PSP Launching Price : Japan 48,000 Yen, UK : ~250 Pounds

[Dave Baumann]

It's also not a leap of faith when the same thing was done on the 90nm node; they clearly stated they sampled a 64MBit SoC on 65nm.

So, why bother with 90nm at all for PSP?

 

[Vince]

So, why bother with 90nm at all for PSP?

Production Capacity is a finite resource; cost-benefit analysis; there are a thousand plausible reasons if you'd stop to think about it.

Or, to turn the tables, why did ATI "bother with" 150nm "at all for" R300? UMC has been yeilding 130nm 3D parts since April 2002.

 

[Dave Baumann]

Or, to turn the tables, why did ATI "bother with" 150nm "at all for" R300?

Precisely. In ATI's opinion 130nm wasn't ready - which is the reasonable conclusion for 65nm and PSP, otherwise they would be going for 65nm.

Panajevs point asserion was that PSP is already sampling on 65nm, but there is evidently some need to release it at 90nm - given the target of PSP is to make a profit on the hardware they evidently feel they can do this at 90nm and not at 65nm (yet). PS3 will be a loss leader so they'll probably use that that get the process mature and then move the PSP across a year or so down the line.

 

[Vince]

Precisely. In ATI's opinion 130nm wasn't ready - which is the reasonable conclusion for 65nm and PSP

WTF?

Panajevs point asserion was that PSP is already sampling on 65nm

Care to quote him for me? I can't seem to see where he explicitly said this.

 

[V3]

So, why bother with 90nm at all for PSP?

Recouping investment. Sony is not in ATI position, they're not outsourcing production. So not using the process is like starving them, not very good in the long run.

 

[Brimstone]

It's also not a leap of faith when the same thing was done on the 90nm node; they clearly stated they sampled a 64MBit SoC on 65nm.

So, why bother with 90nm at all for PSP?

If Sony is moving along ok with 65nm, then I think they will just concentrate on that for the PSP. The PSP at 65nm will make the only real counter attack from Nintendo to be with a lower price. Nintendo really won't be able to out trump them on a SoC fabbed on 65nm 12 months from now.

To me the most brilliant success of Sony with the PS2 is how they were able to keep the price high and maintain great sales. Consumers willing to pay more for Sony technology is Sonys ace up the sleeve. Microsoft and Nintendo should be very affraid if with the PS3 Sony ups the ante and goes with a higher price point. Consumers are willing to pay the extra money from the Sony name brand, but I'm not so sure consumers are willing to pay more for a Microsoft or Nintendo product. They need to maintain this with the PSP, so a high quality product is essential. If the price has to be high so be it.

Sony already rewrote the book on how the game console market works and they can do the same in the portable market. They have only one real enemy at this point and it's themselves.

 

[Dave Baumann]

Precisely. In ATI's opinion 130nm wasn't ready - which is the reasonable conclusion for 65nm and PSP

WTF?

What do you mean "WTF?" - if ATI's opinion was that it was ready for their application they would have done it, however they felt it wasn't so they didn't - if you look at NV30, given the similarities with what they were both trying to achieve, you'd say that their decision was probably a little better, no?

Quite obviously there are other decisions for Sony saying that 90nm is better for PSP for now.

Care to quote him for me? I can't seem to see where he explicitly said this.

Sorry, he said its probably wasn't far behind the EE+GS.

 

[Vince]

Precisely. In ATI's opinion 130nm wasn't ready - which is the reasonable conclusion for 65nm and PSP

WTF?

What do you mean "WTF?" - if ATI's opinion was that it was ready for their application they would have done it, however they felt it wasn't so they didn't - if you look at NV30, given the similarities with what they were both trying to achieve, you'd say that their decision was probably a little better, no?

Quite obviously there are other decisions for Sony saying that 90nm is better for PSP for now.

I mean, WTF as it's becoming tradition in conversing with Dave (analogous to head-banging) that mass problems persists regarding comprehension, I believe their conversation was about aggressive fabrication advances to lower the per unit cost. Panajev stated that just as it's highly probable that the 65nm 32MBit SoC which OTSS recently talked of, is, in fact, a EE+GS@65nm - which, if true, means they could make an analogous move in the near future with the PSP's SoC to reduce costs. DM then made a good point wrt to productiona availability at that node.

You seem to be confused again. I don't see anyone but Brimstone talking about an off-the-bat move to 65nm wrt the PSP's MPU. Which is what you're implying by stating "So, why bother with 90nm at all for PSP?"

