How can Sony further cost reduce the entry-level PS3? *spawn

For your information there are 360 games requiring mandatory installs or disk swapping (read Forza3 and Rage coming up) so yeah, the Arcade got limits too which some people may find cumbersome. So if you have to uninstall a PS3 game at some point in time people may do that and accept the hassle or they just buy a hard drive, again read Shiftys post (Edit: or corduroygt´s post above).

It's not exactly a Mandatory install if you can still play the game without installing is it? :) Being an optional component does not make it mandatory.

I get what you are saying with regards to being able to upgrade the capabilities of a system. I still disagree that an X360 Arcade unit is a good example in light of the original post you responded to. But I see where your thinking was going.

Likewise I don't see SSD prices coming down to a point before the PS3 is replaced by a PS4 that it would be suitable for inclusion in an entry level console.

Regards,
SB
 
Considering that Sony had 20 and 40GB PS3's out there without any of these support "nightmares", I think 32 GB SSD would be fine once it cost less than 20 bucks. If you are a serious gamer, you'll buy a bigger hard drive just like xbox pro users who upgraded their 20GB hard drives to 120GB. The SSD option is always there for eliminating the "fixed cost" of traditional hard drives to the bottom line of PS3 BOM, we just have to wait until they are cheap enough and I'm sure Sony will consider it too.

They had them. The 20 gig went by bye quickly. The 40 gig also went bye bye fast.

Anyway 30GB ssd's cost us $100 or more.For the same price you can get a 500gig 2.5inch drive. whats the real advantage right now ? They can advertise a 500 gig drive for most likely the same cost as the ssd and more people willpick the standard hardrive.

http://www.fudzilla.com/content/view/17385/1/

also considering there were only 10m ssd's shipped in all of 2009 i think supply would def be an issue even next year.

Sony shipped 10m units last year i believe and they will most likely ship more next year.
 
"I think 32 GB SSD would be fine once it cost less than 20 bucks."
No one is talking about right now, this is for 3 years down the road, when the next gen systems should be about to come out. I don't think neither the 360 or the PS3 will be discontinued in 3 years.
 
How can sony reduce costs? Aside from decreasing the actual manufacturing cost, Sony could try deferring it like microsoft.

1. Don't pay the blu-ray licensing fee. Have a user be able to purchase the blu-ray functionality via the store for like $50. The license would of course have to be tied to the console and not the psn user.

2. Have opera or firefox available for download and charge like $20.

3. Remove the power supply to the outside like microsoft. This will save additional generated heat and possibly allow Sony to make more savings on cooling.

4. Maybe, possibly, reengineer the cell to use gddr 5 ram? would this even be possible w/ maintaining current compatibility? I remember that japanese web site was saying that cell 2 would move away from xdr and move to gddr ram.

5. and as others have said, remove the normal hard drive and put in like a 32 gb ssd drive. Have a small section where a user could buy a sony branded hard drive and be able to stick it in for expansions.

6. remove the wireless. Have it as a addon like microsoft. Most hard core games want a wired connection anyways.
 
"I think 32 GB SSD would be fine once it cost less than 20 bucks."
No one is talking about right now, this is for 3 years down the road, when the next gen systems should be about to come out. I don't think neither the 360 or the PS3 will be discontinued in 3 years.

Ummm, yes, but the thread title is about how to cost reduce the entry-level PS3... If the cost of an SSD eventually becomes more cost effective than magnetic HDDs (doubtful for quite a while, I'm betting even 3 years is too soon) after the lifetime of the PS3, then it's rather a moot point.

Regards,
SB
 
5. and as others have said, remove the normal hard drive and put in like a 32 gb ssd drive. Have a small section where a user could buy a sony branded hard drive and be able to stick it in for expansions.

I will say I've always wondered why Sony didn't sell a hard drive upgrade in game stores, and maybe a PS branded keyboard/mouse set. It's ok that they allow 3rd party gear and are industry standard, but I think they are missing a point-of-sale opportunity at places like Wal-Mart, Best Buy, GameStop by not even offering it.

remove the wireless. Have it as a addon like microsoft. Most hard core games want a wired connection anyways.

I think the wireless serves other purposes to be honest. I use a wired connection myself (well 20GB so no choice...), but the PS3 *is* more than a game system, and insofar as folk use it in a home theater setup without ethernet - but may still be enticed to purchase a thing or two through the PSN store, might as well leave it in. And in all honesty for its cost, might be best to leave it in anyway to raise the value proposition of the system.
 
I get what you are saying with regards to being able to upgrade the capabilities of a system. I still disagree that an X360 Arcade unit is a good example in light of the original post you responded to. But I see where your thinking was going.

