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

[MrFox]

The two latest estimates I read were $0.11 and $0.15 per GB in 2020.

If you still think next gen launches in 2018 (which I don't), maybe 0.20 per GB at launch for blanks, plus the cost of replication which is a problem if there is 100GB to copy and verify, compared to a bluray stamper doing over 1000 discs per hour.

Average game size difference between PS3 and PS4 is supposed to be 2.5x for the multi-gen games. If the same progression happens next gen, should we expect 100GB for a AAA game?

 

[Rikimaru]

The two latest estimates I read were $0.11 and $0.15 per GB in 2020.

If you still think next gen launches in 2018 (which I don't), maybe 0.20 per GB at launch for blanks, plus the cost of replication which is a problem if there is 100GB to copy and verify, compared to a bluray stamper doing over 1000 discs per hour.

Average game size difference between PS3 and PS4 is supposed to be 2.5x for the multi-gen games. If the same progression happens next gen, should we expect 100GB for a AAA game?

BDXL 100/128Gb exist since 2010. I think PS4 and Xbox One could read them but they are not as cheap as 2 layer BD.

 

[AzBat]

BYOS - Bring your own storage. The game is already at the store. Just copy it over to your own device(USB, SATA, SD card, mobile phone) or buy an official storage device. It's not unlocked until you pay for it online or enter a code given to you at the store. The only thing MS/Sony/Nintendo have to do to is print the paper cards/gift cards with scratch off codes or the store prints the code on a receipt. Forget shipping the games already on flash storage or whatever kind of medium you can think of. Don't want to wait just over 2 minutes for them to copy 50Gb unto your USB 3.0 device? Then I'm sure you can pay a little extra to get it on something pre-loaded.

Tommy McClain

P.S. Here's something neat

 

[Shifty Geezer]

Also why limit us to SD cards ?

Form factor and price. The actual tech is likely something else.

$13 for 32 gig usb. That's as fast as a 2.5 drive for reads. I really think you guys are over estimating the cost

No, the cost is still the issue. The conversation at large has generally looked a $5 flash storage versus $0.20 disc*. This USB device is ~20-50x more expensive than a BRD. That and the transfer speed, which perhaps might not be too much of an issue. (*edit: not sure what the figures being thrown around are. <$1 is definitely the mass-production cost)

Anyway when we talked about DD games I talked about how nice the plastic cards are for retail. I'm sure something slightly bigger clam shell style would work for a flash type game cart. Retail would love it as they'd be able to fit many more into the space as multiple copies from other companies.

Then why isn't Nintendo doing that already? According to your packaging+distribution cost theory, Nintendo are throwing substantial amounts of money away on oversized boxes. If they could save tens of millions with smaller boxes, why aren't they?

BYOS - Bring your own storage.

I'm pretty sure that was one of the earliest suggestions in this thread! I certainly remember talk of in-store kiosks. Actually, the opening post is quite eye-opening. 256 GB SSD in 2008 and the notion that that'd be $20 in 2013. Also, Rangers just bought a 4GB USB drive for $15. 7 years later we're talking about 32 GB for the same money. 4 GBs can be found for $2 on Amazon. So it'll likely be 5+ years from now before 32 GB at speed is just a few dollars versus a disc, which is still a significant cost. Although I reckon an extra $2 loss per game sold would be worth it with a cool-ass cart distribution system!

 

[MrFox]

BDXL 100/128Gb exist since 2010. I think PS4 and Xbox One could read them but they are not as cheap as 2 layer BD.

BDXL was not available in ROM form until bluray 4K was finalized. They experimented with up to 16 layers, and the conclusion seems to be that it will be 3 layers in the foreseeable future to keep the cost below $1 per disc.

I think the cost was only 10 cents per additional layer, but each layer have 90% yield. That makes 16 layers $15 each, which is unusable. While 1, 2 and 3 layers would be something like $0.80, $1.00 and $1.25.

If they needed more than 100GB they could press the 300GB, 500GB and 1TB discs from the bluray archival format, but without a home video format using it, there is no justification for the additional cost. It's the reason we had BDXL in 2010 but it took 5 years before it appeared in ROM form for bluray 4K.

 

[Rikimaru]

BDXL was not available in ROM form until bluray 4K was finalized.

It was available, just nobody had to use it.

 

[MrFox]

It was available, just nobody had to use it.

Yeah the drives could read that and the specs existed, but singulus and other companies making optical stampers didn't have any machine available to allow mass production. Chicken and egg problem.

 

[Rikimaru]

Also I think majority UHD BDs will use 2 layer 50 and 66 GB discs.

