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

It's like you people can't read. SIZE HAS A COST A smaller box will save money. It's not about aesthetics.

Will it save enough to pay for the extra cost of the media over the life of the console. I don't know, possibly.
I've only said it twice already on this page alone... it's not just the cost of the drive. You have to build a bigger noisier box that uses more power to support it.
So you can save some pennies or even a couple dollars, if you meant excluding the cost of the drive itself. Is that really why you keep arguing the same thing like its the single most important ever?
How smaller and less power-drawing (or even less noisier given that most comes from the fan) do you think the PS3 would be without Blu-Ray? - Keep in mind that your 25W value is off by an order of magnitude.

Its more an argument for handhelds, but not 100W behemoths
 
The PS3 could have launched $100 cheaper and a year sooner, where would Sony be if they had done that? But it's a stupid pointless argument, there's no reason in going there. I wouldn't suggest MS or Sony hold up releasing their console to wait for cheaper flash, if they're forced to ride optical for one more gen because of media costs, so be it. Optical needs to evolve to something much better (holographic or whatever) or die, it's going to be a bigger burden by the end of this next gen than DVD ever has been for MS.

I think you're (and others) seriously underestimating the cost including an optical drive. If your box is 10% smaller and lighter (that's if I used the 2.5W slim numbers that aren't anywhere near reality), shipping is 10% cheaper, but removing the optical drive will cut the size of the slim boxes in later years by much more than that. The optical drive is essentially a fixed cost, you can't make it smaller than the disc, electric motor efficiency isn't likely to improve (enough to matter).
 
So you can save some pennies or even a couple dollars, if you meant excluding the cost of the drive itself. Is that really why you keep arguing the same thing like its the single most important ever?
Well, in this thread it kinda is, seeing as money's the deciding factor. ;)

Alpha's suggestion is the savings in hardware (cost of drive, cost of storage and shipping due to smaller box) offsets the added costs of the games, which is true. He even tries to put a dollar figure on it of $50, an extra $5 per game for ten games. I think that's very optimistic on his part. I doubt the optical drive equates to $50 total cost of a $300 (and eventually $99) console. As the optical drive isn't causing $99 consoles to be lossy, it must cost next to nothing to include as part of the whole machine, and transport, storage, and other costs as a result must be pretty minimal.

Then there's the matter of how you manage that extra $5 per game. Does the console company take a $5 cut on license fees? Do they sell the console $50 cheaper and sell the games higher, offsetting the hardware losses with added game revenue?

I don't think storage capacity will be there, especially early on, and given the price per unit for flash I expect an optical drive and local persistant storage HDD and/or flash to solve the issues of load times. Flash works okay for Vita and DS where battery life and form factor is everything, but in a console the optical drive adds value with movie playback while being the most cost effective distribution format for mass-pressed games.
 
Continually repeating it doesn't make it true.

I don´t think it´s a secret that i would prefer Blu-Ray for both personal and reasons that i think would yield me and everyone else better games than flash. I would on the other hand have no problem with being proving wrong since i am not married to either technology.

But i don´t think you so far has come up with an answer to the "ever repeating point" of how do think multiplatform games can take advantage of flash speeds when they have to abide by lowest common denominator?
 
Sony_XQD_write_speed_chart.png
 
Cell Type and cost per GB:
SLC $3.00
MLC $0.90
TLC $0.60

http://www.anandtech.com/show/5067/understanding-tlc-nand

Now I guess you could argue that a reasonable reduction in cost over the next two years is about 33%, is that fair? That would mean your typical flash disc cost would be in the order of $4 for 10GB plus say another $1 for manufacturing etc.

By say 2019 I guess you ought to expect that the price per GB would have halved again so your standard $5 buck 10GB game from late 2013 ought to have 20GB of data available by 2019.

Some of the cons aren't as bad as they appear. For instance you'd save some of the extra costs from the media itself with lower shipping and handling costs. You'd also have the ability for a game store to 'print on demand' games so that helps with stock control issues as well.

