Planar NAND Development Ends After 26 Years
Lucas123 writes: The non-volatile memory used in thumb drives, SSDs, smartphones and any other mobile device today has at last hit an engineering wall. The major developers of planar NAND this week said now that they've reached 15 or 16 nanometer process technology, they no longer expect to shrink their lithography process any further, as the capacity and economic benefits no longer make sense. Toshiba, which produced the first NAND flash chip in 1989, SanDisk, Intel and Micron said they will turn their engineering efforts to 3D flash trap NAND, 3D resistive RAM and other vertically-stacked non-volatile memories that offer a much longer road map. The manufacturers all said they'll continue to produce planar NAND while developing 3D NAND, which has already doubled previous capacities while also offering two to 10 times the erase-writes of previous non-volatile memories and twice the write performance. Intel and Micron are also producing a 3D NAND, based on floating gate, and a ReRAM that the companies say will increase performance and endurance 1,000 time over planar NAND. Toshiba and SanDisk have come out with a 48-layer 3D NAND that could allow them to produce 400GB microSD cards next year.
I'm still waiting for the unified address space I was promised twenty years ago... So, yes.
Ezekiel 23:20
Despite these advances, IPhone 7's will still be just 32 or 64GB, with a ridiculous upcharge for the 64GB version...
Once you have a unified address space, you'll pretty quickly realise that: 1) the definition of "unified address space" has moved on enough that you still don't have one, and 2) you still want to segment those address spaces for reasons other than performance and the physical limitations of volitility
-- 'The' Lord and Master Bitman On High, Master Of All
I don't know how slow they'll be, but every one of them will incur royalty payments to Microsoft because the SD spec requires all cards larger than 32 GB to be formatted in Microsoft's patented exFAT file system.
I was just thinking to myself how awesome it would be to have a 1 petabyte micro SD card, but then realized, "What could I possibly use that much storage for?" Yes, I know, the supposed "640k is enough for anyone" fallacy. Well, there really is a limit to what a normal human being needs to store. Why aren't MP3 files today 100 times larger than they were 15 years ago? Because the normal human's audio perception cannot tell the difference between a 5 MB MP3 and a 500 MB MP3. So the space required to store 1,000 songs is pretty much the same as 10 years ago, for most people.
In the last few years, we've reached the limits of human perception when it comes to image resolution. The display on my phone and my ultrabook are both so high resolution that I cannot see individual pixels without a magnifying glass. How high of a resolution does a photograph need to be to print it out 8x10 with pixels so small that they cannot be seen? We've already surpassed that resolution a long time ago.
Why don't computer monitors and image formats use 64 bit colors instead of 32 bit color that we've had for 15 years? Because the normal human cannot distinguish shades of color beyond 32 bit RGB.
When everything is in 4k video, why would we need higher resolution (unless people are regularly projecting things on screens as wide as their house)?
The amount of storage we need has already plateaued when it comes to certain kinds of media, and it will soon plateau in the others (video, etc) as well. At that point it's just a matter of quantity. What good would it do me to be able to store 1 million songs, or 1 million pictures on my phone? I certainly cannot produce that many myself, and I cannot even consume them either.
For normal consumers, there will be a limit to the amount of storage we need and thus will pay for. When that occurs, research will slow down as the profit to be gained from selling petabyte of storage vs an exabyte will no longer justify the research. We are quickly reaching the point where speed and longevity are more important than capacity, so I expect, within 5 years, the emphasis will switch from mainly quantity to quality.
Better known as 318230.
Intel and Micron produced a resistance change based 3D NAND chip they call 3D XPoint RAM. This ComputerWorld article explains the improvements in the memory chips leading to this breakthrough. As the description stated, it leans towards heralding the death of 2D NAND. 2D NAND chipsets have been reducing R&D funding for years. The reality is the cost for the latest chips will remain high until manufacturing improvements are made to reduce the 3D NAND technology. 2D NAND will continue to improve even with reduced R&D and costs will continue to decline. 3D NAND is the future, but it will be a several years before you see this technology in everyday devices. Good article, poor description.
Unified raw address space. You can still carve it up into virtual spaces if you want, like protected memory.
