Tuesday, September 10, 2013

How Apple Leverages its R&D

11 Apple products which were realised via strategic acquisitions
Apple leverages its R&D by buying 11 small companies over the last few years as is discussed in the article below.

For example,
"In 2010, Apple unveiled the first iPad. Inside of it was a powerful custom SoC designed by Apple called the A4. Later came the A5, A5X, A6, A6X, ....none of these outstanding SoCs would even be possible if Apple didn’t make a handful of strategic acquisitions over the last few years. In 2008, Apple picked up P.A. Semi ..., 2010 saw the acquisition of Intrinsity for ..., and Passif Semiconductor was purchased earlier this year"

See Apple iPad 4 – A6X teardown


Flash NAND is remaking computers, servers, mobile tablets, and cell phones architecture and Apple realizes it.

 "Anobit Technologies was founded in Israel in 2006. It developed dozens of patents for flash controllers,...Apple uses so many NAND chips in iPads, iPhones, iPods, and Macs, ...Apple acquired ... in December 2011 ... Any competitive edge Apple can find for creating cheaper and better flash-based devices holds the potential to drastically change its production chain. Flash storage is constantly improving, and it’s in Apple’s best interest to stay on the cutting edge of speed and reliability improvements."

HDR photography, face detection, and iAds  will be integrated in future Apple products (see below).

 Ron

Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -
http://www.maltiel-consulting.com/



11 Apple products which were realised via strategic acquisitions

 by Grant Brunner, 09 Sep 2013Features
Apple is widely thought of as a major innovator in the tech industry. While it’s true that Cupertino does employ some of the best designers and engineers in the world, Apple simply wouldn’t be where it is without a large number of strategic acquisitions. Most notably, iOS and OS X wouldn’t even exist without the purchase of NeXT.
Of course, that acquisition also brought Steve Jobs back to Apple, but that serves to highlight the strength of a smart acquisition. Not only will purchasing a company net you products and patents, but you’re gaining an influx of amazing employees as well.
In this article, I’d like to walk through a number of important acquisitions Apple has made in the last decade. From the A6 SoC through multitouch displays to Siri, Apple has relied heavily on acquired technology to build the iPhone, iPad, and iOS. It’s important to remember that purchases like these help make Apple what it is, and even the best engineers in the world need some fresh ideas from time to time.

Multitouch

FingerWorks, founded by doctoral students at the University of Delaware, focused on developing numerous devices like the TouchStream multitouch keyboard. Its quirky products never took off on a large scale, but the core technology and patents were successful enough to garner the attention of Apple. In 2005, Apple purchased the company, and effectively shut down the existing business. Two years later, the iPhone came out, and heavily featured multitouch technology.

Even if Apple was working on multitouch devices before the acquisition, it’s clear that the expertise and patent portfolio that came along with FingerWorks helped make multitouch displays become an industry standard.


App Store Genius

Chomp was founded in 2009, and served as a search engine for iOS and Android apps. The search tool was available on the web and as an app, and it enabled users to discover new and relevant apps more easily than Apple or Google’s built-in search engines. It received millions of dollars of venture capital funding, and even partnered with Verizon to build a backend for the telecom’s app store. Despite its success as an independent company, Apple was able to snatch up Chomp in early 2012 for a reported $50 million (£32 million). The Android and iOS apps disappeared, and the Chomp technology appears to be powering app searches and the App Store Genius functionality currently featured on the App Store.

Custom SoCs

In 2010, Apple unveiled the first iPad. Inside of it was a powerful custom SoC designed by Apple called the A4. Later came the A5, A5X, A6, A6X, and the rumoured upcoming A7 chip. However, none of these outstanding SoCs would even be possible if Apple didn’t make a handful of strategic acquisitions over the last few years. In 2008, Apple picked up P.A. Semi for $280 million (£180 million), 2010 saw the acquisition of Intrinsity for $121 million (£77 million), and Passif Semiconductor was purchased earlier this year for an unknown figure. With this ever increasing pool of talent and patents, Apple continues to make a name for itself in the world of SoC design.


Apple Maps

Apple’s maps have been a bit of a sore spot for the last year. Apple’s contract with Google ran out, and Apple decided to build its own solution. Of course, a few purchases helped get the ball rolling. Placebase was purchased back in July of 2009, Poly9 was added to Apple’s portfolio in July of 2010, and 3D mapping company C3 Technologies was acquired in August of 2011. This amalgam of purchases, alongside partnerships with companies like TomTom, led to the launch of Apple’s own mapping service in 2012. It was received very poorly, but Apple hasn’t stopped trying. WiFiSlam was purchased in March of this year, Locationary and HopStop.com were picked up in July, and Embark was snatched up by Cupertino just a couple of weeks ago. With this much mapping prowess under its roof, Apple is clearly taking this problem very seriously.

