It has become increasingly apparent that flash memory die shrinks have hit a bit of a brick wall in recent years. The issues faced by the standard 2D Planar NAND process were apparent very early on. This was no real secret - here's a slide seen at the 2009 Flash Memory Summit:
Despite this, most flash manufacturers pushed the envelope as far as they could within the limits of 2D process technology, balancing shrinks with reliability and performance. One of the largest flash manufacturers was Intel, having joined forces with Micron in a joint venture dubbed IMFT (Intel Micron Flash Technologies). Intel remained in lock-step with Micron all the way up to 20nm, but chose to hold back at the 16nm step, presumably in order to shift full focus towards alternative flash technologies. This was essentially confirmed late last week, with Intel's announcement of a shift to 3D NAND production.
Intel's press briefing seemed to focus more on cost efficiency than performance, and after reviewing the very few specs they released about this new flash, I believe we can do some theorizing as to the potential performance of this new flash memory. From the above illustration, you can see that Intel has chosen to go with the same sort of 3D technology used by Samsung - a 32 layer vertical stack of flash cells. This requires the use of an older / larger process technology, as it is too difficult to etch these holes at a 2x nm size. What keeps the die size reasonable is the fact that you get a 32x increase in bit density. Going off of a rough approximation from the above photo, imagine that 50nm die (8 Gbit), but with 32 vertical NAND layers. That would yield a 256 Gbit (32 GB) die within roughly the same footprint.
Representation of Samsung's 3D VNAND in 128Gbit and 86 Gbit variants.
20nm planar (2D) = yellow square, 16nm planar (2D) = blue square.
Image republished with permission from Schiltron Corporation.
It's likely a safe bet that IMFT flash will be going for a cost/GB far cheaper than the competing Samsung VNAND, and going with a relatively large 256 Gbit (vs. VNAND's 86 Gbit) per-die capacity is a smart move there, but let's not forget that there is a catch - write speed. Most NAND is very fast on reads, but limited on writes. Shifting from 2D to 3D NAND netted Samsung a 2x speed boost per die, and another effective 1.5x speed boost due to their choice to reduce per-die capacity from 128 Gbit to 86 Gbit. This effective speed boost came from the fact that a given VNAND SSD has 50% more dies to reach the same capacity as an SSD using 128 Gbit dies.
Now let's examine how Intel's choice of a 256 Gbit die impacts performance:
- Intel SSD 730 240GB = 16x128 Gbit 20nm dies
- 270 MB/sec writes and ~17 MB/sec/die
- Crucial MX100 128GB = 8x128Gbit 16nm dies
- 150 MB/sec writes and ~19 MB/sec/die
- Samsung 850 Pro 128GB = 12x86Gbit VNAND dies
- 470MB/sec writes and ~40 MB/sec/die
If we do some extrapolation based on the assumption that IMFT's move to 3D will net the same ~2x write speed improvement seen by Samsung, combined with their die capacity choice of 256Gbit, we get this:
- Future IMFT 128GB SSD = 4x256Gbit 3D dies
- 40 MB/sec/die x 4 dies = 160MB/sec
Even rounding up to 40 MB/sec/die, we can see that also doubling the die capacity effectively negates the performance improvement. While the IMFT flash equipped SSD will very likely be a lower cost product, it will (theoretically) see the same write speed limits seen in today's SSDs equipped with IMFT planar NAND. Now let's go one layer deeper on theoretical products and assume that Intel took the 18-channel NVMe controller from their P3700 Series and adopted it to a consumer PCIe SSD using this new 3D NAND. The larger die size limits the minimum capacity you can attain and still fully utilize their 18 channel controller, so with one die per channel, you end up with this product:
- Theoretical 18 channel IMFT PCIE 3D NAND SSD = 18x256Gbit 3D dies
- 40 MB/sec/die x 18 dies = 720 MB/sec
- 18x32GB (die capacity) = 576GB total capacity
Overprovisioning decisions aside, the above would be the lowest capacity product that could fully utilize the Intel PCIe controller. While the write performance is on the low side by PCIe SSD standards, the cost of such a product could easily be in the $0.50/GB range, or even less.
In summary, while we don't have any solid performance data, it appears that Intel's new 3D NAND is not likely to lead to a performance breakthrough in SSD speeds, but their choice on a more cost-effective per-die capacity for their new 3D NAND is likely to give them significant margins and the wiggle room to offer SSDs at a far lower cost/GB than we've seen in recent years. This may be the step that was needed to push SSD costs into a range that can truly compete with HDD technology.