Yeah, with regard to DLC (game patches, map packs, PSN games, game trailers etc.), the 32 GB unit would still be limited but be far more convenient than the Arcade, so they are different in many ways. But in terms of upgradeable entry level units they would serve the same purpose.

Silent_Buddha said:
Ummm, yes, but the thread title is about how to cost reduce the entry-level PS3... If the cost of an SSD eventually becomes more cost effective than magnetic HDDs (doubtful for quite a while, I'm betting even 3 years is too soon)

Just don´t put to much money on that bet. Sony obviously don´t have to use an off-the-shelf SSD with a lot unnecessary mechanics, just look at the price of the Flash.
Using the currently lowest contract price $6.4 per 4 GB we get 32 GB -> $51 and if we apply Moore´s theorem of half the price every 18 months, the price of 32 GB would be below $13 three years from now. If Sony choose to go with less than 32 GB, well then the entry level unit could show up much earlier.

Silent_Buddha said:
after the lifetime of the PS3, then it's rather a moot point.
Quite the opposite I would say, by the time the next Playstation version is introduced it will be more important than ever to have a noticeable price difference between the old model and the new model and you can bet the next version will not cost $600, that is a rather safe bet in my opinion. :smile:.
 
Just don´t put to much money on that bet. Sony obviously don´t have to use an off-the-shelf SSD with a lot unnecessary mechanics, just look at the price of the Flash.
Using the currently lowest contract price $6.4 per 4 GB we get 32 GB -> $51 and if we apply Moore´s theorem of half the price every 18 months, the price of 32 GB would be below $13 three years from now. If Sony choose to go with less than 32 GB, well then the entry level unit could show up much earlier.

Except you can't do that unless you want a pathetically and pathologically slow SSD with an extremely low (relative to a "real" SSD) write/erase cycle.

It's the controller that allows to implement the various algorhythms that allow MLC to not only show respectable speed (look at the old Jmicron controller for a pathological bad case, and that's BETTER than mass MLC on a flash drive) as well wear leveling for increased lifespan, garbage collection, etc...

In other words, if they are going to use anything that isn't as bad or worse than a good USB stick, it's going to be roughly the same price as the cheapest MLC based SSD on the market.

Obviously they won't be paying retail price, but the same goes for magnetic drives.

And you may or may not be able to apply Moore's law to this. As there are good indications that prices for SSD's this year won't drop even close to half compared to last years prices.

Especially when you get to the low MLC chip drives like the 32 GB models where the cheaper the MLC gets the more significant becomes the cost of the controller, PCB, labor, etc...

Regards,
SB
 
Except you can't do that unless you want a pathetically and pathologically slow SSD with an extremely low (relative to a "real" SSD) write/erase cycle.

It's the controller that allows to implement the various algorhythms that allow MLC to not only show respectable speed (look at the old Jmicron controller for a pathological bad case, and that's BETTER than mass MLC on a flash drive) as well wear leveling for increased lifespan, garbage collection, etc...

In other words, if they are going to use anything that isn't as bad or worse than a good USB stick, it's going to be roughly the same price as the cheapest MLC based SSD on the market.

Obviously they won't be paying retail price, but the same goes for magnetic drives.

And you may or may not be able to apply Moore's law to this. As there are good indications that prices for SSD's this year won't drop even close to half compared to last years prices.

Especially when you get to the low MLC chip drives like the 32 GB models where the cheaper the MLC gets the more significant becomes the cost of the controller, PCB, labor, etc...

Point is that they don´t need a controler to mimic a HD, the API will take care of that, cache handling etc. All they need to do is solder the flash on the motherboard and design a way for the CPU to access it with good enough read/write speed. Shouldn´t be to hard considering that there is already some Flash memory on the motherboard.

The only reason I can see why they would use an SSD instead of flash soldered on the motherboard, would be if they there is a to high overhead cost of having two motherboard designs in production, but considering the production volumes I am doubtful that the overhead would be significant.
 
We know that Sony has also been thinking about implementing remote storage as well, so that's always an option also. It just demands some minimum level of connectedness throughout the buyer pool before it might become feasible.
 
All they need to do is solder the flash on the motherboard and design a way for the CPU to access it with good enough read/write speed. Shouldn´t be to hard considering that there is already some Flash memory on the motherboard.

If the flash is not using the standard hdd iterface, how can the user upgrade to a bigger storage device then?
 
We know that Sony has also been thinking about implementing remote storage as well, so that's always an option also. It just demands some minimum level of connectedness throughout the buyer pool before it might become feasible.

Then it sounds like it would only be practical in Europe/Japan and only for consoles hardwired into the network. They would need a minimum of 80Mbp/s (10 MBp/s) to achieve the expected level of throughput most developers who cached/used installs would expect right?
 
