Alternative distribution to optical disks : SSD, cards, and download*

[eastmen]

its because its a crappy drive

Max br single layer is 6x , max dvd r is 8x
Here is a 12x buffalotech drive

http://www.buffalotech.com/products...iastation-12x-external-usb-30-blu-ray-writer/

Power Consumption Max 24W


A 6x bluray drive will only hit 27MB/s

Current PS3 blu-ray drive does not require cooling and neither does the 360 DVD. The same drives I linked are used internally in laptops without any extra cooling. You are 100% WRONG on blu-ray power requirements, they are less than 5 Watts. Case closed.

Modern bluray drives are using 24w as i've shown above.

Not only that but they increase dead space in the console , increase power supply needs and because of both of them increase cooling needs by increasing power usage and interupting case cooling.

Take the volume of a 5 1/2 inch drive out of modern consoles and you suddenly have alot more internal space to improve cooling / reduce foot print size of the device.

If a product is being sold for way under the regular everyday retail price, it's a loss leader or inventory liquidation. Simple as that. After 22nm, there isn't much further to go by the way.

That doesn't mean much at all. Many times at a store they will eat some of their retail market to move product. If the 8GB cards weren't moving well they might take say 25% of their mark up to move te inventory. Esp if they are getting in 16GB cards that are going to skew the pricing in favor of the higher end product.

Which DRM, the one thats useless because it requires devices to honor the "DRM" (not really DRM, more similar to macrovision on VHS) or the one that requires online activation (bah)?
Unless Sony adds some additional protection it wont take 2 months for some copier devices to show up.
(If DRM on flash is a solved issue then why is every flash based media to date copied with ease...)

The extra cost is lower volume than regular SD Cards which are sold at razor thin profits thanks to very high volume and competition of multiple vendors.

But where will the devices that don't honor the DRM come from ? IF a company goes with their own packaging and doesn't release the information on how to build them someone will have to reverse engineer a reader that will strip the drm.

And thats just one form of DRM thats in the hardware. Console makers create new drm every console generation so i don't see why they would have just one type of drm on the flash

Also bluray has been hacked , the ps3 has been hacked so its not like bluray will save them from piracy.

 

[corduroygt]

In any event its an order of magnitude more power than flash requires.

At a magnitude of more cost * 1 billion software units compared to disc.

And the cases are cooled.

Don't lie. They are not cooled, show me if they are. A 5W device is easily passively cooled.

 

[AlphaWolf]

Don't lie. They are not cooled, show me if they are. A 5W device is easily passively cooled.

First, see eastmen's post above.

2nd passive cooling means the heat goes into your case, where the main fan has to deal with it.

 

[corduroygt]

its because its a crappy drive

No need to have that monster in a console, besides that's a writer. Didn't you say that PS3 blu-ray must have cost a lot of money because it was a slim drive? Now you show a monster PC drive that has nothing to do with what goes in a console because it's convenient for your argument.

The one I showed is plenty fast enough, since cheap $5 SD cards have actually lower than 27MB/s speeds. You have to get a much more expensive one to get proper speeds. 27MB/s is plenty as well.

I'd bet good money on next gen having optical drives.

 

[eastmen]

No need to have that monster in a console, besides that's a writer. Didn't you say that PS3 blu-ray must have cost a lot of money because it was a slim drive? Now you show a monster PC drive that has nothing to do with what goes in a console because it's convenient for your argument.

The one I showed is plenty fast enough, since cheap $5 SD cards have actually lower than 27MB/s speeds. You have to get a much more expensive one to get proper speeds. 27MB/s is plenty as well.

I'd bet good money on next gen having optical drives.

We've been talking about 12x bluray in other posts which is why i pointed to it.I can't find anything for just a 12x reader with no burning.


6x is 27MB/s so even at its fastest it will be the same speed as your cheap flash


I have a 8 gig sd card here that is getting 21.4MB/s read time. Take the nand in two of them and put it in a custom package for raid 0 and your looking at 42.8MB/s

Flash gets very fast by putting more chips in a raid configuration with a decent controller.

Not to mention that each micron process they are able to cram more and more GB/s into the same die size and the speed goes up

 

[AlphaWolf]

Well there's already shipping 128GB flash products that are 100MB/s.

 

[corduroygt]

Fast flash is never cheap, and neither is raiding them.

 

[AlphaWolf]

Fast flash is never cheap, and neither is raiding them.