The final major issue as I see it is the outright level of affordable capacity for any title. Will say 10GB be enough for a major game release on a next generation console in 2013 and how will the cost/capacity factor change as the generation progresses?
 

what is that science-fiction charts?
what I've read lately is that article about playstation vita memory card speeds.
but amazingly it went offline.

http://www.eurogamer.net/articles/df-hardware-how-fast-are-vita-memory-cards

transfer rates reported to be measured were in the 7MB/s to 8MB/s range, down to 4MB/s sometimes for write speeds.

but, here's a whole thread about it, with many comments and a few quotes from the article on page 2.
http://www.neogaf.com/forum/showthread.php?t=465050

tests were done with transfer over USB, which maybe should be a remote reason for slowness but amazingly I learn that the Vita doesn't allow to be seen as mass storage, it is the one that initiates transfers and oversee it no matter what.

people complain of slow loading screens too.
here is it, premium price and bottom of the barrel flash
which shouldn't be surprising, any flash mass distribution will be like that speed wise. cannot be faster than a plain USB stick, not be like one branded "traveller" or "XXX GT force" or what have you. though here it seems slower still, and running an live OS from a USB stick is not that bad of an experience.
 
alright, compact flash is IDE, and I've checked that XQD it uses a new interface based on PCI express. also IBM even used to sell compact flash hard drives. thanks wikipedia article.

so here, I am assuming these figures are for SSD in a compact flash form factor.
price is through the roof and what I assume is CF type A claims being an UltraDMA/133 peripheral.
http://www.sears.com/shc/s/p_10153_...9x00001a&ci_src=14110944&ci_sku=SPM6661469802

if your point was that a ssd or raid 0 like configuration for flash is fast, then it is fast, but not sold for pennies or even dollars right now.
 
Cell Type and cost per GB:
SLC $3.00
MLC $0.90
TLC $0.60

http://www.anandtech.com/show/5067/understanding-tlc-nand

Now I guess you could argue that a reasonable reduction in cost over the next two years is about 33%, is that fair? That would mean your typical flash disc cost would be in the order of $4 for 10GB plus say another $1 for manufacturing etc.

By say 2019 I guess you ought to expect that the price per GB would have halved again so your standard $5 buck 10GB game from late 2013 ought to have 20GB of data available by 2019.

Some of the cons aren't as bad as they appear. For instance you'd save some of the extra costs from the media itself with lower shipping and handling costs. You'd also have the ability for a game store to 'print on demand' games so that helps with stock control issues as well.

The final major issue as I see it is the outright level of affordable capacity for any title. Will say 10GB be enough for a major game release on a next generation console in 2013 and how will the cost/capacity factor change as the generation progresses?

They could also use bigger wafers for cheaper chips also . There is also talk of 1xnm nand going into production this year . Anand doesn't talk about what they were looking at but micron for instance has been using 25nm nand but has switched to 20nm process which apparently has a 30-40% reduction in board space , that was supposed to go into production late 2011 so we may not even be seening the full cost savings of that move yet
 
I don´t think it´s a secret that i would prefer Blu-Ray for both personal and reasons that i think would yield me and everyone else better games than flash. I would on the other hand have no problem with being proving wrong since i am not married to either technology.

But i don´t think you so far has come up with an answer to the "ever repeating point" of how do think multiplatform games can take advantage of flash speeds when they have to abide by lowest common denominator?


I'm not really sure on your personal reasons. I rather just buy a 4k bluray player (although i'm done with discs ) or just wait till the next mpeg format comes out with that support in it and stream it .

Regardless ,


With bluray alot of that omg amazing space will be used to repeate textures and data for optimal reading speeds . You don't need that with flash , random access will be many times faster than with bluray .

So now instead of having 10 gigs of textures repeated multiple times , you can have them stored once.

As for worryin gabout the lowest common denominator , that may be a problem however who knows we might have a psone or ps2 era console where one is a huge clear leader and lots of exclusives get developed for it by default.


Personaly I rather spend $5 more a game and save $50 a console and have a more enjoyable experiance.
 
I'm not really sure on your personal reasons. I rather just buy a 4k bluray player (although i'm done with discs ) or just wait till the next mpeg format comes out with that support in it and stream it .

Regardless ,


With bluray alot of that omg amazing space will be used to repeate textures and data for optimal reading speeds . You don't need that with flash , random access will be many times faster than with bluray .

So now instead of having 10 gigs of textures repeated multiple times , you can have them stored once.