 

[eastmen]

Form factor and price. The actual tech is likely something else.
No, the cost is still the issue. The conversation at large has generally looked a $5 flash storage versus $0.20 disc*. This USB device is ~20-50x more expensive than a BRD. That and the transfer speed, which perhaps might not be too much of an issue. (*edit: not sure what the figures being thrown around are. <$1 is definitely the mass-production cost)

This is with multiple mark ups . Its $13 to sell to a customer. That's like looking at a $30 bluray and saying well that's to expensive for a console maker. I.E it makes little sense . When your buying in the 100s of millions the cost will be much lower

Then why isn't Nintendo doing that already? According to your packaging+distribution cost theory, Nintendo are throwing substantial amounts of money away on oversized boxes. If they could save tens of millions with smaller boxes, why aren't they?

For 3DS Nintendo already has the smallest boxes on the market , they have already introduced eco friendly boxes for both the wii u and 3ds that features even less plastic and thus costs less. With a new system we may see new smaller boxes

I'm pretty sure that was one of the earliest suggestions in this thread! I certainly remember talk of in-store kiosks. Actually, the opening post is quite eye-opening. 256 GB SSD in 2008 and the notion that that'd be $20 in 2013. Also, Rangers just bought a 4GB USB drive for $15. 7 years later we're talking about 32 GB for the same money. 4 GBs can be found for $2 on Amazon. So it'll likely be 5+ years from now before 32 GB at speed is just a few dollars versus a disc, which is still a significant cost. Although I reckon an extra $2 loss per game sold would be worth it with a cool-ass cart distribution system!

Yes we talked about it before. Its interesting to note that we can now find 128 gig SSDs for $40 at retail in 2015.

The problem with a 4GB usb drive is that its using older flash nand that hasn't been shrunk in a very long time because it doesn't make sense to shrink it. So your just seeing old stock that is sitting there. You can see the price drops in the higher catagories . The 64 gig micro sd cards went from $200 in 2012 to $21 in 2015 . This is before nand stacking also.

 

[Rikimaru]

For 3DS Nintendo already has the smallest boxes on the market , they have already introduced eco friendly boxes for both the wii u and 3ds that features even less plastic and thus costs less. With a new system we may see new smaller boxes

Have you seen Vita boxes? I think it's half volume. Wii U cases are not standard BD size.

 

[eastmen]

Have you seen Vita boxes? I think it's half volume. Wii U cases are not standard BD size.

I don't believe those exist at retail !

Honestly I'm not sure how big the vita boxes are. I used to own one a while ago but all my stuff was DD .

 

[MrFox]

Eastmen...
http://www.dramexchange.com

You see the insane fluctuation? When a company needs millions of chips the price goes up. When there's an overstock it's getting dumped. If they need a hundred million chips in a guaranteed steady supply they will not get the dumping price. They will pay a premium for the guaranteed volume.

Emmc contract price is currently averaging 6.20 for 16GB.

 

[eastmen]

Eastmen...
http://www.dramexchange.com

You see the insane fluctuation? When a company needs millions of chips the price goes up. When there's an overstock it's getting dumped. If they need a hundred million chips in a guaranteed steady supply they will not get the dumping price. They will pay a premium for the guaranteed volume.

Emmc contract price is currently averaging 6.20 for 16GB.

I would think someone gearing up to provide nand for their console games which would result in hundreds of millions if not billions of products shipping with it , they would either open their own fab for it in a joint venture or have enough purchasing power to control their prices much better than the dramexchange.

 

[MrFox]

Making nand that is technologically competitive with samsung, and more cost effective, is not just a question of investing in fabs. The cost of entry is gigantic. The R&D is also gigantic.

Otherwise anyone would do it. Even Apple doesn't make their own flash and pay a premium for it (as can be seen in iSuppli BOM analysis).

If it was that easy, Sony would just make their own GDDR5 instead of paying $88 for each PS4.

 

[Shifty Geezer]

I would think someone gearing up to provide nand for their console games which would result in hundreds of millions if not billions of products shipping with it , they would either open their own fab for it in a joint venture...

Why didn't Nintendo create their own ROM pressing fab for producing DS/3DS carts then? Establishing suitable component manufacturing is, as MrFox says, expensive. That's why companies outsource. Sony buys in a lot of components for it's own products, competing with itself for the best value so we've been told.

...or have enough purchasing power to control their prices much better than the dramexchange.

I'd have thought if the market has the demand, it doesn't matter how big a client is. The only time Nintendo or anyone else could command a significantly better price than the market is if there's excess product available. Otherwise, Nintendo will be competing with other consumers. Who's going to sell 10 billion chips to Nintendo at $4 a pop when the market price is $6.20 and there's other customers willing to pay that? A contract for ongoing supply would have some bargaining value where the market is volatile as stability will be worth something. That'd be determined by tech used.

