"so the 256Gbit die alone isn't enough to bring the capacities above 10TB"
But we still don't know what the actual die size will be. Samsung's 3D-NAND is at 40nm, and the manufacturing process might be slightly different.,I believe (logically speaking) that a node shrink on a 3D-NAND die should see relatively much more gains in area than shrinking nodes for traditional NAND... It'll be interesting to see how things turn out (size/area/performance/endurance VS Samsung's) because 3D-NAND is MUCH more flexible in density, process and area, unless there's something I'm not getting here.
The die size isn't really an obstacle when it comes to increasing SSD capacities. The limits lie in die stacking and controller tech as only a limited number of die can be stacked (at least economically) and controllers have a fixed number of die that they can access. That's why the capacity per die is so important as doubled die capacity practically means doubled SSD capacity.
We likely won't see traditional die shrinks (i.e. transitions to smaller lithography) with 3D NAND, but just more layers stacked on top of each other.
The 34nm node is about the most cost effective process right now. Small enough to get a good transistor density but not so small that it requires more expensive techniques to produce. So we may very well see process shrink in the near future, combined with more die stacking.
Did you not read the article? 3x nm processes are cheaper to produce than smaller nodes in that they're less complex ie using only single patterning. Smaller nodes have historically been economical only because everything is 2D and it allows for more useable product per wafer. 3D technologies have entirely different economics because die size no longer directly correlates to density. Now you have die layers contributing to density. Not to mention larger node NAND has better performance and endurance meaning you save space by needing fewer channels/less spare area for equivalent performance to a smaller node.
You wouldn't realize it just looking at Intel and ARM processors, but 28 nm will probably the last die shrink for most products for many, many years due to lack of cost effectiveness. One projection I saw shows 28 nm ramping up until the middle of next decade. If you can stay above 30 nm, most products are going to try. 3D NAND is one way to do so.
As I just said, moving to smaller nodes involves technical challenges that exponentially increase costs. Moreover, when it comes to NAND, going to a smaller process cuts your P/E cycles so the endurance drops a lot. Going back to 32/34nm doubles the endurance of today's NAND so you have TLC with MLC endurance. Also the wear leveling algorithms can be simpler. Shrinking is great for 2D NAND because it's the only way to increase the chip's density even if it comes with caveats (like increased costs and reduces endurance). But for 3D NAND it's pointless because you can get the best of both worlds: endurance and cost effectiveness of 3xnm and the density you get from stacking more layers. No need to shrink.
+10TB SSDs in the 'next couple of years' is all right, but when we see true 2TB or 4TB SSD? I know about 2TB (P3700 & P3600), but they are very expensive and energy hungry. I would like something 4TB with price 4 x Samsung SSD 840 EVO 1TB or 2TB with price 2 x of that Samsung. Of course it would be SATA3 or SATA Express.
Kristian, could you please ask Intel, Samsung or other manufacturer about +2TB SSDs? I am waiting for announced 2TB ADATA SSD, but maybe I should wait for some 3D NAND SSD?
Manufacturers like Intel and Samsung are not really interested in 2TB client SSDs yet because they are such a small niche (yes, I've talked to them). We'll probably see some 2TB SSDs next year as Phison's S10 supports up to 2TB and so will the SF-3700 when it finally comes.
This is probably like wanting to eat my cake and still have it, but a 2TB SSD for under $500 is what I want more than any other upgrade. $0.25/GB... 2015?
The bigger limitation is if consumer SSDs will scale to 2TB by end of 2015. I think we'll be able to buy a 1TB drive for $250 by end of 2015 but not sure if there will be 2TB high performance drives like Samsung 860 Pro. 850 EVO hasn't even launched so it's doubtful that we'll see $500 for 2TB from any reasonably fast drive. The top of the line drives like SanDisk Extreme Pro and 850 Pro are just now hitting $185-275 for 480-512GB sizes.
SandDisk actually believes 10TB by 2016 is too slow:
"Intel is well behind the curve with its 10TB flash drive, slated for 2016, claims a SanDisk enterprise storage bigwig, adding that the flash memory storage provider will have a 16TB drive by then.
Brian Cox, SanDisk’s senior director in marketing for enterprise storage, said the company's 4TB drive technology would develop at a 2X/year capacity pace, with an 8TB drive next year, and 16TB in 2016.
If you want 2x or 4x the capacity of the current 1 TB drives, simply buy 2 or 4 of them. You don't even need so many 2.5" mounts in your case, as SSDs don't really need mounting (as long as they don't short your mainboard). Buy a PCIe controller card if you run out of SATA ports. It will be very cheap compared to those SSDs.
Yeah I'll believe the +10TB SSDs bit when they actually materialize. So far the entire industry has exhibited extraordinarily scant interest in increasing capacity or lowering prices. Similar to Intel's CPU strategy I've been left for years with very little reason or interest in upgrading.
