PS3 to get 45nm cell processor

[j_garcia]

What that means is: lower cost for the unit, it should run cooler and did I mention lower cost?

http://www.engadget.com/2009/02/05/david-reeves-drops-hint-of-sonys-shift-to-45nm-ps3s/

 

[MinusTheBear]

How much of a price drop are we to suspect? They said it will compete "better" price wise but give no figures.

 

[j_garcia]

I wouldn't expect a huge drop, like $100, and it is hard to say with the current market, but I'd expect around a $25-$50 decrease.

 

[Shock]

Yeah, somewhere inbetween there I would expect.

 

[MinusTheBear]

That is what I was thinking as well. Max $50 drop. For now this move seems insignificant.

 

[j_garcia]

Not in that article, but elsewhere I read that it is actually a pretty big step. The 45nm cell has already been proven, so the question was, why did Sony take so long to incorporate it? It opens the door for them to drop the price even further down the road, so since they are saying this may hit mid year, I'd expect good prices on PS3s come 09 holidays. Lower power draw and cooler running should be a plus as well for all those who don't like the fan noise.

 

[lsiberian]

Cooler running on a smaller chip?

I question this logic. The transistors are more tightly packed. How do they make it cooler? This is a clear upgrade in performance, by shortening the transistors the cpu will respond much faster. But I think the only reason it might run faster is because they won't fully utilize its power.

It's kind of like having a more powerful amp on the same speakers.

I guess reducing 200 atoms on each chip will be nice though.

Now it only 450 atoms. Wow. We are getting close to the max.

 

[lsiberian]

Not in that article, but elsewhere I read that it is actually a pretty big step. The 45nm cell has already been proven, so the question was, why did Sony take so long to incorporate it? It opens the door for them to drop the price even further down the road, so since they are saying this may hit mid year, I'd expect good prices on PS3s come 09 holidays. Lower power draw and cooler running should be a plus as well for all those who don't like the fan noise.

I doubt the chip runs cooler. But the PS3 is fairly quiet already.

 

[j_garcia]

A LOT of people complain about the fan noise. It doesn't bother me. I don't think it will run considerably cooler, but with it being smaller, I would expect the system overall will run a little cooler. Some of the articles are discussing possible cooler running, but I don't know that to be a fact.

 

[GlocksRock]

I'd like to see Sony implement the new chip and make a slim PS3.

 

[j_garcia]

I'd like to see Sony implement the new chip and make a slim PS3.

I was thinking that too, but I don't know how much it will allow them to downsize the system. I would imagine they will keep the same case for now just because people recognize it, and eventually shrink it down. I would like to see it a bit smaller also, though it isn't exactly giant currently.

 

[trnqk7]

The smaller chips draw less current, therefore, there is less heat so they run cooler. Just like your cpu's-remember the 90nm P4's drawing like 135W's or more? When the Core2 came out at 65nm, it had a lot less power consumption, <90W (I don't recall the exact number). Same deal with the shift to 45nm and 32nm after that.

Part of it will have to do with shorter "wire" lengths (hard to call it a wire at such short distances, but still functionally the same) so there will be less I^2R losses as heat. There's more to it than that, but I'm not entirely clear on it all myself. I'm sure others can chime in.

 

[lsiberian]

The smaller chips draw less current, therefore, there is less heat so they run cooler. Just like your cpu's-remember the 90nm P4's drawing like 135W's or more? When the Core2 came out at 65nm, it had a lot less power consumption, <90W (I don't recall the exact number). Same deal with the shift to 45nm and 32nm after that.

Part of it will have to do with shorter "wire" lengths (hard to call it a wire at such short distances, but still functionally the same) so there will be less I^2R losses as heat. There's more to it than that, but I'm not entirely clear on it all myself. I'm sure others can chime in.

I'm not sure about that. The shorter transistors would allow for tighter packing of the transistors. This leaves less air in between for heat to dissipate.
I'm not a chip designer though that would be my father. Things are way more complicated in the 100s of atoms than in the 1000s of atoms. Part of the reason the cpu's run more efficient is they don't run it at full speed all the time. They have step down logic for the processor. Depending on load.

Of course I am not an EE so I could be mistaken, but I then blame my Computer Architecture prof.

 

[j_garcia]

I've worked in semiconductors for many years, but in the equipment field not the chips themselves, so I am not sure either, but I'd say that trnqk7 is right in that the lower current draw should result in lower heat generation, even with the smaller architecture. Lower power consumption was one of the reasons given for the switch.

 

[lsiberian]

I think their claims are just claims and have little real merit. I understand that's there argument, but I just don't buy it. I see some of this as a marketing ploy and maybe a way to decrease overall costs. Just my 0.02$ but it's nice that they are using better chips. However the PS3 will be noticeably faster. And that is very nice.

 

[j_garcia]

Not that it is slow now, but I'm sure it will help. It is to reduce THEIR cost, not so much the cost to the consumer

 

[Shock]

I'm not sure about that. The shorter transistors would allow for tighter packing of the transistors. This leaves less air in between for heat to dissipate.
I'm not a chip designer though that would be my father. Things are way more complicated in the 100s of atoms than in the 1000s of atoms. Part of the reason the cpu's run more efficient is they don't run it at full speed all the time. They have step down logic for the processor. Depending on load.

Of course I am not an EE so I could be mistaken, but I then blame my Computer Architecture prof.

The chips run cooler because they should in theory require less power to run them at equal clock speeds.

 

[MinusTheBear]

It is to reduce THEIR cost, not so much the cost to the consumer

I think you nailed it!

 

[lsiberian]

The chips run cooler because they should in theory require less power to run them at equal clock speeds.

Yes but the transistors are packed into a tighter area by the reduced chip size. So whatever gains are made by reducing the power I think will be largely countered by the decreased space between transistors. So I think you fan would have to be as powerful but the heatsink would need to channel the air into a smaller space. In theory that would generate higher noise from the fan, but whoever thinks the fan is loud needs to get a better sound system or do some treatments to their cabinet anyway. I have mine right behind me and hear no sound from it. I bet the noise is from lack of ventilation. Which is a bad idea for the PS3.
My rack is very open and well ventilated.

I hijacked the forum a bit with a chip architecture discussion. I will defer to the electrical engineers and go back to living in the software world.

 

[lsiberian]

I think you nailed it!

Not entirely Sony needs the PS3 to last a long time for them to regain their investment. So they are upgrading it to do that. I wouldn't be surprised to see a terabyte PS3 with a 30nm chip in the future.

This is a long term investment not a short term cost cut. Since 45 nm chips have been around a few years now they are probably cheaper to produce than 60 nm chips.