It would appear that they're talking about future reductionist moves... kinda like we've seen in PlayStation 1 & 2. Which would make sence when Pana bluntly stated:

Let them shrink the puppy to 65 nm ( after the initial launch shipment which will probably sell out in Japan and maybe even in the U.S. if they hype it well at next E3 ) and the cost will fall decisively...

 

[Dave Baumann]

You seem to be confused again. I don't see anyone but Brimstone talking about an off-the-bat move to 65nm wrt the PSP's MPU. Which is what you're implying by stating "So, why bother with 90nm at all for PSP?"

I was talking about timings, of which fab capacity and and estimated production costs on that process are obviously key elements to assessing the timings. If the timescale you metion for 90nm are analoguous at 65nm then they could roughly hit the expected release times at 65nm, which clearly they are not doing, which leaves us with one or all of the options:

• The timescales are not analoguous
• They are not at an equivelent point in the 65nm development process
• Fab capacity isn't sufficient
• costs for the process are expected to be too high for PSP to profit on the hardware

(Plus numerous other things that I can't think of right now)
In which case is it a forgone conclusion that it will be cheaper for the PSP to go to 65nm straight away?

 

[akira888]

Or, to turn the tables, why did ATI "bother with" 150nm "at all for" R300? UMC has been yeilding 130nm 3D parts since April 2002.

Yet the XP4 never launched and quanity issues may have played a role in this. Only now are we seeing this chip, almost two years later - as the XGI Volari V3.

Consumers are willing to pay the extra money from the Sony name brand (snip)

There willing to pay more because PS2 has the games they want, not just the Sony name brand. How long this disparity in developer resources will continue is anyone's guess.

I think the price should not be any higher than $299 which is a nice step down from $450.

Also, pretty much all of the features I listed are directly tied to how well it plays games and how good the game experience is.

For example, a nice screen + fast processors are key to display great looking 3D graphics and a nice Sound DSP is key to allow for great Sound FX and Music in your games.

A portable system will not sell for 300 dollars. Look at NGage - my local Target had a shipment come in around October and last time I went there about 2 of the original six were gone. On the other hand they go through GBASP shipments every week. It's going to be hard to justify a price three times higher than the competitor, regardless of is abilities.

bigger library != better library

Don't extrapolate from your own experiences onto others. For my preferences PS2 has a much better selection of titles. You see this is totally subjective; there is no objective definition of "better library."

 

[Vince]

I was talking about timings...

Well, I'm glad you were since nobody else was. How we go from a discussion about a die shrink from 90nm -> 65nm and instead we get Dave talking about applying analogous time metrics to see the plausability of a statement that nobody made -- funny, I just did that with Intel's metric compared to Sony's schedual in another thread that you can't comprehend. Hypocrisy rules!

 

[Teasy]

Ty

If there is one thing nobody would blame for the failure of Dreamcast it would be its graphics hardware.

Kyro 1 and II didn't fail. Kyro became a popular and well recognised budget graphics solution which sold well (millions of chips sold) and made profit for both IMGTEC and STM. STM's financial problems elsewhere, and there desire for a quick cach injection, were not the fault of Kyro 1 or II.

 

[marconelly!]

I'm betting even ngage would sell better if there was at least one must-have game on it. Since there isn't, it's floundering...

As for those claiming it will be 300 bucks well then sony is just going to be fighting itself . What would you rather own. A cutting edge home system or a cutting edge portable with both being the same price/

PSP comes out at the end of 2004. PS3 comes at the end of 2005 or sometime 2006. By then, they will drop the price. It's not like there's much of a difference in pricing between GBA and GC right now as well, but both are selling (you can sometimes even find a Cube for $80 in Wallmart, which I think is exactly what GBASP goes for) So even if that happens, that practice is not unheard of. I don't quite understand your complaints, really. There you go asking for PSP to have PS2 level hardware, yet you complain that even what is there in PSP (close, but sub PS2) is too expensive...

Naomi 2 was a Sega specific hardware solution, based on previous hardware that had been designed to be cost effective for a fairly cheap console to be released back in 1998. The CPU and GPU's were the same type and speed as those in the Dreamcast. Elan (the T&L unit) was the only new peice of technology.

On the other hand, there were two power VRs on it plus the Elan chip, plus tons of RAM. That doesn't exactly ring 'cheap' to me. Naomi 2 was designed for the arcades, where the cost of the hardware is not an issue.