I don't think that's true. SSDs have been coming down in price over the last couple of years. The Intel 520 SSD is half the price it was just a couple of years ago. A 1TB SSD was $1,000 18 months ago now they are under $500. Generally most drives are well under $1 per GB. I was an early adopter and paid nearly $470 for 100GB OCZ Vertex 2!!!
Not true, SSD prices for 240GB/256GB capacities were totally stuck for over a year at one point (from about Feb/13 onwards), infact they went *up* shortly after autumn 2012. For a brief time it looked like prices would go below 100 UKP equivalent for a 256GB model, but in early 2013 that all changed. Nothing moved until Crucial kicked everyone else up the butt with the MX100 and M500. I was tracking pricing on behalf of some industrial users, the static or rising prices were all too obvious. The most likely reason is proverbial supply & demand: why sell them cheaper when they're selling just fine at existing or even higher prices? One particular retailer in the UK did a special on the Samsung 830, selling hundreds of them in just a couple of weeks for around 125 UKP each. I'm sure this helped convinced Samsung & others that there was simply no need to lower the pricing any further, the demand was just fine at existing levels back then (135+ for a mainstream model like the 830). When the 840 came out, the 830 vanished and pricing went _up_ to around 145 for a while.
So, in other words, whatever an SSD *could* be sold for based on its production cost, if the demand is strong, the end consumer pricing will be higher. It needs a vendor willing to rock the boat to really make things change.
Today the 840 EVO 256GB is around 95 UKP in the UK, but when one considers the 830 256GB was 125 UKP in late 2012, overall that's not really a huge difference in 2 years.
It always happens that prices stutter. The same happens to RAM. Look at where that is now. It's much worse!
You need to look at an average trend line. When you do, you'll se the dropping prices over time, and the increased capacity. This is dependent on the memory manufacturers. When they come out with bigger chips, then we see bigger drives, and lowered pricing.
It also depends on our hers. If people aren't buying big drives because of cost, it will take longer.
That is weird because I bought a 240GB drive on Black Friday for $109 and note that was in Canadian dollars which I think were worth $0.88 American at the time.
That's not true. Capacity has increased, and prices have fallen. It used to be that 500GB drives were over $1,000, not all that long ago. Then they dropped to $600, and now they are below $300.
In turn, 1TB drives were well over $1,000 when the 500's were at $600, now you can get them for less than $500, if you look around.
Sheesh! I remember when 32GB models were over $3,000, a few years ago.
Regarding the use of an Intel fab for the SSD production - this may very well be done in Israel. Intel in Kiryat Gat, Israel, bought Micron, which has a plant right next to the Intel one. At first it was not clear what would happen to the plant and even to Intel's presence in Israel as a whole, but a few weeks ago, Intel signed a long-term deal with the Ministry of Finance, getting nice tax breaks while in return promising to stay for another 20 years or so. It seems like the perfect place for such a production facility.
Micron on 3D NAND flash at the UBS Tech conference on Nov 18;
Steven Chin
... But to make a transition maybe we could talk a little bit about NAND Flash opportunity you brought up the Singapore fab, you transition that from DRAM to NAND Flash. Sound like that will be the initial 3D NAND fab, can you just remind us what the timing is for 3D NAND roll out, is Singapore actually the primary location where you likely do 3D NAND, maybe just share some thoughts and let's around --
Kipp Bedard You bet. Yes. And Singapore will be initially the 3D NAND opportunity for us. Timing wise, we are shipping samples here in Q4. And we've always maintained that the ramp probably is going to take longer from an industry standpoint than people were expecting, these transitions are difficult. It is new technology.
Looking at the odd size of samsungs 3D flash... Is it possible that this is designed to actually be 128Gbit TLC, but then simulated down to MLC (to up the speed and endurance). It would certanly explain odd size.
Intel still loves to throw around that "10" number. Remember the 10GHz P4's we'd be using by 5 years ago? Yeah, like one poster wrote. How about affordable 4GB drives next year or 6GHz processors next year?
The thing I like about the V-NAND (or 3D-NAND, whatever the particular fabber calls it), the the sheer write endurance enabled by the larger process that these chips are fabbed on. The rush to smaller process just had too many tradeoffs with regards to speed and durability that I just didn't understand.
Plus, the increased density enabling larger SSD's is great. I really hope they start releasing some high capacity SSD's (2+ TB). Using those in my media server would give me such piece of mind that I'd probably not even enable RAID.
"The rush to smaller process just had too many tradeoffs with regards to speed and durability that I just didn't understand." The smaller processes make them cheaper, whereas endurance is stil fine for regular use and performance is limited by SATA3 anyway. From my point of view that's easy enough to understand. Don't make everyone pay more (demand more expensive SSDs manufactured with bigger structures) because you don't trust the endurance. If you really need more endurance, simply choose the more expensive drives available today.
And using SSDs for bulk media storage? Well.. that's pretty much throwing money out of the window. And especially there you don't need write endurance: you're putting the media collection there once and that's it. And not using RAID will not protect you from loosing an SSD controller, which is a far more typical failure than wearing out the NAND.
