xbox 360 specs (unofficial, but believable)

Shifty Geezer said:
hugo said:
A tri-core CPU @ 3Ghz does not come cheap.Look at the P4 EE 840 dual core for example and the PowerPC G5 2.7Ghz.I am wondering how they are going to price their consoles below the 500 dollar mark.A 9-core Cell?with a top of the line GPU included?How much losses are these companies going to bear?Next gen would not be cheap.
The 9 cores of Cell aren't comparable to the dual Pentiums cores. Overall Cell occupies as much die space as these dual-core processors, and as I understand it that's the limiting factor. Plus the dual-Pentiums have a massive markup as they're for a specialist market willing to blow money away.

Price of dual-Pentium/PPC is not comparable to price of 9 'core' Cell.

But still it will still be priced about the same or slightly more than that for its capabilities.How much does a dual core P4EE cost today?The last time I checked it was over a thousand and that's just the single core 3.7EE.
You haven't include the price of the GPU,XDR and drives yet.Sum them all up and you'll ponder how MS brings their console down to sub $500 level?As for Sony with the BD Nvidia GPU and Cell included I am wondering how are they going to sustain all that losses.
 
hugo said:
How much does a dual core P4EE cost today?The last time I checked it was over a thousand and that's just the single core 3.7EE.

Don't confuse cost with price.

I've seen quotes on production cost as low as $15 for Prescott dies.

Cheers
Gubbi
 
Gubbi said:
I've seen quotes on production cost as low as $15 for Prescott dies.

Gubbi

Does that mean that the PC market components are overly priced?I am sure IBM would never sell CPUs at that kind of level to MS.Why would TSMC be willing to fab those CPUs for MS when PC components have higher profit margins?Sony would also need to include their investments made for their plants.
 
hugo said:
Gubbi said:
I've seen quotes on production cost as low as $15 for Prescott dies.

Gubbi

Does that mean that the PC market components are overly priced?I am sure IBM would never sell CPUs at that kind of level to MS.Why would TSMC be willing to fab those CPUs for MS when PC components have higher profit margins?Sony would also need to include their investments made for their plants.

U'd be surprised at the kind of profits Intel and others make from their processors.
 
hugo said:
Gubbi said:
I've seen quotes on production cost as low as $15 for Prescott dies.

Gubbi

Does that mean that the PC market components are overly priced?I am sure IBM would never sell CPUs at that kind of level to MS.Why would TSMC be willing to fab those CPUs for MS when PC components have higher profit margins?Sony would also need to include their investments made for their plants.

All it means is that Intel enjoys a healthy profit margin. On top of the $15 you get testing and packaging and one-to-many resellers before the chip reaches retail.

Also, while their top dogs are pricy, the ASP is alot lower, don't forget that their biggest sellers are Celerons which are basically the same die as the high end P4s (cache configurations not withstanding), that sells for $50 and upwards.

IBM has higher production cost than Intel per mm^2, but it's nowhere near the multi-hundreds of bucks per chip mentioned here.

Cheers
Gubbi
 
Gubbi said:
hugo said:
How much does a dual core P4EE cost today?The last time I checked it was over a thousand and that's just the single core 3.7EE.

Don't confuse cost with price.

I've seen quotes on production cost as low as $15 for Prescott dies.

Cheers
Gubbi

Good point, but surley the $15 is only for direct production costs and doesn't include overheads?

If your just talking about the direct costs for the production company to make a die then thats probably not a very good example. I think it would be more appropiate to state the cost for Intel to get the chip to the point where they can sell it to 3rd partys for retail (if thats applicable). Then there's still fixed costs that will have to be taken into account.
 
As I already said (I talk, but no-one listens!)...
Plus the dual-Pentiums have a massive markup as they're for a specialist market willing to blow money away.
Not everyone's going to spend a grand on a CPU, so for them you need a cheaper solution. But you sell more of the cheaper chips.

As a chip designer, you are working on new technologies to maintain performance with competitors. When you have a new tech you want to break it in. You can't make lots of these chips as the fabbing tech isn't up to scratch, say, but can charge maximum bucks for this cutting-edge for those who demand cutting-edge. Once the fabbing gets better you can sell more at a lower price. That's a simplified business model, but you get the idea. Basically you always charge what people are willing to pay, not necessarily relative to what something costs to produce.
 