Flash storage

Anobit Technologies was founded in Israel in 2006. It developed dozens of patents for flash controllers, and quickly gained substantial notoriety and venture funding. Since Apple uses so many NAND chips in iPads, iPhones, iPods, and Macs, it was no surprise when Apple acquired the Israeli company and its patent portfolio in December 2011 for a whopping $390 million (£250 million). Any competitive edge Apple can find for creating cheaper and better flash-based devices holds the potential to drastically change its production chain. Flash storage is constantly improving, and it’s in Apple’s best interest to stay on the cutting edge of speed and reliability improvements.


HDR photography

Apple introduced built-in HDR (High Dynamic Range) photography on the iPhone 4 with the release of iOS 4.1. By taking multiple photos at different exposures, and then merging the visual data together, photographers can end up with better looking pictures with more detail in dark and bright areas. To help improve the core technology, Apple purchased a UK-based company called IMSense for an undisclosed amount of money in September 2010. As the hardware and software of the iPhone have improved, so have the overall quality of HDR images. Now, amazing HDR techniques are available to even novice photographers using iPhones.


Face detection

Faces might just be the most important part of personal photos. Finding and identifying faces is vital for good photography software, and Apple does a decent job on most fronts. iPhoto automatically finds faces, and attempts to guess who is in each photo based on previous input. On iOS, the camera app automatically detects faces, and enables developers to easily implement live photo manipulation. Part of this clever face detection is thanks to the 2010 acquisition of a Swedish company called Polar Rose. With that core technology, any developer on iOS can take advantage of automatic face detection thanks to the built-in Core Image API.


iAds

After Google began to make serious moves against Apple in the smartphone market, Cupertino decided to get into the advertising game – seemingly in retaliation towards its former partner. After a bit of a bidding war, Google was able to secure the purchase of AdMob in 2009. Not to be outdone, Apple purchased Quattro Wireless in 2010 to compete directly against Mountain View. iAds was born out of this purchase, although it never really took off on a large scale. Even so, this acquisition remains a notable point in the fascinating breakdown of the working relationship between Google and Apple.


Read more: http://www.itproportal.com/2013/09/09/11-apple-products-which-were-realised-via-strategic-acquisitions/#ixzz2eUrfkP4e

Thursday, September 5, 2013

Hynix Fab Fire Manufacturing Implications

Yesterday Hynix had a fire at their  at its Wuxi,China, plant

It looks that the fire would have only short term impact on NAND production since it "mainly in the air purification facilities on the rooftop of the fab and created a disproportionate amount of smoke".

If its true that it only impacted the air filtration system, production is dependent on how quickly Hynix can fix/ replace the air purification components, and the magnitude of the damage to the air filtration inside the fab.

"TrendForce said the fire could impact SK Hynix's production procedures "considerably" in the near future, particularly since the Wuxi plant manufactures almost 50 percent of the company's monthly output of 260,000 wafers and contributes over 10 percent of the world's DRAM wafer production. It added that the facility is responsible for producing 100,000 of its PC DRAM and 30,000 mobile DRAM. 
"The potential damages imparted on the supply end should not be underestimated," the research firm said. Should SK Hynix's main production line be stalled, the shipment of almost 11 million laptops and 10 million smartphone units will be affected within the span of a month, Trendforce said. "....
http://www.zdnet.com/sk-hynix-china-plant-fire-wont-affect-chip-production-7000020281/

Short term, Hynix production will depend on how quickly Hynix can get a new system installed and running.

Longer term impact depends on;

1. Market demand for year end for consumer NAND, SSD (tablet /iPad growth vs . PC stagnation, and mobile phones)

2. The real short term pace of demand for server farms, cloud SSD products.

Ron


 Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -
http://www.maltiel-consulting.com/

Monday, August 26, 2013

Longest Battery Life Smartphone: Moto X (Teardown)

Below are some snapshot from Ifixit teardown of Google/ Motorola new cell phone. One of the key features of Motorola cell phones have been their very long battery life. Is it due to the combination of Qualcomm processor and the power management chip, some of the other chips, or their custom system architecture (software/ hardware)?

The key chips in Moto X are listed below.