I'm still waiting for sata express drives. We have had the connector available for like at least a year i want to say and still no drive to connect to it. And Samsung is still the only vendor I will buy an ssd from. Samsung has complete vertical integration of the entire SSD product. They alone control the nand, the dram cache, the controller, the firmware. This is why their ssd's perform so well. They can tweak everything to synergize extremely well together. You can't beat that. I love my 256GB samsung 840 pro and I would like to get a 1TB samsung 860 pro sata express SSD when they finally start releasing some sata express drives (don't know what it will actually be called but you get the idea). I like to keep 25% of the SSD drive free to keep performance nice and maxed out and I'm running out of room.
Hmmm. After the EVO firmware debacle I'm steering clear of Samsung for a while, especially for new NAND tech - early adopter syndrome can be too costly even if it looks cheaper on paper.
It's nothing to do with being "cheap"! Not everyone wants, needs or can afford "Pro" editions. You make it sound like a snobbish choice, rather than people making informed decisions on how to prioritise their spending. The EVOs are drives designed as mass-market consumer items, that got rave reviews and recommendations from just about everyone, including here at AT, for their ground-breaking price/performance ratio (until the MX100 came out, anyway). This is the new battleground for companies as we all wait for cheaper and more dense NAND tech so that price/GB can drop further the way we all want.
The fact that they screwed the EVO up by not testing or validating their firmware properly in a rush to market with the cheap TLC design is concerning. Admitting they had a problem doesn't excuse the original failing. That's why I disagree strongly with the original poster that ONLY Samsung are trustworthy SSD vendors, something that's patently absurd in light of the EVO problems. They may not be unreliable on the scale of OCZ as was, but they still have to be called out on this, especially when they are likely to be the first out with a 3D NAND device that everyone will be expecting very big things from. It's even more difficult to risk being an early adopter if problems only start to show up months after first usage, as was the case with the EVO.
Just because you were lucky to be able to afford something MUCH more expensive, don't discount the bad experience of the many that bought EVOs expecting a base reliability which the vast majority of current SSD products share now (after the numerous issues with OCZ, SandForce etc. in the past which badly damaged this market) but who got burnt on the back of the EVO's hype in the press, and the assumption that Samsung couldn't screw up on something like this.
And the fixed the bug so everyone with an evo drive is having no problems. It was a slight inconvenience people had for a couple months and samsung admitted their problem and fixed it promptly. Everyone currently using an evo drive is running great now with no problem. If they didnt fix it I could see where u are coming from but they fixed it therefore the issue you are speaking of is non existent.
And how did they get burnt? The problem is fixed. It no longer exists so they are getting the performance they paid for. Do you not get the problem is fixed now. You act like since there was a problem everyone who bought the evo is still screwed over and all the money was wasted when in reality it's fixed and there is nothing to whine about now.
I'm not "acting" like anything. Read what I wrote instead of complaining about something I didn't write.
The original thread poster said he didn't trust anyone EXCEPT Samsung to produce reliable drives. I replied that this was a dubious position because Samsung HAD produced an unreliable drive with a serious problem. Whether they fixed it or not is irrelevant to the false premise that ONLY Samsung can be trusted to make reliable drives.
Debacle? How long have you been working on a technical level with technology? Samsung has had amazing tech support/response times compared to the rest of the industry. I still have a hard drive made by western digital 10 years ago that has a firmware bug (that causes data corruption) and has never received an update. Samsung SSDs are probably the fastest and most reliable out there. We use them for our desktops and our servers. I can forgive a company for an unforseen bug. I cannot forgive a company that refuses to fix said bugs.
Boys. Boys! Silence! :D Now I go back from 2025 and stick my super-duper faster USSB 16.0 with 20PB volume and speed of light read/write. This stick have trademark "DJ波波" who's makers also of doner of every corner. Intel turnip to eat and now must to begin much more R&D work to do, for not to be overtaken by someone worked in your garage Thai.
"it's safe to assume that the lithography is in the order of 30nm or 40nm because the whole idea of 3D NAND is to move away from multi-patterning to cut costs and with today's technology the smallest pitch of single-patterning is somewhere between 30nm and 40nm"
I've been frustrated wondering why Samsung didn't aim for 32 layers on a 2X node. They could completely disrupt the current SSD market and bring per GB costs to be on par to HDDs.
I've been waiting eagerly for a competitor to bottom out the market. The idea of a 1 TB flash drive as the minimum on a future iPad or really any mobile device is extremely exciting.
did u not read? it is cheaper to be on 30 to 40 due to single patterning of the lithography. It's cheaper to manufacture at that node. if 20nm was cheaper for them to make it they would charge the same price and pocket and even bigger profit margin. They aren't stupid every company wants to maximize thier profits.
The reason there was an aggressive push to smaller nodes is because without stacking the only way you can increase density is by going smaller cells. Now that they can stack they can use cheaper single patterning and still keep high densities.