If IT had been designing a chipset to go in a console in 2000, for the kind of budget the PS2 had, I think it's reasonable to expect that it would have significantly outperformed Naomi 2 on the GPU end. Maybe even T&L - Elan was a 100mhz part (to sync with the 100mhz memory I believe) after all.

Maybe, maybe not... Maybe we would just get souped-up Dreamcast, which would again be better than PS2 in some things and worse in others. It's all up to speculation really at this point, really...

I wouldn't say they would have no control whatsoever. Sega are supposed to have requested specific features within the hardware, and AFAIK weren't held to an Nvidia style buying of ready made chips agreement (didn't NEC manufacture the chips for Sega, rather than for PVR who then sold them fixed price to Sega?). With the talk of the "DC on a chip" in late 2000 it certainly looks like they weren't forced to take hardware any way people chose to give it to them.

Still, if you could create something yourself and have a clear vision of the return on investment, would you do it, or would you contract others to do it for you, adding another variable to an already complex equation. There are many advantages to what Sony decided to do with PS2. They starting making money fairly quickly (while look what happens to Microsoft who contracted every piece of hardware), or they could implement PS1 compatibility due to their in-house design GPU. It will allow them to make PS3 backwards compatible as well, without any headaches (again, look at MS...) There are more advantages that Panajev already mentioned earlier, and Sony does contract where it really makes sense, like memory for example.

 

[jvd]

PSP comes out at the end of 2004. PS3 comes at the end of 2005 or sometime 2006. By then, they will drop the price. It's not like there's much of a difference in pricing between GBA and GC right now as well, but both are selling (you can sometimes even find a Cube for $80 in Wallmart, which I think is exactly what GBASP goes for) So even if that happens, that practice is not unheard of. I don't quite understand your complaints, really. There you go asking for PSP to have PS2 level hardware, yet you complain that even what is there in PSP (close, but sub PS2) is too expensive...

we don't know exactly when psp will come out or ps3. We all don't know if the psp will drop in price in only a year .

Both the gc and gba sell well because you can have both for under 200$ .

You wont be able to have both a ps3 and a psp for 300$ it will be at least 500$ or more .

Also as of now there is no reason to have both. Where a gba helps alot with a gc in certian games .

 

[Panajev2001a]

at 65 nm that would probably turn out to be a ~44 mm^2 chip with low power consumption.

I'm not sure about this specific chip, but it is rare for highly complex chips to be able to be that small because they become pad limited.

I am not 100% sure about this chip either and that is why I said "could"

44 mm^2 is not what I could call incredibly smal though... I mean some ARM and MIPS cores ship at about 2+ mm^2 with cache.

 

[marconelly!]

Also as of now there is no reason to have both

And again the weird logic that a home console somehow replaces the functionality of a handheld. I don't follow that, and I don't think any amount of explanation will help me understand that logic...

 

[jvd]

Also as of now there is no reason to have both

And again the weird logic that a home console somehow replaces the functionality of a handheld. I don't follow that, and I don't think any amount of explanation will help me understand that logic...

The logic that i'm using is for the same price would you rather have an extremly powerfull home console or a handheld . That is the choice that many will have to deal with .

Also if nintendo undercuts sony again you may have a choice of an extremly powerfull home system for much less than portable.

 

[Panajev2001a]

I only said "let them shrink the puppy to 65 nm... etc... and the price will fall down" and unless you have reasons to believe that they will never do it I do not see what the issue with my statement was.

SCE wants to launch PSP in Q4 2004 and I do not think that their 65 nm lines ( the first one co-owned with Toshiba, Oita #2 ) would be ready in time to mass-produce PSP SoCs to meet that launch date.

 

[marconelly!]

The logic that i'm using is for the same price would you rather have an extremly powerfull home console or a handheld . That is the choice that many will have to deal with .

Well, that's not the same as the quote I quoted. You said there that thre's no reason to buy both, which means that even if one has money, there's no reason.

 

[jvd]

The logic that i'm using is for the same price would you rather have an extremly powerfull home console or a handheld . That is the choice that many will have to deal with .

Well, that's not the same as the quote I quoted. You said there that thre's no reason to buy both, which means that even if one has money, there's no reason.

I never said there was no reason. I said which would you rather have at that price. Not everyone can afford or want to spend 600$ to have both . That is assuming that sony launches at 300$ for both of them.