 

[Lalaland]

No-one is entering into a joint venture to produce NAND for their console, if the console flops you're left with a giant NAND plant that will now be flooding the market with NAND. Sony already has a number of fabs but those largely supply the low volume, high margin imaging sensor business there's no incentive for them to get into low margin, high volume NAND trade.

There have been rumours for years that Apple was going to get into the semi business for their devices but every time they look at the numbers and say 'naw let's contract this'. If Apple who can pretty much sell a brick-in-a-box as long as they put their logo on it think semi is too much risk that should tell you a lot. As Shifty has pointed out even Nintendo, whose mobile business has been untouchable for decades, have never thought about building their own fabs (or establishing joint ventures). In using contracts they get to use the latest processes at far lower levels of risk and walk away if things go south.

 

[eastmen]

Why didn't Nintendo create their own ROM pressing fab for producing DS/3DS carts then? Establishing suitable component manufacturing is, as MrFox says, expensive. That's why companies outsource. Sony buys in a lot of components for it's own products, competing with itself for the best value so we've been told.
I'd have thought if the market has the demand, it doesn't matter how big a client is. The only time Nintendo or anyone else could command a significantly better price than the market is if there's excess product available. Otherwise, Nintendo will be competing with other consumers. Who's going to sell 10 billion chips to Nintendo at $4 a pop when the market price is $6.20 and there's other customers willing to pay that? A contract for ongoing supply would have some bargaining value where the market is volatile as stability will be worth something. That'd be determined by tech used.

Well I said joint venture. There are already joint ventures formed like intel and micron for nand capacity.

You would need to invest a lot of money into the joint venture sure. Apple hasn't done it for nand yet but they do design their own chips now and are bringing more and more in house to save costs.

one thing you have to remember is time frame. Yes in 2015 they could get $6.20 cents for it instead of $4 but who are they going to sell those chips to in 2020 when people have moved onto greater capacities and formats.

ITs the same reason a company is out there making the xdr ram in playstation 3s.


I also doubt you'd have to worry about the nand plant. MS would be able to use it for console games , surface , windows phone and a host of other devices they sell.

 

[MrFox]

one thing you have to remember is time frame. Yes in 2015 they could get $6.20 cents for it instead of $4 but who are they going to sell those chips to in 2020 when people have moved onto greater capacities and formats.

ITs the same reason a company is out there making the xdr ram in playstation 3s

Elpida made XDR for sony... and they filed for bankrupcy in 2012 with about 6 billion in debts. Wow, why doesn't Sony make their own memory? Or imagine if Sony went into a joint venture with Elpida? They would have an additional 3 billion in debts over the PS3. Now instead they just continue the contract with Micron, who purchased Elpida assets in japan.

If a company ends up with 6 billion losses making memory, it means they were selling the memory at a lower price than their operating costs.

Sony makes the best imaging sensors in the world, but the PS4 camera is using omnivision sensors. The sensor fabs of Sony make high-margin parts, and they'd lose money if they made the ps4 camera sensor themselves.

 

[eastmen]

Elpida made XDR for sony... and they filed for bankrupcy in 2012 with about 6 billion in debts. Wow, why doesn't Sony make their own memory? Or imagine if Sony went into a joint venture with Elpida? They would have an additional 3 billion in debts over the PS3. Now instead they just continue the contract with Micron, who purchased Elpida assets in japan.

If a company ends up with 6 billion losses making memory, it means they were selling the memory at a lower price than their operating costs.

Sony makes the best imaging sensors in the world, but the PS4 camera is using omnivision sensors. The sensor fabs of Sony make high-margin parts, and they'd lose money if they made the ps4 camera sensor themselves.

Mostly because they picked the wrong tech .

XDR ram was used in what ? The ps3 ? It was a huge failure just like cell was.

Nand is used across a huge amount of devices and money helps develop newer versions of the tech. Getting another large company involved in a venture to improve and manufacture the tech will drive costs down for the venture and increase funding for development of advances. It works fine in a large number of joint ventures.

Sony failing is just what sony has done for the last half a decade or more. Everything with PS3 was a horror show in terms of execution.

 

[Shifty Geezer]

It works fine in a large number of joint ventures.

Joints ventures are about companies with separate skills coming together to use their expertise together in a joint venture. A company that already can do something has little need of creating a joint venture. What would a RAM manufacturer have to gain by partnering with Nintendo or MS? In those cases you'd look more to just investment or a contract. But certainly regards numerous joint ventures

Sony failing is just what sony has done for the last half a decade or more

What is Sony failing in this instance and what are their last half decades of failures?