Ron
Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -
http://www.maltiel-consulting.com/



Motorola Moto X Teardown

 
Image #1

Step 1 — Motorola Moto X Teardown 

  • The love child of Motorola and Google is here, and we are dying to crack open the little Motoroogle.
  • Technical Specifications:
    • Dual-core 1.7 GHz Qualcomm Snapdragon S4 Pro processor
    • 2 GB RAM
    • 4.7-inch 1280x720 pixels AMOLED display
    • 16 or 32 GB internal storage
    • Qualcomm Adreno 320 GPU
    • .........
    • Image #1

      Step 9 

      • Motorola claims the Moto X battery can power through an amazing 24 hours of "mixed usage."
      • How does the Moto X accomplish such a feat with a 3.8 volt, 2200 mAh Lithium ion battery? The secret is in the X8 Mobile Computing System.
      • The Motorola X8 Mobile Computing System is comprised of a Qualcomm Snapdragon S4Pro family processor, a natural language processor and a contextual computing processor.
      • Motorola developed a custom system architecture, which, when coupled with eight processor cores, allows for the delegation of processing power:
        • 4 graphics processor cores for "stunning clarity"
        • 2 application processor cores for "swift action"
        • 2 low-power cores—"awaiting your next command"
      Image #2

      Step 10 

      • Out comes the upper midframe panel, housing the speaker, headphone jack, more antennas, and pressure contacts.
        • Yay, pressure contacts! We like spring pressure contacts because they don't require any work to disconnect.
      • This is possibly the most modular headphone jack we've ever seen. It pops right out of the upper midframe panel, spring contacts and all.

      10 MP rear-facing camera

    • Image #1

      Step 15 

      • Notable ICs on the motherboard:
        • Toshiba THGBMAG7A2JBAIR 16 GB eMMC NAND Flash
        • SK Hynix H9TKNNNBPDAR RAM (we assume that the Snapdragon S4 Pro is also layered under this IC)
        • Qualcomm PM8921 Power Management IC
        • Texas Instruments TMS320C55 Digital Signal Processor
        • NXP 44701 NFC Chip
        • Skyworks 77619-12 Multiband Multimode Power Amplifier Module for Quad-Band GSM / EDGE and Penta-Band (Bands I, II, IV, V, VIII) WCDMA/ HSDPA/ HSUPA/ HSPA+/ LTE
        • Texas Instruments MSP430 F5259 Mixed Signal Microcontroller
      Image #1

      Step 16 

      • Additional ICs:
        • Qualcomm WCD9310 Audio Codec
        • Qualcomm WCN3680 802.11ac Combo Wi-Fi/Bluetooth/FM
        • NXP TFA9890 High Efficiency Class-D Audio Amplifier
        • Skyworks 77737 SkyHi™ Power Amplifier Module for LTE Bands 12/17 (698-716 MHz)
        • EPCOS 7 959 Wireless LAN / Bluetooth Filters (IF)
        • 0V00660 A56G 1B

Wednesday, August 14, 2013

Samsung 3D Stacked NAND Flash has Engineering Samples

Yesterday Samsung announced 3D flash V-NAND at Flash Summit


A key advantage of 3D vertical scaling is that device and process development issues of silicon based technology are better understood than other future flash approaches that depends on integrating brand new materials. 

Two other strong flash vendors are also developing 3D Flash

Toshiba and Hynix

There is an interesting discussion regarding 3D flash from 2009 between Samsung and Toshiba
3D Cells Make Terabit NAND Flash Possible

Based on Samsung keynote announcement at the flash summit:

 1. Samsung is having already engineering sample now and it will be in production in 2014.

2. Unless there is some unexpected development, it sound like Samsung's 3D NAND (and similar flavors by its competitors) will be the mainstream future NAND technology.

3. A key issue which they did not explain is erase cycle. They only said that erase had to be optimized with specific circuits. 

4. Samsung has been developing it since 2003. Initially they just developed the CTF memory cell technology as a standard planar NAND (see my previous comments  and at http://maltiel-consulting.com/Samsung's_32-gigabit-Gbit_40-nm_CTF-NAND_uses_high-k.htm). However, that product was not a commercial sold.

6. The first 3D test product they made was 16G in 2011, the current one is 128Gb which is built using a stack of 24 layers.

7. Samsung is building it based on a standard known 30nm silicon technology, they said it is cost competitive (or cheaper?) than planar technology. I am not sure what are their assumptions are as far as cost, but they are probably correct that it will be cheaper than competing future technologies.






Ron
Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at - http://www.maltiel-consulting.com/

Wednesday, August 7, 2013

Samsung’s 1Tb SSD: 3D Vertical NAND

The article below discuss using CTF instead of floating gate to create 3D flash memory with up to 1T SSD product next year.

" new V-NAND is manufactured at a 10nm process size, and it starts at a density of 128Gb per NAND chip. The NAND chips are constructed in layers, stacking up to 24 individual NAND cells on top of each other...

Samsung is claiming that at minimum, the CTF-based V-NAND has at least a 2x increase in lifespan over floating gate NAND, and perhaps as high as 10x. Additionally, write performance is doubled over floating gate NAND."