As for worryin gabout the lowest common denominator , that may be a problem however who knows we might have a psone or ps2 era console where one is a huge clear leader and lots of exclusives get developed for it by default.


Personaly I rather spend $5 more a game and save $50 a console and have a more enjoyable experiance.

Since i already have the PS3(4) connected to a big screen another 4k separate player or just standard player is a waste for me. Streaming is just not an option, lots of reasons, most important, i want to own a physical copy and i think streaming quality is pretty bad as is, with 4K i would expect it to be worse.

My point about Harddrive installations seem to have been ignored :)
 
Cell Type and cost per GB:
SLC $3.00
MLC $0.90
TLC $0.60

http://www.anandtech.com/show/5067/understanding-tlc-nand

Now I guess you could argue that a reasonable reduction in cost over the next two years is about 33%, is that fair? That would mean your typical flash disc cost would be in the order of $4 for 10GB plus say another $1 for manufacturing etc.

By say 2019 I guess you ought to expect that the price per GB would have halved again so your standard $5 buck 10GB game from late 2013 ought to have 20GB of data available by 2019.

Some of the cons aren't as bad as they appear. For instance you'd save some of the extra costs from the media itself with lower shipping and handling costs. You'd also have the ability for a game store to 'print on demand' games so that helps with stock control issues as well.

The final major issue as I see it is the outright level of affordable capacity for any title. Will say 10GB be enough for a major game release on a next generation console in 2013 and how will the cost/capacity factor change as the generation progresses?

It will be like in the old days, where big games were expensive, already as today we have huge games that easily break the 10GB limit, and some of those already had to compromise.

LA Noire is a 20GB game, Final fantasy XIII is around 18GB, so those games would just "$5" by dollars to the price.
 
It is funny I waded into this thread as I have a Word DOC open right now and my bullet points all align with AlphaWolf. I don't think anyone is pretending that a Flash format trumps BDR or other optical formats across the board (hardly). But looking at some of the challenges of upcoming consoles and then looking at the cost/benefit of Flash vs. Optical I think creates a scenario where solid media has some merits.

Major Issues:

* Load Times. Optical drives at this point in technology are not addressing load time issues. There are probably *serious* end-user experience benefits to a platform that put near-instant load times as a platform experience priority. With consoles with 2GB (or more) load times must be a massive concern in the MS/Sony camps. Even the fastest optical drives are slow; designing around load times is a design issue and developers will need to get used to streaming in chunks and designing with massive loads ahead but that doesn't mean this should not be address on the HW side to some degree. Streaming installs are one path yet while the PS3 BDR was CAV that doesn't necessarily address the issue of a lot of small files, which leads to...

* Seek Time. Packing and redundancy play a part in disk layout as a ton of small file seeks really slows down an optical drive. A solid Flash media would be at least a minimal step in the opposite direction.

* Power Draw. Power is a big issue with new consoles--while transistor density is doubling every ~ 2 years the power consumption is not reducing at a similar rate (often in the 30-40% range). This means if you are capped at TDP, and not die area, you are getting only a 30-40% performance gains for TDP every 2 years or so. I could be wrong but it looks like Blu Ray movie playback is about 10W (cf. the 1080p network stream to the Blu Ray play back). For gaming this will be higher as higher speed drives will be needed to transfer data for the assumed larger memory pools to accommodate next generation assets.

* Size & Cooling. Optical drives take up a lot of room. Optical drives do not undergo significant size reduction. This issue slices two ways. There is the "Wii" sized console where removing the optical drive results in a smaller console which means a smaller, lighter price that (a) appeals to certain consumers and (b) has real impact on shipping and shelf space. The other direction, in-line with the issues of power, is cooling. Keep the volume footprint of the optical drive and shift it toward better cooling layouts.

* Costs & Licensing. Optical drives have a lot of attendant costs. There is the manufacturing and assembly. There are the associated costs the space they consume in the design which ends up being a relevant percentage volume packaging space and weight. And then there are the licensing issues. As a project leader if you were staring down $50 for an optical format for bulk product costs and that this consumes 17-12% of your product budget ($300-$400 product) with the slowest reductions you have to at least consider what else that could be spent on and how it may benefit the user experience.