Well smaller cells or increasing the overall die size but increasing the die size is an expensive poor option so they pushed smaller and smaller nodes on 2d nand to increase the max capacity. Even though manufacturing at the smaller nodes was more expensive it was still cheaper than pushing up the density any other way. Now they can cut manufacturing costs down using single pattern litho cells and still have nice 1TB size with 2TB already possible and right around the corner. I'm sure we will see consumer 2TB SSD's by the end of 2015.
Also since they are back to 40nm cell size write endurance is right back through the roof too. 40nm MLC is incredibly durable compared to the tiny 1xnm companies are using now. With 40nm even when samsung switches to TLC it will have more endurance than the 16nm and 19nm MLC nand. With 34nm or 40nm 32 layer stacking + 3 bit per cell and new controllers that can handle it, 4TB ssd's will become affordable. There is no point to go to more expensive double and triple patterning litho when we will already be able to get 4TB ssd's at the larger cell sizes with better endurance. Maybe for the 10TB intel is talking about.
Uh... Die size is also a major cost, as well as the number of layers. A smaller node lets you use smaller dies and significantly reduce the number of layers.
Right now I think there are so few companies with the 3D NAND tech that no one wants to rock the price boat yet.
im sure they did a cost comparison of what would cost them more. Using a larger die and more layers with single patterning litho or using smaller die and less layers with double or triple pattern litho. Defect rates go up with the smaller litho as more steps = more chances for defects. I'm sure they did the math and chose the best mix of characteristics to give them the lowest manufacturing costs. If 20nm was cheaper they would of done it I'm sure and charged the same price and pocketed an even higher percentage of net profit.
Do you really think they would just decide to use a process that cost them more money and make less profit on purpose? When 20nm v nand becomes the cheapest option then they will do it. If this EUV tech would get figured out than they could single pattern 20nm with it and leap frog 20nm into the cheap sweet spot.
Why do these companies scrap their old production that is builtup and efficient?
Intel and other companies can have two sections one from the new production process and use the older process for additional memory.
Let say Intel was going to introduce 512GB only of this 3D memory. They can have 512GB plus 1TB of the older production process. The benchmarks can test the new interface for speed.
That is weird because I bought a 240GB drive on Black Friday for $109 and note that was in Canadian dollars which I think were worth $0.88 American at the time.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
58 Comments
Back to Article
lilmoe - Tuesday, November 25, 2014 - link
"so the 256Gbit die alone isn't enough to bring the capacities above 10TB"But we still don't know what the actual die size will be. Samsung's 3D-NAND is at 40nm, and the manufacturing process might be slightly different.,I believe (logically speaking) that a node shrink on a 3D-NAND die should see relatively much more gains in area than shrinking nodes for traditional NAND... It'll be interesting to see how things turn out (size/area/performance/endurance VS Samsung's) because 3D-NAND is MUCH more flexible in density, process and area, unless there's something I'm not getting here.
Kristian Vättö - Tuesday, November 25, 2014 - link
The die size isn't really an obstacle when it comes to increasing SSD capacities. The limits lie in die stacking and controller tech as only a limited number of die can be stacked (at least economically) and controllers have a fixed number of die that they can access. That's why the capacity per die is so important as doubled die capacity practically means doubled SSD capacity.We likely won't see traditional die shrinks (i.e. transitions to smaller lithography) with 3D NAND, but just more layers stacked on top of each other.
close - Tuesday, November 25, 2014 - link
The 34nm node is about the most cost effective process right now. Small enough to get a good transistor density but not so small that it requires more expensive techniques to produce. So we may very well see process shrink in the near future, combined with more die stacking.III-V - Tuesday, November 25, 2014 - link
Source? Honestly, that sounds ridiculous. The entire point of moving to higher densities is for the improved economics.garadante - Tuesday, November 25, 2014 - link
Did you not read the article? 3x nm processes are cheaper to produce than smaller nodes in that they're less complex ie using only single patterning. Smaller nodes have historically been economical only because everything is 2D and it allows for more useable product per wafer. 3D technologies have entirely different economics because die size no longer directly correlates to density. Now you have die layers contributing to density. Not to mention larger node NAND has better performance and endurance meaning you save space by needing fewer channels/less spare area for equivalent performance to a smaller node.MrSpadge - Tuesday, November 25, 2014 - link
Ridiculous? That's exactly why the GPUs are still at 28 nm, despite 20 nm being available at TSMC.TelstarTOS - Friday, December 5, 2014 - link
It's not that simple with big die sizes such as GPU...saratoga3 - Tuesday, November 25, 2014 - link
You wouldn't realize it just looking at Intel and ARM processors, but 28 nm will probably the last die shrink for most products for many, many years due to lack of cost effectiveness. One projection I saw shows 28 nm ramping up until the middle of next decade. If you can stay above 30 nm, most products are going to try. 3D NAND is one way to do so.close - Wednesday, November 26, 2014 - link
As I just said, moving to smaller nodes involves technical challenges that exponentially increase costs. Moreover, when it comes to NAND, going to a smaller process cuts your P/E cycles so the endurance drops a lot. Going back to 32/34nm doubles the endurance of today's NAND so you have TLC with MLC endurance. Also the wear leveling algorithms can be simpler.Shrinking is great for 2D NAND because it's the only way to increase the chip's density even if it comes with caveats (like increased costs and reduces endurance). But for 3D NAND it's pointless because you can get the best of both worlds: endurance and cost effectiveness of 3xnm and the density you get from stacking more layers. No need to shrink.
