What does this have to do with anything? The bandwidth figure is all that matters in this instance. The clock speed is just a means to achieving the bandwidth.

Please, just drop the silliness. The thread ought to move on from this.

in fact iirc Cell's ring was designed with Rambus/XDR spec in view and demoed at the ISSCC in 2005 and didn't hit production until about the time of the PS3 and that was one of its very few uses?

What might have happened was that they weren't betting on conventional memory tech to increase in bandwidth to a similar degree; recall it was mostly only DDR1 and DDR2 circa 2001-2005 with GDDR3 appearing late 2005, and by then it would have been too late as they were fully committed. So by the time they were done, things changed...
 
From reading around it doesn't look like XDR was in significant production before the PS3 and was only produced by a few companies (Elpida and Qimonda, others?) for Rambus/Sony. Rambus is fabless and pretty much acts like NV, ARM chip makers, former ATI and now AMD, etc and contracts space.

If XDR1 was essentially only used in 1 product (the PS3) I don't see why a company, if they so wished, could not go the same route with XDR2 and get a fab to produce it for them. I am not saying it is the best option but XDR2 was announced in 2005, the spec was established in spring 2008. I am sure Rambus has working samples.

So other than it not being a mainstream memory part (which is a concern for cost) is there any other reason to believe it isn't a viable part to produce for future consoles? i.e. What is different about the current situation versus the PS3/XDR1?

Edit: XDR2 7.2GHz samples in January 2009 with production available April 2009 from the now brankrupt Elipda. So not so much vaporware as an issue of no demand? Which moving over to a low-volume part that would require, say, a GPU maker to create a totally new memory interface and XDR2 while (my understanding) is they have theoretically higher bandwidth lose a bit of potential due to the signally (correct?) so I am not argueing for XDR2, just trying to understand the negativity toward availability in contrast to 2006/XDR1. Looks like it could be made and deployed if a party was interested.
 
Just hypothetically speaking, what do you think the cost would be for a part that is fabbed just for said console versus conventional memory that is already supported by the rest of the industry?

edit:

Anyways, you might want to consider the practical size of the memory bus interface. Differential signalling requires 2n wires per bit. Single-ended is n+1. The 64-bit XDR I/O on Cell is pretty large, and perhaps ballpark comparable to 128-bit GDDR3. So... I mean it's great that XDR has higher per pin bandwidth, but um...yeah... do you see where I'm going with this? (double b/w per pin, but double the physical size...)
 
What was the timeline for XDR before PS3? iirc it was not a volume part either, in fact iirc Cell's ring was designed with Rambus/XDR spec in view and demoed at the ISSCC in 2005 and didn't hit production until about the time of the PS3 and that was one of its very few uses? I am foggy on all the details... I should wiki after dinner as I may be remembering wrong.

Toshiba started sampling XDR in 2003.
 
Just hypothetically speaking, what do you think the cost would be for a part that is fabbed just for said console versus conventional memory that is already supported by the rest of the industry?

edit:

Anyways, you might want to consider the practical size of the memory bus interface. Differential signalling requires 2n wires per bit. Single-ended is n+1. The 64-bit XDR I/O on Cell is pretty large, and perhaps ballpark comparable to 128-bit GDDR3. So... I mean it's great that XDR has higher per pin bandwidth, but um...yeah... do you see where I'm going with this? (double b/w per pin, but double the physical size...)


going to gddr memory would be a step back for sony. when they are faster more efficient ones out there. who gives a damn about the industry? these same people are the ones who are happy for sony to have mediocre generic pc specs for consoles. screw that.

sony has been working with rambus for all these years. they have deals and licenses in place. Better to work with partners that served you well rather than going with generic stuff like your competitors who are not even hardware manufactuerers in the first place


i hope sony continues with their hardware ethos and make trail blazing stuff that lasts till 2020. 8 gigs of superfast XDR2 ram is the way to go.
 