If the flash is not using the standard hdd iterface, how can the user upgrade to a bigger storage device then?

Not only that but you are STILL going to need the cost of a good MLC controller to get good performance and good lifespan. Otherwise that flash soldered onto the mainboard of the PS3 is going to be far slower than a magnetic HDD with again, crappy write/erase cycles.

I suppose they could go with SLC, but then the price is going to be quite a bit more and you'll still need a controller.

Regards,
SB
 
If the flash is not using the standard hdd iterface, how can the user upgrade to a bigger storage device then?
the gameos could transparently handle using either the flash or an optional HDD as primary drive? (the flash could then be mounted similar to an USB-stick/drive to allow copying your data)

Anyway, whats wrong with just omitting the HDD/flash storage on the cheapest modell and let the customer choose and not end up with a redundant HDD? Even if it means you end up spending more in total, most consumers just look at "lowest cost entry" for relating console-prices... they might not be play a single game but thats not the point... Im surprised Sony hasnt done that right from the start.
 
Then it sounds like it would only be practical in Europe/Japan and only for consoles hardwired into the network. They would need a minimum of 80Mbp/s (10 MBp/s) to achieve the expected level of throughput most developers who cached/used installs would expect right?

You could have a minimal cache presence onboard, say 4GB, for those purposes. I don't think you would need it in real-time; the storage would be just that, storage. But anyway you do make a good point relative game installs... which of course is probably one amongst several reasons they haven't gone that route. :)

(But, they have thought about it!)
 
the gameos could transparently handle using either the flash or an optional HDD as primary drive? (the flash could then be mounted similar to an USB-stick/drive to allow copying your data)

Anyway, whats wrong with just omitting the HDD/flash storage on the cheapest modell and let the customer choose and not end up with a redundant HDD? Even if it means you end up spending more in total, most consumers just look at "lowest cost entry" for relating console-prices... they might not be play a single game but thats not the point... Im surprised Sony hasnt done that right from the start.

You can't sell a console that won't work out of the box. Same reason they don't sell cars without wheels&tires, so anyone wanting to upgrade theirs ends up with a second set.
 
You can't sell a console that won't work out of the box. Same reason they don't sell cars without wheels&tires, so anyone wanting to upgrade theirs ends up with a second set.
It doesnt work out of the box currently, atleast mine came without TV.
 
Not only that but you are STILL going to need the cost of a good MLC controller to get good performance and good lifespan. Otherwise that flash soldered onto the mainboard of the PS3 is going to be far slower than a magnetic HDD with again, crappy write/erase cycles.

I suppose they could go with SLC, but then the price is going to be quite a bit more and you'll still need a controller.

Regards,
SB

:?:

There is already MLC Flash memory on the Playstation motherboard which requires a Flash memory controller, it is probably integrated in the NorthBridge.

Flash memory controler IPs can be bought and integrated in the configuration you want. Buying IP is what the electronics business is very much about today, you don´t buy discrete chips for every function you need.

Does the current Flash memory controller offer a bandwidth comparable to a hard drive? I don´t know, but if it´s not it can be fixed by accessing multiple memory units in parallel (expanding the data path).

And what is so magic about letting the OS SW mimic a hard drive behind the APIs, would be the simplest thing in the world in my opnion.
 
:?:

There is already MLC Flash memory on the Playstation motherboard which requires a Flash memory controller, it is probably integrated in the NorthBridge.

Flash memory controler IPs can be bought and integrated in the configuration you want. Buying IP is what the electronics business is very much about today, you don´t buy discrete chips for every function you need.

Does the current Flash memory controller offer a bandwidth comparable to a hard drive? I don´t know, but if it´s not it can be fixed by accessing multiple memory units in parallel (expanding the data path).

And what is so magic about letting the OS SW mimic a hard drive behind the APIs, would be the simplest thing in the world in my opnion.

Controllers to make MLC suitable as a HD is the hottest thing right now because it is extremely difficult to get it right and make it work well. That's the big deal.

If anyone could do it, Jmicron, a company that makes it their business wouldn't have been having such extraordinary problems getting out one that didn't perform like complete poo (IE - worse than magnetic HDD) for the past 2-3 years.

The Intel, Indilinx, and a new controller coming soon (can't remember name of company), aren't exactly cheap. Intel doesn't doesn't even let another other than Kingston market a drive based on their chip. And Samsung is reliable but entirely average.

Flash for firmware isn't exactly going to have to need high random IO performance, nor is it going to go through many write/erase cycles. Designing a drive to effectively use MLC and work around the deficiencies of MLC isn't easy.

Regards,
SB
 
Back
Top