As density increases, which it will for the foreseeable future (intel has 4nm on their roadmap in 2020), they will get faster. I'm not saying 128GB flash is cheap atm, but the speed is more than necessary, and so is the capacity. In 2 years (for launch) it will be 1/2 the price and there will be faster ones. 16GB is probably enough capacity for launch (but its certainly not a limit), and 40MB/s would probably be adequate as well. 6 years from now (well into the next gen) that 128GB 100MB/s drive would probably be palatable on 15nm or less.

Raiding them doesn't really have a specific cost, its just a matter of using more chips and more pins. A few more cents on each device, with a controller required on the console. I wouldn't be surprised if there was a raid controller on the chipsets they are using now, probably just a matter of enabling it.

 

[manux]

I'm willing to bet that flash and/or blu-ray rom read speed has little to do with next gen consoles and decision to go either flash or blu-ray. What matters is the cost of manufacturing the media at factory(time is also money if write speed is slow).

It's easy enough for sony/microsoft to put reasonable amount of flash memory as cache. That cache eclipses both blu-ray and flash cartridges in speed and price as there is need to put only 1 of those cache to each device rather than make each game media fast. Also this cache will help should the console have slower HDD for downloaded games and cache is needed to improve load times.

6x is 27MB/s so even at its fastest it will be the same speed as your cheap flash

6x is 3 times faster than blu-ray on ps3 and 8-2 times faster than dvd drive on xbox360. Assuming game binary sizes don't get three times bigger then it's quite easy to see better performance out of this 6x drive than on current gen ps3 optical media. Make that binary size growth 2x compared to xbox360. If the game sizes do get three times bigger or more that doesn't look good for flash either as you start to need rather large capacity flash memories for each game made(not sold, but made)...

I think 8 to 12x drive would be feasible next gen but it makes much more sense for console manufacturer to take reasonable speed optical media and put that 1 cache chip inside console.

Assuming SD cards or some such are fast enough(which many seem to say they are) the flash cache could just be a memory card slot I personally don't believe that SD cards would be as fast as a dedicated solution built into the console. If other manufacturer is using SD card and another one is using built in optimized flash+io controller it might be significant differentiation factor on the system. Also using standardized chips with known minimum performance is easier for developers.

I for one would prefer to buy a console next gen which plays my old blu-rays and dvd's. I rather would not jump to standalone players and livingroom electronics clutter. I'm using ps3 as my main and only media device currently if netbook and cellphone are not taken into equation.

 

[AlphaWolf]

Assuming flash chips are fast enough(which many seem to say they are) the flash cache could just be a memory card slot I personally don't believe that SD cards would be as fast as a dedicated solution built into the console.

I for one would prefer to buy a console next gen which plays my old blu-rays and dvd's. I rather would not jump to standalone players and clutter(using ps3 as my main media device currently...)

Flash media is already as fast as blu-ray is ever going to get. The lack of moving parts gives it a distinct advantage in this area.

I played a DVD once in my 360, just to see if it could do it. Even my ps2 only every played a DVD a few times. There's pretty much zero chance nintendo and MS will use blu-ray, the net benefit for them is nothing.

 

[manux]

Flash media is already as fast as blu-ray is ever going to get. The lack of moving parts gives it a distinct advantage in this area.

I played a DVD once in my 360, just to see if it could do it. Even my ps2 only every played a DVD a few times. There's pretty much zero chance nintendo and MS will use blu-ray, the net benefit for them is nothing.

Did you not read my post? I was saying the optical media bottleneck is taken away by using flash as cache similarly to what xbox360 and ps3 use hd in current gen games. I was also giving reference that next gen using optical for installation media would not cause a bottleneck compared to current gen performance gamers have accepted in millions.

 

[AlphaWolf]

Did you not read my post? I was saying the optical media bottleneck is taken away by using flash as cache similarly to what xbox360 and ps3 use hd in current gen games. I was also giving reference that next gen using optical for installation media would not cause a bottleneck compared to current gen performance gamers have accepted in millions.

Well it's not taken away. It's just moved to the beginning. You're basically looking at a multiple minute install every time you hit play.

 

[manux]

Well it's not taken away. It's just moved to the beginning. You're basically looking at a multiple minute install every time you hit play.

