Sony's Next Generation Portable unveiling - PSP2 in disguise

Could it be that the four cortex A9 NEON simd-s would be used also for some graphic efects like the cells SPE and RSX combo in PS3 ?
Or could it alow more geometry in games ?

At theoretical 1 GHz thats quite a lot of CPU power for a handheld. At least nintendo 3DS dual core 200MHz ARM11 sounds realy cheap compared to this.(unlike the price they ask for it ;))
 
As far as gaming goes- I have tried 3g once for fps on my laptop and it was horrible :D

I've been using 3G for multiplayer games like Lost Planet on X360 for the last 2 years without much problems...gotta have a strong and stable 3G signal though. With that said for PSP2 the data plan has to be unlimited and AFAIK none of the wireless carriers except Virgin Mobile offer unlimited 3G plans.
 
Even though they were planning for around 512 MB, they should really move that to a gigabyte considering phones will have already reached around half of NGP's performance by the time it launches.

NGP's inevitably superior environment for game development will keep its graphics looking a lot better for a while, yet the RAM would help to solidify that edge.
 
Even though they were planning for around 512 MB, they should really move that to a gigabyte considering phones will have already reached around half of NGP's performance by the time it launches.

NGP's inevitably superior environment for game development will keep its graphics looking a lot better for a while, yet the RAM would help to solidify that edge.

I wonder if the NGP is on the wrong side of a micron drop. If this is on 40nm now and chips move to 28nm early next year it may get eclipsed really fast
 
I wonder if the NGP is on the wrong side of a micron drop. If this is on 40nm now and chips move to 28nm early next year it may get eclipsed really fast

I was thinking the exact same thing the other day. I don't think it's plausible to expect a refresh chip (higher frequencies) for the NGP later down the road, but I'd like to stand corrected.

Assuming NGP is on 45/40nm, how much delay would it had meant for it, if it would had been laid out for 28nm from the start? A year maybe or even less?
 
I was thinking the exact same thing the other day. I don't think it's plausible to expect a refresh chip (higher frequencies) for the NGP later down the road, but I'd like to stand corrected.
What are the chances that Sony is trading off power consumption for higher frequency - well above 200 MHz - on 40 nm, to get the former back with a die shrink later on?
 
What are the chances that Sony is trading off power consumption for higher frequency - well above 200 MHz - on 40 nm, to get the former back with a die shrink later on?

I can't know from where I stand. If rumors are indicating as has been posted in this thread 200MHz then I consider it unlikely that they'll opt for a much higher frequency in the end.

Nathans Fortune mentioned something about 5-6 hours battery life for it; now I can't know with what frequency that could be, but if it's with 200MHz as the MP frequency, increasing frequency should mean that battery life will be affected by a significant margin.
 
I was thinking the exact same thing the other day. I don't think it's plausible to expect a refresh chip (higher frequencies) for the NGP later down the road, but I'd like to stand corrected.

Assuming NGP is on 45/40nm, how much delay would it had meant for it, if it would had been laid out for 28nm from the start? A year maybe or even less?

They'll just win back form factor and battery. NGP won't be eclipsed very fast, because there will be very few other devices that allow you to program to the metal. Eventually, they could use extra power availabilty to put in and power a 3D screen at a resolution that is compatible with the 2D version of the NGP.
 
No one mentioned it but I highly doubt, I don't think that it would be healthy for the battery life.

Maybe not a rumble feature for force feedback functions, but I consider the capability of having haptic feedback quite important (and stuff like e-mail\psn messaging notifications, there's 3G after all), even more with all the importance they're giving to touch interfaces this time.
 
They'll just win back form factor and battery. NGP won't be eclipsed very fast, because there will be very few other devices that allow you to program to the metal. Eventually, they could use extra power availabilty to put in and power a 3D screen at a resolution that is compatible with the 2D version of the NGP.

Well here comes the trick question the topic is actually about: assume NGP could be eclipsed within theoretical time-frame N (under 40/45nm); wouldn't a smaller manufacturing process (like 28nm) give them a theoretical time-frame N*X instead?
 
That's what she said.
Sorry, couldn't resist..


I for one think that a well implemented haptic feedback replaces the "urge" of a tactile feel.


Which reminds me of another feature that's absent from the specs: rumble.
Does the NGP have it?
Nintendo decided to keep it out of the 3DS (and DSi, DSXL), even though there was a rumble add-on for the DS/lite.

The killzone video show the player shaking the system or looks like its getting feedback from the explode, maybe thats a hint or could be one of the feature that is still yet to be decided.
 
Sorry, but that really sounds like a pure marketing comment. Power draw wasn't even a variable in my question.
I don't know what the clocks are in that SGX543MP4, but I have a hard time believing it can beat the 500MHz Robson with 80 ALUs, 8 TMUs, full DX11 shaders and programmable tesselator.

I'm going to question whether the SGX543MP4 is really faster than the C-50's 280MHz core.