Samus - Tuesday, November 25, 2014 - link
Well it's good to hear they're only a year or so behind Samsung.Greg100 - Tuesday, November 25, 2014 - link
+10TB SSDs in the 'next couple of years' is all right, but when we see true 2TB or 4TB SSD? I know about 2TB (P3700 & P3600), but they are very expensive and energy hungry. I would like something 4TB with price 4 x Samsung SSD 840 EVO 1TB or 2TB with price 2 x of that Samsung. Of course it would be SATA3 or SATA Express.Kristian, could you please ask Intel, Samsung or other manufacturer about +2TB SSDs? I am waiting for announced 2TB ADATA SSD, but maybe I should wait for some 3D NAND SSD?
Kristian Vättö - Tuesday, November 25, 2014 - link
Manufacturers like Intel and Samsung are not really interested in 2TB client SSDs yet because they are such a small niche (yes, I've talked to them). We'll probably see some 2TB SSDs next year as Phison's S10 supports up to 2TB and so will the SF-3700 when it finally comes.nathanddrews - Tuesday, November 25, 2014 - link
This is probably like wanting to eat my cake and still have it, but a 2TB SSD for under $500 is what I want more than any other upgrade. $0.25/GB... 2015?RussianSensation - Thursday, November 27, 2014 - link
It's possible by 1H 2016. Right now the 1TB Samsung 840 EVO is going for $360.http://www.amazon.com/Samsung-2-5-Inch-SATA-Intern...
The bigger limitation is if consumer SSDs will scale to 2TB by end of 2015. I think we'll be able to buy a 1TB drive for $250 by end of 2015 but not sure if there will be 2TB high performance drives like Samsung 860 Pro. 850 EVO hasn't even launched so it's doubtful that we'll see $500 for 2TB from any reasonably fast drive. The top of the line drives like SanDisk Extreme Pro and 850 Pro are just now hitting $185-275 for 480-512GB sizes.
SandDisk actually believes 10TB by 2016 is too slow:
"Intel is well behind the curve with its 10TB flash drive, slated for 2016, claims a SanDisk enterprise storage bigwig, adding that the flash memory storage provider will have a 16TB drive by then.
Brian Cox, SanDisk’s senior director in marketing for enterprise storage, said the company's 4TB drive technology would develop at a 2X/year capacity pace, with an 8TB drive next year, and 16TB in 2016.
What did he think of Intel's 10TB drive news? “I think it'll be late," he said, while talking to El Reg at a recent Gartner Data Centre Summit in London."
http://www.theregister.co.uk/2014/11/25/sandisk_in...
Michael Bay - Saturday, November 29, 2014 - link
Pffft, it`s just a suit talking shit and making promises which you never should believe.Let them put out at least 4Tb first.
rriiicchh - Sunday, November 30, 2014 - link
You mean like the 4Tb drive they released last May!http://www.sandisk.co.uk/enterprise/sas-ssd/optimu...
MrSpadge - Tuesday, November 25, 2014 - link
If you want 2x or 4x the capacity of the current 1 TB drives, simply buy 2 or 4 of them. You don't even need so many 2.5" mounts in your case, as SSDs don't really need mounting (as long as they don't short your mainboard). Buy a PCIe controller card if you run out of SATA ports. It will be very cheap compared to those SSDs.orionz - Tuesday, November 25, 2014 - link
Unless you are talking about a laptop, or a tablet, or a console, or servers stacked in racks with every bay already full.Magichands8 - Tuesday, November 25, 2014 - link
Yeah I'll believe the +10TB SSDs bit when they actually materialize. So far the entire industry has exhibited extraordinarily scant interest in increasing capacity or lowering prices. Similar to Intel's CPU strategy I've been left for years with very little reason or interest in upgrading.cactusdog - Tuesday, November 25, 2014 - link
I don't think that's true. SSDs have been coming down in price over the last couple of years. The Intel 520 SSD is half the price it was just a couple of years ago. A 1TB SSD was $1,000 18 months ago now they are under $500. Generally most drives are well under $1 per GB. I was an early adopter and paid nearly $470 for 100GB OCZ Vertex 2!!!mapesdhs - Tuesday, November 25, 2014 - link
Not true, SSD prices for 240GB/256GB capacities were totally stuck for over a year at one
point (from about Feb/13 onwards), infact they went *up* shortly after autumn 2012. For
a brief time it looked like prices would go below 100 UKP equivalent for a 256GB model,
but in early 2013 that all changed. Nothing moved until Crucial kicked everyone else up
the butt with the MX100 and M500. I was tracking pricing on behalf of some industrial
users, the static or rising prices were all too obvious. The most likely reason is proverbial
supply & demand: why sell them cheaper when they're selling just fine at existing or even
higher prices? One particular retailer in the UK did a special on the Samsung 830, selling
hundreds of them in just a couple of weeks for around 125 UKP each. I'm sure this
helped convinced Samsung & others that there was simply no need to lower the
pricing any further, the demand was just fine at existing levels back then (135+ for a
mainstream model like the 830). When the 840 came out, the 830 vanished and
pricing went _up_ to around 145 for a while.