Let's assume for example that we could have 64GB of cache memory or ~20GB flash card for each game. So we would soak 1 time cost of about 3 game medias to each console. If you would cache let's say 3-4GB of stuff per game and stream rest from the slower memory be it optical or flash card you could go with 15 games installations and leave some 4 gigs for savegames and somesuch. Not bad? If one assumes any sd card/usb stick is fast enough then it could be left up to the user how much storage/cache space he wants to have and the console would not need to contain cache memory.

Some ps3 games like uncharted1 and 2 work with rather small cache and fill it in smartly. Could work for next gen too stretching that cache to work with much more than 15 games I think XBox360 also has this gradual filling in the cache rather than installs?

But what I'm really saying is that it will be the price of flash cart versus pressed media that decides the media used for games. It's not going to be read speed. And to that matter I think we will see multiple sku's where one is DD only. Fundamentally I would prefer to buy a blu-ray based sku but should that not be available then I will go whatever is being sold For the pricing blu-ray for 25GB discs has become cheap and commodity, it remains to be seen what the cost of similar flash cards is once next gen comes.

Seeing the NGP flash card pricing, sizes and performance would be super interesting...

 

[Squilliam]

Another thing to bear in mind is that Samsung, Intel and TSMC are all transitioning to 450mm wafers from 2012. By 2014 this should have important implications for the price per chip and thus the price per GB of flash chips. By the end of the next generation, say in 2019 if we assume the next generation effectively starts it's first full year in 2014 we ought to be in the position where flash chips are cheaper and faster with higher capacity than any comparable flash chips today. We also should have had the internet infrastructure in most game buying nations to download large files swiftly and without issue.

 

[3dilettante]

TSMC's plan is a 450mm pilot line in 2013-2014. It would be another two years before production.

The question is what those companies have done to convince the equipment manufacturers to agree to the timetable, or if they are currently keeping the error bars on their roadmaps deliberately wide to allow for multi-year delays.

edit:
Similarly, Intel's 450-capable fab will start out at 300mm in 2013 and will need to upgrade at some point in the future if it is to support 450mm.

 

[AlphaWolf]

intel's building a fab that's compatible with 450mm but they don't expect tools till 2018.

http://www.eetimes.com/electronics-news/4211367/Intel-confirms-450-mm-fab-plans

 

[Silent_Buddha]

What about those of us that prefer to keep their games for eternity?

Quick reply before trying to catch up from 2 days of backlog here.

Since the media can be distributed on any media available you can burn it to your own BDR's if you want. And the game will STILL be cheaper including the cost of the blank BDR with a reuseable distributio media system (combined with DD as primary).

Hell, store it a Raid-0 array in a storage server if you want. Put it on a 99 USD 2 TB external drive if you want. Choice is entirely up to the end user.

Either way even if you choose not to, you can redownload it at anytime if you have uncapped broadband. Copy it from a friend if they kept a copy and you didn't. Whatever. You can spend as little (straight DD) or as much (from a physical location for a fee + cost of permanent media) and still have the potential to come in far cheaper than the current 1 game = 1 media distribution paradigm.

Regards,
SB

 

[Silent_Buddha]

http://www.anandtech.com/show/4078/intels-ssd-310-g2-performance-in-an-msata-form-factor

you can go with something in this form factor. Its write are slow but its using an older intel controller. A sand force controller should hit around 230MB/s .

Sandforce controller based SSDs would actually be a really really bad choice for this. They only excel in area's that feature data that can be compressed on the fly.

Game data is usually already packed and redistributable game's will be even slightly more compacted, if possible.

http://techreport.com/articles.x/20087/5

Shows that when copying typically compressed data nothing can touch the Mavell based Crucial C300. Same situation when playing games as game data is usually compressed.

After disregarding people "mentioning DD" as if that means anything, the only hard numbers you have is the questionable newzoo report that's probably just as accurate as vgchartz, and NPD report that PC game sales majority but not revenue majority have moved to DD.

My argument was not about whether the publishers want to move to DD, because of course they would. That's never in question. What's in question is that how much of the massive video games market is DD.

Here's a link as it pertains just to the US market, which is by far the most widespread DD market out there:
http://www.zacks.com/stock/news/46055/Video+Game+Sales+Decline+in+2010

From this piece, we see that PC software sales are insignificant, and a lesser dollar amount was spent on DD according to the NPD report (57/43) as well. It also says that used game sales, DD, casual games, facebook, and other digital content combined makes up about half of new game sales. Given that a vast majority of that is used games plus iOS games and facebook, this absolutely refutes your ridiculous claim that "DD has surpassed physical media in revenue for "traditional" style games." It's false, not even true for PC where it's close.

