PlayStation 4 (codename Orbis) technical hardware investigation (news and rumours)

Status
Not open for further replies.
If that really is the long term plan, then you shouldn't buy any of your games on physical disks. Gaikai will be the carrot to go all digital.

Only if it changes completely - the way it currently is it would never replace anything, while it serves other things too, it's still "main point" is still streaming games which will never replace anything
 
Tessellation should be no different as it doesn't increase the theoretical triangle rate limit.


A Sony specific feature won't be put in a Microsoft console and vice versa, but that doesn't mean useful features won't make their way to the PC. Some things like security features will remain exclusive and others are determined on a case by case basis.

That very interesting. Were these the changes you were talking about?

The three "major modifications" Sony did to the architecture to support this vision are as follows, in Cerny's words:

"First, we added another bus to the GPU that allows it to read directly from system memory or write directly to system memory, bypassing its own L1 and L2 caches. As a result, if the data that's being passed back and forth between CPU and GPU is small, you don't have issues with synchronization between them anymore. And by small, I just mean small in next-gen terms. We can pass almost 20 gigabytes a second down that bus. That's not very small in today’s terms -- it’s larger than the PCIe on most PCs!

"Next, to support the case where you want to use the GPU L2 cache simultaneously for both graphics processing and asynchronous compute, we have added a bit in the tags of the cache lines, we call it the 'volatile' bit. You can then selectively mark all accesses by compute as 'volatile,' and when it's time for compute to read from system memory, it can invalidate, selectively, the lines it uses in the L2. When it comes time to write back the results, it can write back selectively the lines that it uses. This innovation allows compute to use the GPU L2 cache and perform the required operations without significantly impacting the graphics operations going on at the same time -- in other words, it radically reduces the overhead of running compute and graphics together on the GPU."

Thirdly, said Cerny, "The original AMD GCN architecture allowed for one source of graphics commands, and two sources of compute commands. For PS4, we’ve worked with AMD to increase the limit to 64 sources of compute commands -- the idea is if you have some asynchronous compute you want to perform, you put commands in one of these 64 queues, and then there are multiple levels of arbitration in the hardware to determine what runs, how it runs, and when it runs, alongside the graphics that's in the system."

http://www.gamasutra.com/view/feature/191007/inside_the_playstation_4_with_mark_.php?page=2

Seem everyone of these are now in amd chips.
 
This sounds like a plausible theory. And seeing how small the PS4 is, now that according to Mr Fox the 22nm Cell is about to become a reality, they could attach the PS3 motherboard to the PS4 and create an entire new box from it. It's not like they don't have experience with a similar approach.

This is something that I too was thinking of for backwards compatibility... Instead of shipping with a big box at a high price with both PS3 and PS4 hardware, they could add b/c later on in a premium model, when they do a price decrease for the basic model. This would be a much better approach compared to the PS3, where they had to remove B/C later down the road to be competitive. Native B/C later down PS4's lifecycle would be much more valuable and may even significantly boast sales even from people who already own PS4's.

Of course they would be tight lipped about it as an anticipation for a B/C model could cannibalize early sales of the system.
 
....The 22nm cell is almost confirmed, and there's a rumor of an RSX at 28nm...

Intel are the only people fabbing at 22nm right now and AFAIK no-one else has even begun fab construction for 22nm. This is a big part of the reason why Intel can out perform everybody, they spend billions annually to be at least 18-24 months ahead of everybody else in fab tech.
 
Intel are the only people fabbing at 22nm right now and AFAIK no-one else has even begun fab construction for 22nm. This is a big part of the reason why Intel can out perform everybody, they spend billions annually to be at least 18-24 months ahead of everybody else in fab tech.

ummm..... clue bat please.....................
 
This is something that I too was thinking of for backwards compatibility... Instead of shipping with a big box at a high price with both PS3 and PS4 hardware, they could add b/c later on in a premium model, when they do a price decrease for the basic model. This would be a much better approach compared to the PS3, where they had to remove B/C later down the road to be competitive. Native B/C later down PS4's lifecycle would be much more valuable and may even significantly boast sales even from people who already own PS4's.

Of course they would be tight lipped about it as an anticipation for a B/C model could cannibalize early sales of the system.