Subject: General Tech | November 24, 2014 - 12:40 PM | Jeremy Hellstrom
Tagged: ssd, Intel, 3d nand
Great news flash fans, Intel will be increasing the storage density of SSDs quite significantly over the next few years. They will be using the 3D NAND technology we have just been introduced to to stack flash memory with 32 planar layers for 32GB per cell with MLC and 48GB per cell if TLC flash is used. This increased density could lead to 10TB SSDs by 2017 as well as mobile devices with 1TB of local memory that runs at higher speeds than the current generations as well. As The Register noted this will have to be accompanied by price reductions as at $1.00/GB no one would even dream of a 10TB drive and even at $0.50 it would be far too expensive. Perhaps Ryan's dreams of low cost flash storage are not as far out there as some seem to feel, indeed he may not be aiming low enough for price per GB. You can also get a peek at what Samsung, Hynix and Sandisk will be up to in the same article.
"IMFT, Intel Micron Flash Technologies, a partnership between Intel and Micron, has a 3D MLC NAND technology, which will be used to build 10TB SSDs in two years."
Here is some more Tech News from around the web:
- Linux distributor SUSE delves into software-defined storage @ The Inquirer
- Intel decides to keep tablet subsidies, say sources @ DigiTimes
- Mozilla remembers 2013. Y'know, back when it still gobbled at the Google money-trough @ The Register
- Digitimes Research: Samsung, Apple, LG rank as top-3 smartphone vendors in 3Q14 @ DigiTimes
- KitGuru visits Logitech’s G Labs in Switzerland
- First in line to order a Nexus 6? AT&T has a BRICK for you @ The Register
- Tech ARP 2014 Mega Giveaway Contest
Subject: Processors | November 20, 2014 - 01:31 PM | Josh Walrath
Tagged: amd, APU, carrizo, Carrizo-L, Kaveri, Excavator, Steamroller, SoC, Intel, mobile
AMD has certainly gone about doing things in a slightly different manner than we are used to. Today they announced their two latest APUs which will begin shipping in the first half of 2015. These APUs are running at AMD and are being validated as we speak. AMD did not release many details on these products, but what we do know is pretty interesting.
Carrizo is based on the latest iteration of AMD’s CPU technology. Excavator is the codename for these latest CPU cores, and they promise to be smaller and more efficient than the previous Steamroller core which powers the latest Kaveri based APUs. Carrizo-L is the lower power variant which will be based on the Puma+ core. The current Beema APU is based on the Puma architecture.
Roadmaps show that the Carrizo APUs will be 28 nm products, presumably fabricated by GLOBALFOUNDRIES. Many were hoping that AMD would make the jump to 20 nm with this generation of products, but that does not seem to be the case. This is not surprising due to the limitations of that particular process when dealing with large designs that require a lot of current. AMD will likely be pushing for 16 nm FinFET for the generation of products after Carrizo.
The big Carrizo supposedly has a next generation GCN unit. My guess here is that it will use the same design as we saw with the R9 285. That particular product is a next generation unit that has improved efficiency. AMD did not release how many GCN cores will be present in Carizzo, but it will be very similar to what we see now with Kaveri. Carrizo-L will use the same GCN units as the previous generation Beema based products.
I believe AMD has spent a lot more time hand tuning Excavator instead of relying on a lot of automated place and route. This should allow them to retain much of the performance of the part, all the while cutting down on transistor count dramatically. Some rumors that I have seen point to each Excavator module being 40% smaller than Steamroller. I am not entirely sure they have achieved that type of improvement, but more hand layout does typically mean greater efficiency and less waste. The downside to hand layout is that it is extremely time and manpower intensive. Intel can afford this type of design while AMD has to rely more on automated place and route.
Carrizo will be the first HSA 1.0 compliant SOC. It is in fact an SOC as it integrates the southbridge functions that previously had been handled by external chips like the A88X that supports the current Kaveri desktop APUs. Carrizo and Carrizo-L will also share the same infrastructure. This means that motherboards that these APUs will be soldered onto are interchangeable. One motherboard from the partner OEMs will be able to address multiple markets that will see products range from 4 watts TDP up to 35 watts.
Finally, both APUs feature the security processor that allows them access to the ARM TrustZone technology. This is a very small ARM processor that handles the secure boot partition and handles the security requests. This puts AMD on par with Intel and their secure computing solution (vPro).
These products will be aimed only at the mobile market. So far AMD has not announced Carrizo for the desktop market, but when they do I would imagine that they will hit a max TDP of around 65 watts. AMD claims that Carrizo is one of the biggest jumps for them in terms of power efficiency. A lot of different pieces of technology have all come together with this product to make them more competitive with Intel and their process advantage. Time will tell if this is the case, but for now AMD is staying relevant and pushing their product releases so that they are more consistently ontime.