Ragemare said:
Good point, but surley the $15 is only for direct production costs and doesn't include overheads?

It's from sand going in to dies coming out.

Add testing and packaging. Testing is not expensive, but yield < 100% will increase average price on good dies of course. Packaging is *cheap*

So not alot of cost added. But don't forget Intel enjoys a fairly unique profit margin on their products.

Cheers
Gubbi
 
GwymWeepa said:
Wtf, you expect the ability to record video and play games at the same time? I for one certainly don't.

That's the point, it doesn't have enough horsepower to function as an HDTV PVR and play games simultaneously.

It would be lame if they said yes it has PVR but make sure you don't play any games while it's suppose to be recording in the background. Or don't schedule any recordings when you think you might be playing games.

Either they do it right and put dedicated hardware for the PVR portion or they don't offer the PVR functionality at all.
 
wco81 said:
Either they do it right and put dedicated hardware for the PVR portion or they don't offer the PVR functionality at all.

Who ever said it would have a PVR?
 
corysama said:
blakjedi said:
[

I agree.. i'm actually disappointed if these`specs are real...

You are disappointed by the prospect of 9GHz machine with a better-than-top-of-the-line video card for $300? Are you really going to give MS are hard time because they didn't invest enough in time travel to pull technology out of the labs of next year (when the PS3 ships) and hand it to you on a silver platter?

Anyone with a passing familiarity with technology would expect a console that comes out at a much later date for approximately the same price will certainly be more advanced than the one that launches earlier. If the PS3 was launching at the same time for the same price as the 360 then MS would deserve some criticism, but that is not the case.

When the 360 comes out for Christmas, it will be as good as the technology of Earth circa late 2005 can deliver for $300.


Ahh... I would have agreed with you if... I didnt know that all three systems were being developed simultaneously. PS3 development started in 1999. N5 development I believe began in 2000/2001 timeframe with X360 starting a lil later. How is it that the last console in development, using the expertise of the same companies that their competitors are using ends up with the "weakest" technology?

I say all that to say that Sony COULD drop the PS3 in 2005 if they wanted to... if Sony did then THAT would represent the latest in Earth technology AVAILABLE. And MS would look even worse. Instead Sony CHOSE to release their console in 2006 (really only four months later Nov - X360, March - PS3).

In fact in many ways PS3 tech is older than than X360 tech cause no one has seen any X360 tech yet but we have seen a working Cell chip. And for all intents and purposes, it seems that the Cell is more advanced and powerful than whatever is in the X360. In 2005 no less.

I dunno, I'm rambling maybe...
 
See, saying Cell is more powerful but older...its never as clear cut as that. Cell represents a shift in programming and processing design. It's not so much 'better' as 'different'. If Sony had gone with a conventional approach, CPU, GPU, etc, their bleeding-edge tech would be limited to the same performance as the others. By sidestepping into a new field, they get different yields per tansistor count on different functions.

The limiting factors aren't so much technology, but money and time. Sony set up straight after PS2 to work on its sequel, it seems. They've thrown stupid amounts of money and forged alliances to drive a more powerful solution for their needs. Comparatively MS have spent less time and less money, so why should they get better or similar performance? You get out what you put in. If MS had put in the reading, decided on a similar model to Cell, and paid out for development, they too would have a comparable processor.

Cell is tooted as a new superchip, but remember there have been similar designs before, with similar super-performance compared to existing techs. They didn't get anywhere because they didn't have the same sort of backing as Cell. The key point is Cell's tech isn't new, isn't cutting edge per se. It's just a hard-work paying off implementation of old ideas.

Cell is no more cutting edge than XenonCPU. Both use the best design methods available. The differences come from different implementations which are nothing to do with cutting edge technology.
 
^^^^^ very good assessment Shifty.

People have to remember that Cell isn't just a chip going into the PS3, it's a whole new architectural model, so expect some evolution, and probably a next-gen Cell design to go into PS4. Assuming Cell as a concept hasn't become outmoded by that time, that should be a decent advantage to Sony already having the IP ready, the fabs ready, and the framework of design ready. Will hopefully allow them to eschew the billions in investment and tight time-tables measured in years for the next consoles CPU.
 
version said:
cell with 16 SPE far better than xenon's cpu(about 8 times)
and nvidia gpu 2-4 times faster than ati

What the...?

Version what are you babbling about here? :)
 
Back
Top