Probably at first. The first 128GB microSD cards by SanDisk were/are painfully slow. Lexar finally has a 128GB card now and it runs circles around the SanDisk. Took about a year for anyone else to produce a 128GB sized card after SanDisk released theirs. They have a 200GB version now, last time I checked it was over 4x's the price of the 128. I'd pay 4x's the price for a 400GB card, but not for just 72GB difference. I at least wand 256GB to come out, my current 128 is full :(.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
But seriously, UDF already fills the role perfectly.
Except for lack of support on devices other than desktop and laptop computers. If it is mandatory for SDXC certified devices to use exFAT, a lot of lazy device makers won't test anything else. Besides, some operating systems recognize UDF only on a drive that has a partition table, others only without a partition table. (Source)
You can map everything in the virtual memory space, what do you need more?
Until you realize that physical layout of memory is a much bigger issue than 'I have unified raw address space rrrraawwwwrrr.' Sure, you can have 2TB of memory. Some is faster than others, some is waaaay faster than others. Good luck.
The problem comes when a consumer wants to become a power user. Without a reasonable upgrade path, the sticker shock of replacement might deter consumers from becoming power users in the first place.
Oh, pish posh.
You don't need more than 640Kb.
I'm interested to know that, too. At 10MB/s, try to fill up a 400GB card. Worse if you're writing hundreds files (music) with a noticeable "pause" between every of them
I'd very much like my old gopros to be able to use more than 32GB cards. 128GB would be a good start but more is better.
I use the cameras in the race car that I drive in endurance events, a 12 hour race is common, I'm doing a 24 hour race at the end of the month. It would sure be nice to not have to swap cameras every couple of hours. At some point it gets screwed up and there's never video to figure out how it all went so poorly.
And no, wifi won't work. We need at least a mile of non-line-of-sight range, it's just not going to happen at 2.4GHz.
You could always spin up one of those System i machines still ;).
I use the cameras in the race car that I drive in endurance events, a 12 hour race is common, I'm doing a 24 hour race at the end of the month. It would sure be nice to not have to swap cameras every couple of hours. At some point it gets screwed up and there's never video to figure out how it all went so poorly.
You have lots of other options, though, since this is actually inside of a vehicle. You could use an itsy bitsy PC with a camera connected, for example — or, in fact, multiple cameras. If you have two cameras it could easily take up no more volume than using two gopros.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
If you don't like what Apple is selling, then buy something else.
Most computers never get upgraded. Apple thus made the reasonable tradeoff to sacrifice upgradability to make smaller, simpler, and more durable products. If you think this was a bad tradeoff, then you are free to buy something else. I'm perfectly happy with non-upgradable HW and I consider myself an informed shopper, not a gullible blind sheep. Different people care about different things.
Maybe it's time to stop caring so much how other people choose to spend their money.
"We need at least a mile of non-line-of-sight range, it's just not going to happen at 2.4GHz."
Nice job at failing to understand power transmission. It can happen, you simply lack the power (and regulative authority) to go to the power levels needed for such transmissions.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
If you think of the giant corpus of content that people access online all the time, with every click tracked and correlated by spy agencies and advertisers, almost all of it would be better off in local storage if it weren't for capacity and bandwidth issues. Wikipedia alone (counting Commons images) is around 20TB. A petabyte isn't that much storage these days: there are off-the-shelf data center appliances with that much capacity and companies use them all the time. The Internet Archive has tens of PB now. I'd be delighted to have it all on my portable computer instead of having to get it over the internet. Plus, lots of copies keep stuff safe.
Yeesh.
Stacking at 2x the depth produces 2x the heat per square unit of die size. We already have heat and power budget issues that give rise to "dark silicon". Therefore, there isn't very much vertical room to expand either. (And to think that the entire human brain runs at about 0.5 Watt...)
Thus the original remark about "information storage and retrieval [being] the thing holding computing back", perhaps?
Maybe a new transfer bus sometime soon too? I mean, SATA has been a limit for a long time now and PCIe simply isn't a viable choice for most. Even the most recent version of SATA (3.2) will be capping transfer speeds in a few years on single drive setups after some advancements in storage device technology.
Okay!!!11
If you could at least use a 500GB 2.5" single platter hard disk drive, that would be quite good. It's even a fair bit more than 12 x 32GiB.