Subject: General Tech, Processors, Mobile | November 19, 2014 - 07:36 PM | Scott Michaud
Tagged: x86, restructure, mobile, Intel
Last month, Josh wrote about Intel's Q3 earnings report. The company brought in $14.55 billion USD, of which they could keep $3.31 billion. Their PC group is responsible for $9 billion of that revenue and $4.12 billion of that profit, according to the Wall Street Journal. On the other hand, their mobile division is responsible for about $1 million – and it took over a billion to get that million. This has been the trend for quite some time now, as Intel pushes their square battering ram into the mobile and tablet round hole. Of course, these efforts could benefit the company as a whole, but they cannot show that in a quarterly, per-division report.
And so we hear rumors that Intel intends to combine their mobile and PC divisions, which Chuck Mulloy, an Intel spokesperson, later confirmed in the same article. The new division, allegedly called the “Client Computing” group in an internal email that was leaked to the Wall Street Journal, will handle the processors for mobile devices but, apparently, not the wireless modem chipsets; those will allegedly be moved to a “wireless platform research and development organization”.
At face value, this move should allow Intel to push for mobile even more aggressively, while simultaneously reducing the pressure from investors to give up and settle for x86 PCs. Despite some differences, this echos a recent reorganization by AMD, where they paired-up divisions that were doing well with divisions that were struggling to make a few average divisions that were each treading water, at least on paper.
The reorganization is expected to complete by the end of Q1 2015, but that might not be a firm deadline.
Subject: General Tech | November 18, 2014 - 12:53 PM | Jeremy Hellstrom
Tagged: Intel, tablet, smartphone
Intel's smartphone and tablet divisions are being rolled up under Kirk Skaugen's PC Business division in what was described as an attempt to "accelerate Intel's opportunity in the marketplace". This is definitely needed as Intel is having great difficulties garnering market share in the mobile industry, the only successes they have had are with larger more expensive tablets. One possible benefit of this merger that was mentioned in the Inquirer's article was Intel's plans to leverage the growing use of LTE in both tablets and laptops, perhaps we shall see a 4G card become far more common in even basic models. Intel implied that they are not planning any layoffs at this point although unless their ultramobile division can pick up the pace it seems inevitable that some will feel the axe. We shall see over the coming year if Intel's focus on low powered silicon can help in their battles against incumbents such as ARM and Qualcomm.
"INTEL HAS ANNOUNCED plans to merge its smartphone and tablet operations with its PC division in a bid to make the two businesses more efficient, and to further compete with rivals in the mobile semiconductor market."
Here is some more Tech News from around the web:
- A first taste of Lollipop on Nvidia's Shield Tablet @ The Tech Report
- The optic NERVE of it: Intel declares WAR on InfiniBand @ The Register
- Lenovo unveils a water-cooled dual-socket Intel Xeon E5 server @ The Inquirer
- Asustek said to be adopting Qualcomm solution for new ZenFone; looking for price cut from Intel @ DigiTimes
Subject: Storage | November 14, 2014 - 02:00 PM | Jeremy Hellstrom
Tagged: btrfs, EXT4, XFS, F2FS, 530 series, Intel, raid, unix humour, linux
When you use Linux you have a choice as to which file system you wish to use, a choice that never occurs to most Windows users but can spark an argument every bit as vicious as the eternal debate over EMACS versus VIM versus whichever text editor you prefer. There has not been much SSD benchmarking done on alternate files systems until now, Phoronix has benchmarked the Intel 530 series SSD in numerous configurations on Btrfs, EXT4, XFS, and F2FS. With four of the 120GB model available they were able to test the speed of the drives in RAID 0, 1, 5, 6, and 1+0. There is obviously still some compatibility issues as some tests failed to run in certain configurations but overall these drives performed as expected. While the results did not vary widely it is worth reading through their article if you plan on building a high speed storage machine which will run Linux.
"Following the recent Btrfs RAID: Native vs. Mdadm comparison, the dual-HDD Btrfs RAID benchmarks, and four-SSD RAID 0/1/5/6/10 Btrfs benchmarks are RAID Linux benchmarks on these four Intel SATA 3.0 solid state drives using other file-systems -- including EXT4, XFS, and Btrfs with Linux 3.18."