Last edited by a moderator:
that's what they said about xdr too. i doubt sony cares who samples what. they went with xdr for the ps3 because other memory options didnt have the bandwidth the cell needed

Bandwidth? The GDDR-3 in the PS3 had 87.5% of the bandwidth of the XDR. Clocked a little higher it could have easily supplied the bandwidth. High end video cards of the time using GDDR-3 has 2.5x the bandwidth of the XDR in the PS3. These days GDDR-5 on high end video cards delivers over 10x the bandwidth of the XDR in the PS3.

The reason for pairing Cell with XDR was lower latency, not bandwidth. You should try calculating actual bandwidth instead of just looking at frequencies.
 
xdr2 is available if sony chooses to do so. rambus and sony have been working together for years. xdr is not smoke and mirrors. it's the only memory fast enough for the current ps3 processor.

gddr3/5 are simply not fast enough

:???: GDDR5 peaks out today at around 96/GB/s on a 128bit bus which is similarly sized to the bus used in PS3 to XDR.

Would you mind explaining in what way that's "simply not fast enough" compared to the 25.6GB/s currently feeding Cell in the PS3?

Incidentally there were PC graphics cards shipping before PS3 launched running GDDR3 that would have achieved 27.2GB/s on a 128bit bus.
 
these same people are the ones who are happy for sony to have mediocre generic pc specs for consoles. screw that.

How do you define "mediocre generic pc specs"? Regardless of whether it's PC derived or not, if the PS4 were to match the specs of a top end PC from today (nevermind a year or two from now) you should consider yourself very lucky. Odds are it will not.
 
MS must really hate used games to go solid state.
Can't really argue with them considering it has worse effect for devs and themselves than pirating.
Instead of using personal anecdotes why not find something that shows us how HD streaming is cutting in Blu-Ray sales?
You do realize that graph shows nothing about how streaming affects BD sales, right? It just tells us that BD gets adopted slowly over DVDs.
Bandwidth? The GDDR-3 in the PS3 had 87.5% of the bandwidth of the XDR. Clocked a little higher it could have easily supplied the bandwidth. High end video cards of the time using GDDR-3 has 2.5x the bandwidth of the XDR in the PS3. These days GDDR-5 on high end video cards delivers over 10x the bandwidth of the XDR in the PS3.
All that using as many lanes on PCB as PS3 did for XDR? Also, what's the point in comparing XDR1 that's years old and (comparatively) brand new GDDR5 when XDR2 is (theoreically) availiable?

Also, what would be the bandwidth of XDR2 when using as many lanes as 128bit GDDR5? If I were to believe Rambus own data then i would be providing slightly more than twice the bandwidth while using ~30% less lanes.
 
You do realize that graph shows nothing about how streaming affects BD sales, right? It just tells us that BD gets adopted slowly over DVDs.
.

I am not the one claiming that Blu-Ray is as dead as SACD, i am just showing how it´s taken over the physical market from DVD´s. The idea that personal use and "my friends" usage pattern can explain why it´s a good idea to drop Optical from the next XBOX is classic, there is clear tendency that Blu-Ray is taking over the Physical market, so someone friends must be watching them.

And the streaming vs Physical media is even more classic. It´s always been rental vs streaming, why should people that usually buy movies stop doing that just because they can stream? Of course there will be a good amount of people that take the couch approach and just streams/rent the movies instead of buying them, but unless someone can come up with some solid numbers on how does that it will be us guessing.

Anyway, people adjust their arguments to fit their idea of being right.. "i don´t need a blu-ray play i got a standalone", "i need a console with a blu-ray player it saves me space and it´s cheap" etc etc

On this forum it was very clear with the PS3, there was people claiming that people didn´t buy consoles to watch movies (HD-DVD vs Blu-Ray and how the PS3 didn´t matter), and people that claimed the reason the software sales on the PS3 was so low was that most bought it just to watch movies (attach rate discussion).. and i think a few people made both claims :)
 
I am not the one claiming that Blu-Ray is as dead as SACD, i am just showing how it´s taken over the physical market from DVD´s
So having quarter of the market share qualifies as "has taken over" then? :)

Don't get me wrong. I also think that BD is here to stay for quite a while and is far from dead. Just that I'm not sure if it'll dominate the movie (game?) distribution nearly as much as DVD did/does.
 
Status
Not open for further replies.
Back
Top