IMO the utility of BC is largely early in the cycle when software is somewhat limited, the initial launch games aren't likely to be as good AAA titles released for 360 and PS3 this holiday. A year from now that starts to change as the systems get more variety.
 
IMO the utility of BC is largely early in the cycle when software is somewhat limited, the initial launch games aren't likely to be as good AAA titles released for 360 and PS3 this holiday. A year from now that starts to change as the systems get more variety.

Good point. Still to me, tho, early adopters of next gen consoles are the ones that probably who already have current gen consoles, and those early adopters are probably targeting next-gen games. It would be best if Sony came up with a premium model with B/C and one without at the beginning of the console cycle, but if they are indeed planning to do it later, it's best for them [at least understandble of them] to not mention it as I said it would cause more people to wait for the B/C model.
 
Last edited by a moderator:
Good point. Still to me, tho, early adopters of next gen consoles are the ones that probably who already have current gen consoles, and those early adopters are probably targeting next-gen games. It would be best if Sony came up with a premium model with B/C and one without at the beginning of the console cycle, but if they are indeed planning to do it later, it's best for them to not mention it as I said it would cause more people to wait for the B/C model.

Several generations ago I would have agreed with you but due to Nintendo's success with getting people to buy the same games multiple times Sony and MS are starting to work on HD do-overs which allow them to cash in a second time which makes them a lot more money than they ever will with hardware. Console hardware is sold at loss or breakeven so why would Sony or MS want to give us a new console with backwards compatibility and lose the opportunity to sell us HD copies of older games? Its not like they can sell BC at any sort of a premium which would justify the investment in engineering and actually make them money.

To put it another way if BC is a must buy a PC, you'll be much happier in the long run.
 
ummm..... clue bat please.....................

What? My point is that 22nm Cell is a hell of a long way off if it ever materialises at all from Sony or their fab partners. Doing a bit more research it seems TSMC are going with 20nm and have apparently inked a deal with Apple for the 'A8' for late 2013/2014 although they seem to have only one process at this node which is damn odd given they have 4 at the 28nm node. Global Foundries are offering 20nm later this year but both companies are offering planar processes saving FinFET for the next process node.

Suggesting that Sony are going to drop a ton of cash to spin up 20nm Cell + RSX for a 'Pro PS4' with b/c is just nuts, it would be far cheaper to just drop a voucher for a PS3 into the PS4 box. After the launch window no-one cares about b/c, I appreciated it on my launch phat PS3 but when it died two years later and I got a refurb slim without b/c I was past caring.
 
If the specific node is 22nm and not 20, the closer match for Cell would be IBM's 22nm PD-SOI process, which they disclosed the outlines of at the end of last year.

At least the use of PD-SOI and hints in IBM's involvement in the shrink of its other PD-SOI console processor show a possible path of lesser resistance than involving the bulk 20nm nodes.
 
The slides indicate 22nm is the last node PD-SOI will be used for. The rest of the slides discuss FinFETs, mostly in conjunction with fully depleted SOI.
 
Richard needs to "talk" to Blow to get this confirm. Pretty major difference if this is all true.

Guess number could go down over the life of the console. Still even at 1GB is a ton of ram compare the < 60MB they use on the ps3.

60mb is A TON considering the PS3 only had 256mb to begin with that's nearly 25% of the ram being used. Considering the PS4 is more complex and it will allow you to resume playing right when you turn on the console and chat and watch your friends play 1gb is not a lot. 1gb/8 is only 12.5%.
 
60mb is A TON considering the PS3 only had 256mb to begin with that's nearly 25% of the ram being used. Considering the PS4 is more complex and it will allow you to resume playing right when you turn on the console and chat and watch your friends play 1gb is not a lot. 1gb/8 is only 12.5%.

The correct number is 512mb. That 60mb was not just system ram.
 
About the FreeBSD used in the Ps4

Why 9.0 if 9.1 has good improvements like AVX instructions.

And about filesystems ? Maybe it will use the default UFS
In many benchmarks, ZFS not only is more advanced it is also faster.

I remember how many people and close friends had trouble with data corruption with the UFS2 used on the Ps3.
 
Status
Not open for further replies.
Back
Top