Here are some more Storage reviews from around the web:
- Samsung XS1715 (1.6TB) @ The SSD Review
- Angelbird SSD2go Pocket USB 3.0 External Solid State Drive @ eTeknix
- Silicon Power Thunder T11 120 GB @ techPowerUp
- Silicon Power Armor A80 2TB USB 3.0 Portable Hard Drive Review @ NikKTech
- Patriot Memory Stellar Boost XT 64GB USB 3.0 OTG Drive Review @ Madshrimps
- Synology DiskStation DS1815+ @ Legion Hardware
- Western Digital Red Pro (WD4001FFSX) 4 TB @ Tech ARP
Subject: General Tech | November 13, 2014 - 03:19 PM | Ken Addison
Tagged: podcast, video, Intel, core m, core m 5y70, Broadwell, broadwell-y, Lenovo, yoga 2 pro, yoga 3 pro, assasins creed unity, ubisoft, farcry 4, p3500, gskill blade
PC Perspective Podcast #326 - 11/13/2014
Join us this week as we discuss Intel's Core M 5Y70, Assassin's Creed Unity, Intel P3500 and more!
The URL for the podcast is: http://pcper.com/podcast - Share with your friends!
- iTunes - Subscribe to the podcast directly through the Store
- RSS - Subscribe through your regular RSS reader
- MP3 - Direct download link to the MP3 file
Hosts: Ryan Shrout, Jeremy Hellstrom, Josh Walrath, and Allyn Malventano
Program length: 1:09:49
Subject: Storage | November 12, 2014 - 04:44 PM | Allyn Malventano
Tagged: ssd, pcie, NVMe, Intel, DC P3500
Since we reviewed the Intel SSD DC P3700, many of you have been drooling over the idea of an 18-channel NVMe PCIe SSD, even more so given that the P3500 variant was to launch at a $1.50/GB target price. It appears we are getting closer to that release, as the P3500 has been appearing on some web sites in pre-order or out of stock status.
ShopBLT lists the 400GB part at $629 ($1.57/GB), while Antares Pro has an out of stock listing at $611 ($1.53/GB). The other two capacities are available at a similar cost/GB. We were hoping to see an 800GB variant, but it appears Intel has stuck to their initial plan. Here are the part numbers we’ve gathered, for your Googling pleasure:
- 400GB: SSDPEDMX400G401
- 1.2TB: SSDPEDMX012T401
- 2TB: SSDPEDMX020T401
2.5” SFF-8639 (*not SATA*):
- 400GB: SSDPE2MX400G401
- 1.2TB: SSDPE2MX012T401
- 2TB: SSDPE2MX020T401
We did spot a date of December 12th in an Amazon listing, but I wouldn't count that as a solid date, as many of the listings there had errors (like 10 packs for the price of one).
Subject: Storage | November 11, 2014 - 05:32 PM | Allyn Malventano
Tagged: Intel, ssd, dc s3500, M.2
Today Intel refreshed their Datacenter Series of SSDs, specifically their DC S3500. We have reviewed this model in the past. It uses the same controller that is present in the S3700, as well as the SSD 730 Series (though it is overclocked in that series).
The full line of Intel Datacenter SSDs (minus the P3700). DC S3500 is just right of center.
Todays refresh includes higher capacities to the S3500, which now include 1.2TB and 1.6TB on the hign end of capacity. This suggests that Intel is stacking 20nm dies as many as 8 to a package. IOPS performance sees a slight penalty at these new higher capacities, while maximum sequentials are a bit higher due to the increased die count.
Also announced was an M.2 version of the S3500. This packaging is limited to only a few capacity points (80GB, 120GB, 340GB), and is p;rimarily meant for applications where data integrity is critical (i.e. ATM's, server boot partitions, etc).
A standard press blast was unavailable, but full specs are listed after the break.
Core M 5Y70 Specifications
Back in August of this year, Intel invited me out to Portland, Oregon to talk about the future of processors and process technology. Broadwell is the first microarchitecture to ship on Intel's newest 14nm process technology and the performance and power implications of it are as impressive as they are complex. We finally have the first retail product based on Broadwell-Y in our hands and I am eager to see how this combination of technology is going to be implemented.
If you have not read through my article that dives into the intricacies of the 14nm process and the architectural changes coming with Broadwell, then I would highly recommend that you do so before diving any further into this review. Our Intel Core M Processor: Broadwell Architecture and 14nm Process Reveal story clearly explains the "how" and "why" for many of the decisions that determined the direction the Core M 5Y70 heads in.
As I stated at the time:
"The information provided by Intel about Broadwell-Y today shows me the company is clearly innovating and iterating on its plans set in place years ago with the focus on power efficiency. Broadwell and the 14nm process technology will likely be another substantial leap between Intel and AMD in the x86 tablet space and should make an impact on other tablet markets (like Android) as long as pricing can remain competitive. That 14nm process gives Intel an advantage that no one else in the industry can claim and unless Intel begins fabricating processors for the competition (not completely out of the question), that will remain a house advantage."
With a background on Intel's goals with Broadwell-Y, let's look at the first true implementation.