* Reliability. Optical drives and the media are not the most reliable formats. Scratched and smudged disks are annoying; broken consoles due to drive errors can turn consumers away.

* Noise
. A very fast optical drive, which a new console will need, is going to be loud.


I would not argue Flash as a panacea for these issues. Memory across the board is going to be a major design issue on future consoles. We may very well see one company go a tiered format (1GB very fast memory, 2GB slower memory, a local storage Flash/SSD/HDD, and then media storage (optical, flash)) and another go with a single larger pool (2-4GB) and then local storage and media storage (optical, flash). We don't know the new console time lines but we do know that DRAM densities (thus number chips) as well as pad limits on the chips themselves pose issues. If things play out much longer there may be some solutions (Silicone Interposers, Stacked Memory, 3D chips with memory stacked and connected with TEVs, etc; DDR4 is on the horizon) but none-the-less memory systems are a major concern across the board because of manufacturing issues, costs, reduction prognosis, and performance pitfalls.

Physical memory could play some part in the solution. Flash increases density about 30-40% a year and this projected to continue pretty much throughout much of the lifetime of the next console (assuming 5-8 year cycle). I see, in early 2012, 16GB consumer purchasable Flash products at $10. They are junk but looking at a 2014 console and the advantage of a major gaming company ordering tens of millions a month to spec has advantages. Ok, that is a big cost, no way around that.

But lets be crazy. Lets say I am MS (not Sony, as I am tied to Blu Ray) and I am aiming at this sleek, quiet console and leveraging a prompt gaming experience as a core feature. On the one side you just pin your similar consoles at the same price point, reap the advantage of saving $50 a unit for no optical drive, and use them as a media allocation where you offer certain subsidies for the game-media (e.g. $5/game; 10 game attach through a generation and you are looking at a pretty level cost). There would be issues of manufacturing, overhead costs, etc that are very real, no doubt. You really are looking at a 4x cost of goods on manufacturing. On the other hand you just opened the door for games not as large to shave some costs and some high end games going larger / faster media. It would be an interesting dynamic.

If you take the crazy plunge there are some not-so-crazy augmentations you could do. One is that Flash, unlike Optical, is re-writable. Why not allow users to bring their Flash Drive into GameStop/Walmart/Target/BestBuy and download the content and purchase a "game card" in store (or unlock at home)? Maybe this opens the door for a snazzy demo-kiosk where instead of the huge isle of blank game disks you have 6-8 32" screens with all sorts of running demos, previous, adverts, and "purchase" slots. Walk up, select your game, insert your drive, purchase (% going to the store) and done.

This is only really an extension of DLC anyways.

Which is another major component: Who is buying a $400 console in NA in 2014? Most likely someone with solid internet. I know not everyone, but this shifts back to the cost/benefit of the platform design. If you are strongly targeting online consumers of games, social gaming, movies, music, etc who are happy to purchase content digitally do you (a) cut huge manufacturing costs to shift toward peppier performance and larger drives and (b) inflate brick and mortar costs (to you and consumer) due to a Flash media storage?

If it means cutting my load times by 70% I at least consider this scenario.

I am not saying these ideas work as is. I am not saying they would work at all.

I would say that, from a gaming perspective with some of the issues ahead, two platforms that have near identical budgets for chips and box-volume, removing the optical drive and using those budgets for faster drives and faster chips (or shifted completely over to another product concept, e.g. user input devices) could be a game changer. You do lose BC for physical games and the device is no longer a standalone media-center for those invested in BDR/DVD (yet I constantly see people with a DVD or BDR plus a Wii, PS3, and or Xbox 360 all on their entertainment center) and this would be very unpopular for many due to physical media cost increases, the console being seriously aimed at those with strong internet connections, a big blow to used content, not being that all in one media box, and so forth.

Yet I find it sobering that a lot of the concerns of losing an optical drive are less gaming related. End user cost for a physical media really sticks out, but who would pay $5 more per game for better load times and general performance improvements due to a better media format?