So, in other words, whatever an SSD *could* be sold for based on its production cost,
if the demand is strong, the end consumer pricing will be higher. It needs a vendor
willing to rock the boat to really make things change.
Today the 840 EVO 256GB is around 95 UKP in the UK, but when one considers
the 830 256GB was 125 UKP in late 2012, overall that's not really a huge difference
in 2 years.
Ian.
melgross - Tuesday, November 25, 2014 - link
It always happens that prices stutter. The same happens to RAM. Look at where that is now. It's much worse!You need to look at an average trend line. When you do, you'll se the dropping prices over time, and the increased capacity. This is dependent on the memory manufacturers. When they come out with bigger chips, then we see bigger drives, and lowered pricing.
It also depends on our hers. If people aren't buying big drives because of cost, it will take longer.
earl colby pottinger - Saturday, December 6, 2014 - link
That is weird because I bought a 240GB drive on Black Friday for $109 and note that was in Canadian dollars which I think were worth $0.88 American at the time.It is in the computer I am using right now!
earl colby pottinger - Saturday, December 6, 2014 - link
Ouch, I just looked up the exchange rate, you guys are getting hosed.$109 CDN => 61.21 UKP! for a 240GB drive.
melgross - Tuesday, November 25, 2014 - link
That's not true. Capacity has increased, and prices have fallen. It used to be that 500GB drives were over $1,000, not all that long ago. Then they dropped to $600, and now they are below $300.In turn, 1TB drives were well over $1,000 when the 500's were at $600, now you can get them for less than $500, if you look around.
Sheesh! I remember when 32GB models were over $3,000, a few years ago.
odedia - Tuesday, November 25, 2014 - link
Regarding the use of an Intel fab for the SSD production - this may very well be done in Israel.Intel in Kiryat Gat, Israel, bought Micron, which has a plant right next to the Intel one. At first it was not clear what would happen to the plant and even to Intel's presence in Israel as a whole, but a few weeks ago, Intel signed a long-term deal with the Ministry of Finance, getting nice tax breaks while in return promising to stay for another 20 years or so. It seems like the perfect place for such a production facility.
Sources:
http://www.haaretz.com/business/1.549524
http://www.breakingisraelnews.com/22008/israel-int...
bhd2 - Tuesday, November 25, 2014 - link
Micron on 3D NAND flash at the UBS Tech conference on Nov 18;Steven Chin
... But to make a transition maybe we could talk a little bit about NAND Flash opportunity you brought up the Singapore fab, you transition that from DRAM to NAND Flash. Sound like that will be the initial 3D NAND fab, can you just remind us what the timing is for 3D NAND roll out, is Singapore actually the primary location where you likely do 3D NAND, maybe just share some thoughts and let's around --
Kipp Bedard
You bet. Yes. And Singapore will be initially the 3D NAND opportunity for us. Timing wise, we are shipping samples here in Q4. And we've always maintained that the ramp probably is going to take longer from an industry standpoint than people were expecting, these transitions are difficult. It is new technology.
http://seekingalpha.com/article/2693885-micron-tec...
I would guess that the first Intel samples are out of Singapore too!
hojnikb - Tuesday, November 25, 2014 - link
Looking at the odd size of samsungs 3D flash... Is it possible that this is designed to actually be 128Gbit TLC, but then simulated down to MLC (to up the speed and endurance). It would certanly explain odd size.bhd2 - Tuesday, November 25, 2014 - link
Now that Samsung has launched their 128 Gb TLC 3D NAND, I think that is exactly the case.Hulk - Tuesday, November 25, 2014 - link
Intel still loves to throw around that "10" number.Remember the 10GHz P4's we'd be using by 5 years ago?
Yeah, like one poster wrote. How about affordable 4GB drives next year or 6GHz processors next year?
The Von Matrices - Tuesday, November 25, 2014 - link
4GB drives are not affordable? It's extremely easy to find USB flash drives with double the capacity, 8GB, for under $5.mkozakewich - Tuesday, November 25, 2014 - link
"I'm not sure if I buy Intel's clam..."Oh baby, yes.
bill.rookard - Tuesday, November 25, 2014 - link
The thing I like about the V-NAND (or 3D-NAND, whatever the particular fabber calls it), the the sheer write endurance enabled by the larger process that these chips are fabbed on. The rush to smaller process just had too many tradeoffs with regards to speed and durability that I just didn't understand.Plus, the increased density enabling larger SSD's is great. I really hope they start releasing some high capacity SSD's (2+ TB). Using those in my media server would give me such piece of mind that I'd probably not even enable RAID.