(Posted this on SemiAccurate as well)

-GMA 500 with Atom gets 80-90 in 3DMark06
-GMA 600 is the same SGX535 with 2x the clock. According to Intel, the E680 with the GMA 600 at its max 400MHz clock gets roughly double the 3DMark06 score of Z530 and GMA 500(SGX 535 @ 200MHz)
-16x GMA 500 = 1600 3DMark 06, which is slightly less than C-50s 1700

Contrast with the confusing results elsewhere:
-GMA 500 on WinXP with Quake 3 Arena 640x480 Normal settings: 25 fps
-iPad: 50-60 fps
-Moorestown low power version: ~60 fps
-Moorestown high performance version: ~100 fps

The graphics in Moorestown with its top 400MHz clock, is the same as the E680. So the confusion stems from how they achieved 4x performance out of a 2x hardware improvement? 3DMark06 shows the expected 2x improvement.

The only difference between the Moorestown results and the E680/GMA 500 one is the OS, so the OS could be that much more efficient.
 
Well here comes the trick question the topic is actually about: assume NGP could be eclipsed within theoretical time-frame N (under 40/45nm); wouldn't a smaller manufacturing process (like 28nm) give them a theoretical time-frame N*X instead?

In theory, perhaps. I do believe I read somewhere that a lot of factors don't scale as well anymore now with the process changes, but I believe it really doesn't matter. I think what they are doing currently is already expensive enough. They may not be able to do a global launch at the end of the year. If 28nm comes sooner, it just means that they can start dropping in price faster. I don't believe there is a tangible benefit for them ... the whole handheld economy scales similarly in that respect, and they maintain the strong advantage of being able to have a dedicated system that can be programmed to the core, and has dual analog sticks. There won't be many other types of devices, including phones and pads, that will sport those I'm willing to bet. Like Nintendo, I think they've safely cornered themselves a decent piece of the market with this one.
 
I'm going to question whether the SGX543MP4 is really faster than the C-50's 280MHz core.

(Posted this on SemiAccurate as well)

-GMA 500 with Atom gets 80-90 in 3DMark06
-GMA 600 is the same SGX535 with 2x the clock. According to Intel, the E680 with the GMA 600 at its max 400MHz clock gets roughly double the 3DMark06 score of Z530 and GMA 500(SGX 535 @ 200MHz)
-16x GMA 500 = 1600 3DMark 06, which is slightly less than C-50s 1700

Contrast with the confusing results elsewhere:
-GMA 500 on WinXP with Quake 3 Arena 640x480 Normal settings: 25 fps
-iPad: 50-60 fps
-Moorestown low power version: ~60 fps
-Moorestown high performance version: ~100 fps

The graphics in Moorestown with its top 400MHz clock, is the same as the E680. So the confusion stems from how they achieved 4x performance out of a 2x hardware improvement? 3DMark06 shows the expected 2x improvement.

The only difference between the Moorestown results and the E680/GMA 500 one is the OS, so the OS could be that much more efficient.
You have to factor one very important element while doing this type of comparisons. The Windows drivers for Intel integrated graphics are horrible, not just bad.
 
They'll just win back form factor and battery. NGP won't be eclipsed very fast, because there will be very few other devices that allow you to program to the metal.
Perhaps even more than that, there won't be a unified gaming platform for AAA titles. Even if smartphones at the higher price range can outperform NGP within a year of release, who's going to be writing Uncharted and FIFA and the like for them with their non-gaming controls and limited, fractured install base? That level of performance is only needed for the hardcore game experience, and those gamers are going to value library and controls which smartphones will likely not compete with for several years even if something like PSS takes off and becomes a standard universal platform. Only if NGP fails will AAA titles on highend smartphones make sense as that'll be where the hardcore portable gamer goes.
 
Well here comes the trick question the topic is actually about: assume NGP could be eclipsed within theoretical time-frame N (under 40/45nm); wouldn't a smaller manufacturing process (like 28nm) give them a theoretical time-frame N*X instead?

The marketing timeframe takes high importance too. i.e. they decided that they needed to show PSP2 in advance of DS3D launch to hold some buyers from buying DS3D, and to show PSP2 now, they need it on the shelves before the year is out.

28nm might give them a platform that survives at the top end longer, but thats of little use if they deem that waiting 9-12 months longer for launch means the opportunity is totally missed.
 
Perhaps even more than that, there won't be a unified gaming platform for AAA titles. Even if smartphones at the higher price range can outperform NGP within a year of release, who's going to be writing Uncharted and FIFA and the like for them with their non-gaming controls and limited, fractured install base? That level of performance is only needed for the hardcore game experience, and those gamers are going to value library and controls which smartphones will likely not compete with for several years even if something like PSS takes off and becomes a standard universal platform. Only if NGP fails will AAA titles on highend smartphones make sense as that'll be where the hardcore portable gamer goes.

There already is FIFA for iOS, priced just under $10 for iPhone and just over $10 for iPad. Not sure that it sold too well at those prices.

But EA continues to develop, because the installed base is too big to ignore.

The number of NGP units shipped at release is going to be very limited. PSP sold 200k at launch in Japan in a couple of days. They delayed US launch to the following March. I think maybe they hit 1 or 2 million in the first year?

NGP will be launching against smart phones which will ship tens of millions a quarter -- over 10 million in some months. Meanwhile, supplies could be constrained by components like that OLED screen.

Even if the demand for $50 AAA games is there, they may have a hard time getting enough hardware out.
 
Back
Top