I think all consoles will have optical drives next gen. While I will continue to chuckle at those who point out the small number of 360 games that needed more than one disk (no one ever said games would never need spanning! just it wasn't going to be a systemic limiting issue in the projected 5 year life-cycle which has proven to be one of the most true predictions this entire generation for the general industry) the reality is easy to see that games on faster hardware will be routinely pushing well beyond the DVD limits and up to and beyond 25GB BR disk limits if care is not taken (sure, there are work arounds like like DLC HD packs meh meh meh and some of the benefits of better media packing but again Flash is no panacea). And although a rough start Megatexture and the like seem like a viable software design concept which shifts the burden to media. Optical drives may be slow and still require a HDD install but it isn't hard thinking of a next gen product using Megatextures stretching out toward 50GB.

That said I would probably root hard for a console that went with a Solid State drive of some sort, with the goal of addressing load times and performance issues, and sacrificed optical media one way or another.

Just my 2 cents.
 
Btw, I am old enough to remember when the larger N64 games cost more. You may pay $55 for one game but $65 for one that was on larger carts. But the market has really changed where, again, consumers may be able to re-use media, DLC, etc.

Interestingly, considering how married one side is to media (Sony with CD, DVD, and then BDR) which has fostered a certain ecosystem among their users I would say if the other party, MS, dropped optical it would be the single most interesting/defining/acrimonious issue on forums :p

Cell Died. Intel ran away from everyone but bottled themselves up into their market segments. Everyone ran to AMD for console graphics leaving NV in the rain (supposedly). Fabs got spun off by Sony and AMD and everyone is kicking the wheels on contracted fab work. The big idea deep vault products died, KK and Allard are gone. Media and ecosystem nuances flourish and there remains this tension as media moves digitally away from formats but also away from consumer control (which will happen with physical media as well imo, sadly) and there is the future (everything digital) and the traditional (physical media) and there is this fuzzy transition and what makes sense for products projecting 2014-2022 life spans.

I am biased (I don't own a Blu Ray drive and I am happy with HD streaming for my limited media access) and I see load time issues as a major end-user-killing experience but I still think every console will get an optical drive. Yet I think this will be an interesting territory to review, especially in, say, 2017.

I have been told by 3 developers now: You are crazy for wanting more than 2GB of memory with an optical drive. If systems (a) go with less memory due to the practical transfer limits of optical and (b) inflated load time issues due to these optical drives I think that is just a horrible design choice. Ok, so you then add solid storage and begin caching everything or increase product costs with hybrid HDDs, etc.

Maybe someone could convince me how an optical drive makes a console experience a LOT better and how an optical drive address load times, could justify 4GB of memory (you have to populate it somehow--and don't tell me in 2017 developers will be happy with 2GB and it doesn't impact games and how it was a great decision to limit it to shorten load times because it is bull and points toward constraining the platform for a design choice), etc. The obvious solution is Optical>>Longterm Storage / cache / System memory and pretty much turn consoles into PC installs and create guidelines that force developers to push forward a gaming experience while the games install in the background. What a mess any choice will be :p
 
Maybe someone could convince me how an optical drive makes a console experience a LOT better and how an optical drive address load times, could justify 4GB of memory (you have to populate it somehow--and don't tell me in 2017 developers will be happy with 2GB and it doesn't impact games and how it was a great decision to limit it to shorten load times because it is bull and points toward constraining the platform for a design choice), etc. The obvious solution is Optical>>Longterm Storage / cache / System memory and pretty much turn consoles into PC installs and create guidelines that force developers to push forward a gaming experience while the games install in the background. What a mess any choice will be :p

Convince.. well that is a tough one, but as i mentioned earlier, if we take for granted that next gen will have harddrives i see the harddrives speed as the limiting factor, there is no reason why background installation and just outright installation should be a problem with games on a console, it´s already a part of PS3 games today. If we take the proposed average sized flash game of 10GB, and the proposed 10 games attach rate, that collection would take up 100GB on a harddrive. Even 10 x 25GB games would only be 250GB.

So the Optical drive would not be the limiting factor, you would have to compare flash to hard drives even with optical drives. Which imho takes some of the flash out of flash memory.

The compromises already being done today in order to squeeze down games to 6.5GB would just get bigger as the costs is raised as the game sizes grow. Didn´t final fantasy leave Nintendo because of the Cart limitation?
 