Therefore your outrageous claim that "DD has surpassed physical media in revenue for "traditional" style games" is absolutely 100% false. It's not even close, and not even true for PC.

Really that doesn't exactly say a lot. We already know that the survey data that NPD uses to project DD sales as well as the survey data that NewZoo uses to project DD sales are highly variable otherwise there wouldn't be a 9 billion USD difference between the two. As well you realize that the 9.36 billion represents PS2, PS3, PSP, Wii, DS, and X360? PC retail according to that is roughly 750 million. Going by the first half split between DD and Retail (which had changed by the time end of 2010 arrived) would put that at 1.32 billion. And that's the worst case scenario. Even NPD puts in a disclaimer that their estimates of DD are a work in process as they work to make its projections more accurate. And it becomes more obvious that it's a very conservative estimate as you start looking at various Publisher breakdowns of PC versus individual console revenue.

The NewZoo could be considered a best case scenario. And there DD+Retail PC is almost half of total console software spend.

BUT, and here's the key thing you continue to ignore.

There is no alternate DD distribution available for consoles yet. Thus there is no data as to how well DD may or may not do. But IF (yes, that's an IF) it is similar to PC, then it should take off fairly rapidly. PSP2 is going to be our first real look at how well console gamers take to DD as that is the primary distribution method and isn't being deliberately castrated as the PSP Go was.

You're going purely by a 6 month old NPD report using preliminary methods of DD estimate calculations. I'm using a combination of various sources up to the end of 2010, versus only using 1st half 2010 where DD having same price as retail does in fact negate some of the attraction for some people. Of course, december sales for Steam are a bit extreme in that even if we move to DD only it won't be at December sales prices. It'll be somewhere in between. Thus unit sales would be a more accurate portrayal of a DD only world. DD games will be cheaper than current retail games, there absolutely no doubt in my mind about that. As pointed out publishers/developers can instanly chop off 50% of the current retail price of games and still make MORE per unit sale than they do now.

Console makers want to move to DD, Publishers want to move to DD, Developers want to move to DD. PC users have shown they don't mind moving to DD.

So in a DD only world, the NPD data for PC revenue could quite likely drop, and more copies would be sold, and publishers/developers would more than double what they make now.

As mentioned the only remaining roadblock is finding a way to physically distribute the games such that it doesn't make it impossible to reduce the sale price of DD. And optical media, use once flash carts, etc. will ALL end up in a likely price increase. Publishers and console makers will have no choice but to raise the price of games if they do not transition to DD + reuseable distribution media.

So, would those beloved BRDs of yours still sell as well at 70 USD? 80 USD? 90 USD? Versus DD at 30 or 40 USD? With Retail redistribution services potentially adding another 10 USD on top of that if you don't have uncapped broadband and don't have any friends? An 80 or 90 USD BRD game should make the publisher/developer almost as much money as a 40 USD DD game.

Regards,
SB

 

[eastmen]

The facts of the matter is that alot of tech is coming down the pipeline that will drive prices of flash nand down while driving speed and capacity up.

450mm wafers , 3bit and perhaps 4 bit per cell tech , 22nm 18nm 12nm and so on. Prices of flash today in 2011 is not what you'd pay in 2012 or 2013 at the start of the next gen.


Flash ram has continued to move foward in terms of speed. It has gone from a few MB/s to 20-30MB/s and will go up.

Packaging costs will be lower with flash ram
Shipping costs will be lower with flash ram


The costs with bluray will stay pretty much the same.

 

[Silent_Buddha]

Oh and before someone claims that Steam only moves high numbers of small to mid budget games...

http://news.bigdownload.com/2011/01...ops-retakes-top-spot-on-steams-top-10-pc-gam/

And that's not even with sales to boost those as they are after the December push. And I think it was either Activision or Steam or an analyst that said COD: BO sold more on Steam than it did on retail (and that's completely ignoring the other DD services where it was sold).

Heck 7 of the 10 current top sellers are ~40+ USD. 1x at ~30 USD, ~15 USD, and ~10 USD. Once retail 1 media = 1 game retail distribution ends, I wouldn't be surprised to see that to change to 7 out of the top 10 to be ~30 USD or less. With similar types of titles.

Regards,
SB