MrSpadge - Tuesday, November 25, 2014 - link
"The rush to smaller process just had too many tradeoffs with regards to speed and durability that I just didn't understand."The smaller processes make them cheaper, whereas endurance is stil fine for regular use and performance is limited by SATA3 anyway. From my point of view that's easy enough to understand. Don't make everyone pay more (demand more expensive SSDs manufactured with bigger structures) because you don't trust the endurance. If you really need more endurance, simply choose the more expensive drives available today.
And using SSDs for bulk media storage? Well.. that's pretty much throwing money out of the window. And especially there you don't need write endurance: you're putting the media collection there once and that's it. And not using RAID will not protect you from loosing an SSD controller, which is a far more typical failure than wearing out the NAND.
jjj - Tuesday, November 25, 2014 - link
In February or so Micron should have it's winter analysts conference and we might get some details then.Laststop311 - Tuesday, November 25, 2014 - link
I'm still waiting for sata express drives. We have had the connector available for like at least a year i want to say and still no drive to connect to it. And Samsung is still the only vendor I will buy an ssd from. Samsung has complete vertical integration of the entire SSD product. They alone control the nand, the dram cache, the controller, the firmware. This is why their ssd's perform so well. They can tweak everything to synergize extremely well together. You can't beat that. I love my 256GB samsung 840 pro and I would like to get a 1TB samsung 860 pro sata express SSD when they finally start releasing some sata express drives (don't know what it will actually be called but you get the idea). I like to keep 25% of the SSD drive free to keep performance nice and maxed out and I'm running out of room.asmian - Tuesday, November 25, 2014 - link
Hmmm. After the EVO firmware debacle I'm steering clear of Samsung for a while, especially for new NAND tech - early adopter syndrome can be too costly even if it looks cheaper on paper.Laststop311 - Wednesday, November 26, 2014 - link
debacle? it wasn't even that bad they admitted problem and fixed it. And if you werent cheap and bought the pro editions it didn't affect youasmian - Wednesday, November 26, 2014 - link
It's nothing to do with being "cheap"! Not everyone wants, needs or can afford "Pro" editions. You make it sound like a snobbish choice, rather than people making informed decisions on how to prioritise their spending. The EVOs are drives designed as mass-market consumer items, that got rave reviews and recommendations from just about everyone, including here at AT, for their ground-breaking price/performance ratio (until the MX100 came out, anyway). This is the new battleground for companies as we all wait for cheaper and more dense NAND tech so that price/GB can drop further the way we all want.The fact that they screwed the EVO up by not testing or validating their firmware properly in a rush to market with the cheap TLC design is concerning. Admitting they had a problem doesn't excuse the original failing. That's why I disagree strongly with the original poster that ONLY Samsung are trustworthy SSD vendors, something that's patently absurd in light of the EVO problems. They may not be unreliable on the scale of OCZ as was, but they still have to be called out on this, especially when they are likely to be the first out with a 3D NAND device that everyone will be expecting very big things from. It's even more difficult to risk being an early adopter if problems only start to show up months after first usage, as was the case with the EVO.
Just because you were lucky to be able to afford something MUCH more expensive, don't discount the bad experience of the many that bought EVOs expecting a base reliability which the vast majority of current SSD products share now (after the numerous issues with OCZ, SandForce etc. in the past which badly damaged this market) but who got burnt on the back of the EVO's hype in the press, and the assumption that Samsung couldn't screw up on something like this.
Laststop311 - Wednesday, November 26, 2014 - link
And the fixed the bug so everyone with an evo drive is having no problems. It was a slight inconvenience people had for a couple months and samsung admitted their problem and fixed it promptly. Everyone currently using an evo drive is running great now with no problem. If they didnt fix it I could see where u are coming from but they fixed it therefore the issue you are speaking of is non existent.Laststop311 - Wednesday, November 26, 2014 - link
And how did they get burnt? The problem is fixed. It no longer exists so they are getting the performance they paid for. Do you not get the problem is fixed now. You act like since there was a problem everyone who bought the evo is still screwed over and all the money was wasted when in reality it's fixed and there is nothing to whine about now.Michael Bay - Saturday, November 29, 2014 - link
Do they pay you for all this verbal barf, at least?asmian - Wednesday, December 17, 2014 - link
If you don't want to read long comments, go troll Twitter instead.xkiller213 - Thursday, December 4, 2014 - link
the vanilla 840 suffers from that problem, and they didn't bother to fix it...http://i.imgur.com/NLAMGmp.png
asmian - Wednesday, December 17, 2014 - link
I'm not "acting" like anything. Read what I wrote instead of complaining about something I didn't write.The original thread poster said he didn't trust anyone EXCEPT Samsung to produce reliable drives. I replied that this was a dubious position because Samsung HAD produced an unreliable drive with a serious problem. Whether they fixed it or not is irrelevant to the false premise that ONLY Samsung can be trusted to make reliable drives.
