[citation][nom]_horse[/nom]Wouldn't selling PSP2 games on the Memory Stick Duo platform make it that much easier for people to just go buy a large card and put a bunch of the games on them instead of buying them?[/citation]
Then again, if people weren't really buying your disc-based games in the first place, you're losing either way. At least with digital distribution, you're going to be minimizing the amount of investment to publish a game, hence making the losses less painful on the times they do happen. If memory serves, the PSP, during the UMD-only era, had an utterly abysmal attach rate, which has improved (albeit not enough to get the rate comparable to the DS) since they switched to MemoryStick games.
And besides, I believe the attach rate STILL sucks, and perhaps that particular division of Sony is smart enough to realize that the PSP
[citation][nom]Tindytim[/nom]Why not use mini-BDs? They're 8cm and they can hold 7.5GB single layered.I wonder if they'll have BC? Might be able to sell a peripheral that would allow one to transfer PSP1 games.[/citation]
The problem with that is that, if memory serves, the height of the PSP is only a little over 7cm. 8cm may be notably smaller than the full-size 12cm discs, but they are really too bulky for a "portable" machine, where the largest format used was the 6.3cm UMD, and then even smaller for Game Boy cartridges, and down to 3.8 cm for DS cartridges.
Plus, of course, it's quite arguable that non-solid-state data storage serves really little point for portable machines, given that they are infamous for sucking down batter power to spin up the disc, (which also means higher latency to read data, and of course, the format has lower transfer rates) as well as run a laser to read them. They also require moving parts, (which are prone to failure, especially in something that even just gets regularly held and carried around) and altogether are more expensive to make.
Basically, the only argument optical media has is that it's cheap to produce; in home systems, this makes tons of sense, but for portable systems, once you add in something to make them less-than-not-idea for frequent transport, their costs go up, and given the rediculously low price of solid-state non-volatile memory these days, one can readily get plenty enough space from solid-state chips that don't run too high in price, especially, again, once you compare them to the optical media that has to be encased or provided similar protection from wear and damage.
[citation][nom]Tindytim[/nom]I'd guess the PSP2 will have a cell processor that's lower clocked, and has fewer SPEs (4?). 480p screen (854x480?).[/citation]
At such a small size, 854x480 might be impractical, to be honest. And there is zero chance of them actually using an electricity-guzzling Cell, unless the "portable" in "PSP" means that you can carry it in your pocket when it's not plugged in.
I should note that the only practical reason for the Cell, which is horrible at handling the instruction-intensive FP64 used in gaming, even has a place in the PS3 is because it happened to be perfect for decoding high-definition media streams, (i.e, Blu-Ray) which are very instruction-light (data intensive) and done in FP32. Basically, as a chip for gaming, a Cell is worse than most contemporary PC CPUs, including ones that are cheaper and consume less power. It's strength lies as a media processor, which is what the SPEs are good at.
Since the PSP2 most certainly won't have a high-definition screen, it won't be playing high-definition video, so it most certainly won't need that sort of media-processing power that the Cell would provide. What many people miss is that the Cell isn't exactly a "God chip" that produces phenomenal amounts of power for whatever has it; it's a CPU with its own design strengths and weaknesses like any other. It has its purposes; I imagine more and more home media devices will pack it or similar chips, especially as new fabrication processes give it die shrinks and bring the price down; 8-10 years later, the 22nm verion will likely run well under $10US, and hence could pop up in almost any half-decent Blu-Ray player, offering cheap players that run smoothly, and perhaps can even afford some quality filters. However, portable electronics is not one of the places for the Cell.
As a last note, if backwards compatability is desired, it would most certainly be best to keep a similar CPU architecture, so it's incredilby likely that Sony will stick with something on the same MIPS architecture that was used by the original PSP. Of course, to be honest, it won't need all that much more CPU power, since game consoles really don't rely too heavily on it. (in fact, I'd not be surprised that with perhaps 90% of Xbox 360 and PS3 games, they don't even use more CPU power than even the Wii can provide) Rather, graphics power is what is heavily needed, and I imagine the PSP2 would likely have a much-advanced graphics subsystem; the first PSP's GPU had, as I judged, to be 2 ROPs, 4 TMUs, and a single T&L unit, making it comparable to a GeForce 256, or GeForce 2MX/4MX.
Likely, some more T&L power will be added, along with more texturing power, and likely some pixel shading capability. Of course, the main limiting factor will likely be the PSP2's screen resolution; you don't exactly need a terribly powerful GPU for even getting 60fps at low resolutions.