To clarify, are your suggesting taking the already generally poor load time issues, tossing in the PS3 installation load-time experience (which many loath as you know), compounded it by larger memory pools and larger games for next gen, and deposited that as a convincing argument. You don't see why that would be a problem on consoles :p

The HDD limiting scenario makes sense if you first assume the optical drive. Flip the scenario with some sort of fast solid mass storage I am not sure that works anymore. I am not saying it isn't a solution, or even one that will be taken (I think it will), but you are turning a console into a PC at that point. Which doesn't address one of the major design issues consoles are facing. Basically it is throwing up of hands for the instant-gaming / ease of use approach.
 
* Costs & Licensing. Optical drives have a lot of attendant costs. There is the manufacturing and assembly. There are the associated costs the space they consume in the design which ends up being a relevant percentage volume packaging space and weight. And then there are the licensing issues. As a project leader if you were staring down $50 for an optical format for bulk product costs and that this consumes 17-12% of your product budget ($300-$400 product) with the slowest reductions you have to at least consider what else that could be spent on and how it may benefit the user experience.
This is just plain unrealistic. An optical drive can be included in a proftable sub $100 console. It can be included in a £20 cheapo DVD player. The cost of assembly, distribution, storage, that adding optical contributes, has to be minimal and not $50 (eventually at least, assuming your not launching a new format). Making all subsequent cost considerations plain wrong. DVD in XB360 didn't cost MS $50. BRD in PS4/XB720 won't cost $50.

So the $5 extra per game that you might absorb suddenly starts costing by the 5th game, and loses you an extra $25 per unit over the life of the platform.

There are of course options, like increasing game costs and whatnot, but some of the considerations being presented don't seem very grounded to me. Optical drives costing $50 and savings on storage and transport being substantial are views that just don't seem supported looking at the real markets. Real saving are perhaps more like $30 in total for not including optical drives (excluding warranty costs which are hard to gauge) and the real costs of flash are probably more like $7 more over disc depending on what games end up looking like. If we have megatexturing as standard to save on RAM costs, we'll need optical storage and installs, plain and simple, unless games get even shorter!

Which I think sums up the debate. The issue cannot be resolved in isolation, but has to be considered in relation to platform strategy. Do you go with a more expensive box box with local SSD storage and optical distribution and maybe less RAM, or more RAM and more expensive games on flash? And various other options.
 
You are now not even talking about the same situation. You think a high speed BDR drive in 2014, plus licensing, and the associated design and assembly costs compared to an optical-free design is far less than $50? This isn't a Walmart special min-spec Blu Ray player but essentially a high performance box with ~200W of chips (and required cooling), a mass storage device (HDD), ports, etc. that you then want to toss in a space hogging optical format. Which, while only retail, the faster drives are not cheap. $50 may be high but iSupplies folks often indicated how much people under-balled these things. Call it $40 or $35, whatever, but this is not trying to push a cheap-o commodity drive like the 360/Wii had at launch; for reference compare how the Xbox just 4 years earlier had a *fee* to unlock DVD playback due to licensing. That situation, in 2001, is more similar to 2014 with Blu Ray than 2005 with the Xbox 360.

We may just not see eye to eye above, but in terms of long term platform reduction the HDD and the Optical Drive merit special consideration as they consume space and undergo the least aggressive cost reductions. Those are big hurdles for companies who may fancy their console as a cheap set top box down the line. Once you commit to optical you are stuck with it; once you commit to mass storage you are stuck with it (less of a problem now with cheap onboard flash solutions). It isn't just about cutting whatever costs out of the launch but also market movement down the line. (Which the safe betting man looks at the costs, merits, and risks and throws one in no matter what just to avoid missing the check box and the fury of those scorned as the major players are all really attached to fulfilling the market instead of aggressively moving in new directions like a more disruptive product like an iPad).

As for Megatextures, how many games are implementing such (not many)? What are the draw backs (a lot for some methods)? RAM is very cheap, fast optical and HDD are not. Optical drives are not even a great Megatexture solution as you are resigning yourself to large HDDs for HDD installs because the poor optical drive performance.

I don't think it will happen this gen, but at some point a platform (probably when online infrastructures mature) a platform will cast optical drives off and I think it will be readily apparent the large albatross they are in terms of performance and costs. Right now the technology is actually there to do such, just not the infrastructure for either MS or Sony to risk losing customers just based on access and the requisite negative word of mouth.
 
Back
Top