Laststop311 - Wednesday, November 26, 2014 - link
and yes admitting the problem doesnt excuse the original failing but promptly fixing it does.eek2121 - Thursday, November 27, 2014 - link
Debacle? How long have you been working on a technical level with technology? Samsung has had amazing tech support/response times compared to the rest of the industry. I still have a hard drive made by western digital 10 years ago that has a firmware bug (that causes data corruption) and has never received an update. Samsung SSDs are probably the fastest and most reliable out there. We use them for our desktops and our servers. I can forgive a company for an unforseen bug. I cannot forgive a company that refuses to fix said bugs.Pork@III - Tuesday, November 25, 2014 - link
Boys. Boys! Silence! :D Now I go back from 2025 and stick my super-duper faster USSB 16.0 with 20PB volume and speed of light read/write. This stick have trademark "DJ波波" who's makers also of doner of every corner. Intel turnip to eat and now must to begin much more R&D work to do, for not to be overtaken by someone worked in your garage Thai.sonicmerlin - Tuesday, November 25, 2014 - link
"it's safe to assume that the lithography is in the order of 30nm or 40nm because the whole idea of 3D NAND is to move away from multi-patterning to cut costs and with today's technology the smallest pitch of single-patterning is somewhere between 30nm and 40nm"I've been frustrated wondering why Samsung didn't aim for 32 layers on a 2X node. They could completely disrupt the current SSD market and bring per GB costs to be on par to HDDs.
I've been waiting eagerly for a competitor to bottom out the market. The idea of a 1 TB flash drive as the minimum on a future iPad or really any mobile device is extremely exciting.
Laststop311 - Wednesday, November 26, 2014 - link
did u not read? it is cheaper to be on 30 to 40 due to single patterning of the lithography. It's cheaper to manufacture at that node. if 20nm was cheaper for them to make it they would charge the same price and pocket and even bigger profit margin. They aren't stupid every company wants to maximize thier profits.Laststop311 - Wednesday, November 26, 2014 - link
The reason there was an aggressive push to smaller nodes is because without stacking the only way you can increase density is by going smaller cells. Now that they can stack they can use cheaper single patterning and still keep high densities.Laststop311 - Wednesday, November 26, 2014 - link
Well smaller cells or increasing the overall die size but increasing the die size is an expensive poor option so they pushed smaller and smaller nodes on 2d nand to increase the max capacity. Even though manufacturing at the smaller nodes was more expensive it was still cheaper than pushing up the density any other way. Now they can cut manufacturing costs down using single pattern litho cells and still have nice 1TB size with 2TB already possible and right around the corner. I'm sure we will see consumer 2TB SSD's by the end of 2015.Also since they are back to 40nm cell size write endurance is right back through the roof too. 40nm MLC is incredibly durable compared to the tiny 1xnm companies are using now. With 40nm even when samsung switches to TLC it will have more endurance than the 16nm and 19nm MLC nand. With 34nm or 40nm 32 layer stacking + 3 bit per cell and new controllers that can handle it, 4TB ssd's will become affordable. There is no point to go to more expensive double and triple patterning litho when we will already be able to get 4TB ssd's at the larger cell sizes with better endurance. Maybe for the 10TB intel is talking about.
sonicmerlin - Wednesday, November 26, 2014 - link
Uh... Die size is also a major cost, as well as the number of layers. A smaller node lets you use smaller dies and significantly reduce the number of layers.Right now I think there are so few companies with the 3D NAND tech that no one wants to rock the price boat yet.
Laststop311 - Wednesday, November 26, 2014 - link
im sure they did a cost comparison of what would cost them more. Using a larger die and more layers with single patterning litho or using smaller die and less layers with double or triple pattern litho. Defect rates go up with the smaller litho as more steps = more chances for defects. I'm sure they did the math and chose the best mix of characteristics to give them the lowest manufacturing costs. If 20nm was cheaper they would of done it I'm sure and charged the same price and pocketed an even higher percentage of net profit.Do you really think they would just decide to use a process that cost them more money and make less profit on purpose? When 20nm v nand becomes the cheapest option then they will do it. If this EUV tech would get figured out than they could single pattern 20nm with it and leap frog 20nm into the cheap sweet spot.
DerekZ06 - Thursday, December 4, 2014 - link
>Do you really think they would just decide to use a process that cost them more money and make less profit on purpose?Companies Compete.
cointelpro123 - Wednesday, November 26, 2014 - link
Why do these companies scrap their old production that is builtup and efficient?Intel and other companies can have two sections one from the new production process and use the older process for additional memory.
Let say Intel was going to introduce 512GB only of this 3D memory. They can have 512GB plus 1TB of the older production process. The benchmarks can test the new interface for speed.
Rockfella.Killswitch - Thursday, November 27, 2014 - link
Will it be better than Samsung 850 Pro's 3D nand drives? I just bought the 128GB yesterday.earl colby pottinger - Saturday, December 6, 2014 - link
That is weird because I bought a 240GB drive on Black Friday for $109 and note that was in Canadian dollars which I think were worth $0.88 American at the time.It is in the computer I am using right now!