Already in 2006 Samsung discussed CTF memory cell for NAND chips CTF for 40nm 32Gb .

More on CTF in March 2012 (Micron/ Intel 20-nm 64G MLC NAND Flash Memory Reverse Engineered).

At MemCon on August 6, 2013 Samsung said that more details will be presented at Flash Summit next week.

Ron
Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at - http://www.maltiel-consulting.com/


Samsung’s “3D Vertical” NAND crams a terabit on a single chip

Longer life, higher reliability, more performance—what's not to like?

SSD enthusiasts know all about SLC, MLC, and TLC, but there are some new acronyms in SSD town: V-NAND and CTF. Samsung announced in a press release last night that it has begun mass production of "3D Vertical NAND," a type of flash that it claims overcomes the existing limits on the design and production of existing NAND types. When we looked at those limits about a year ago, they seemed pretty significant; Samsung's V-NAND aims to neatly sidestep most of the issues.
Enlarge / Samsung's 3D Vertical NAND stacks up to 24 NAND elements on top of each other.
 
 
 
The new V-NAND is manufactured at a 10nm process size, and it starts at a density of 128Gb per NAND chip. The NAND chips are constructed in layers, stacking up to 24 individual NAND cells on top of each other. This lets S amsung scale the chip's capacity up without having to add more NAND cells in a series, or "planar scaling," as the traditional "just shrink 'em and add more cells" method is called.
The other acronym, CTF, stands for "Charge Trap Flash." Traditional NAND flash records zeros and ones by storing charge in a set of floating gate transistors, with the presence or absence of charge corresponding to a 0 or a 1 in single-level cell NAND, and the amount of charge corresponding to different multibit values in multi- and triple-level cell NAND (we have an extremely in-depth primer on the inner workings of SSDs if you want more details). However, Samsung's new V-NAND dispenses with floating gate transistors and uses a different method:
Samsung's CTF-based NAND flash architecture, an electric charge is temporarily placed in a holding chamber of the non-conductive layer of flash that is composed of silicon nitride (SiN), instead of using a floating gate to prevent interference between neighboring cells.
The longevity and reliability problems with standard floating gate transistor-based NAND have a lot to do with the large amounts of power required to perform erasures. Without taking too large a digression, each time a NAND transistor undergoes a program/erase cycle, it retains some additional electrons in its dielectric layer. Eventually, these trapped electrons alter the transistor's resistance to the point that it can no longer be reliably read. The problem grows worse as the NAND cell manufacturing process shrinks—smaller cells become useless at lower levels of retained charge.
Smaller NAND transistor gates means it takes a smaller retained charge to overwhelm the gates' ability to quickly and reliably change state.
Aurich Lawson

The switch from floating gate to Charge Trap Flash appears to negate a lot of these issues. Samsung is claiming that at minimum, the CTF-based V-NAND has at least a 2x increase in lifespan over floating gate NAND, and perhaps as high as 10x. Additionally, write performance is doubled over floating gate NAND.
Samsung predicts that V-NAND will scale up to 1Tb per individual NAND chip. Most SSDs use at least eight NAND chips in parallel, so V-NAND could lead directly to low dollar-per-GB 2.5-inch form factor SSDs of 1TB and beyond—capacities which many Ars commenters have said repeatedly that they desperately want. At that size, concerns over installing an operating system and a few games on a fast SSD and stashing non-speed-critical files on a larger HDD are moot, and most folks can simply use the SSD exclusively without worry. There's no word yet on exactly when a consumer-level SSD filled with V-NAND will become available, but Samsung's vertical integration likely means that the first V-NAND SSD will be a Samsung-branded product with a Samsung-branded SSD controller. 

Tuesday, July 30, 2013

Server Storage 85% Faster w/o PCI Express

Latest advance in flash storage by changing system architecture and how flash storage is integrated are discussed in the article below. The potential for changes in flash storage architecture was already discussed in May 2012

 "Diablo’s Memory Channel Storage (MCS) architecture, expected to show up in servers shipping later this year, allows flash storage components to plug into the super-fast channel now used to connect CPUs with memory. That will slash data-access delays even more than current flash caching products that use the PCI Express bus...

Diablo estimates that MCS can reduce latencies by more than 85 percent compared with PCI Express SSDs (solid-state disks)...

The connection is designed to be used by many DIMMs (dual in-line memory modules) in parallel, so each component doesn’t have to relinquish the bus for another one to use it. That saves time, as well as CPU cycles that would otherwise be used managing the bus"


Ron
Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at - http://www.maltiel-consulting.com/



Flash breakthrough promises faster storage, terabytes of memory

In the ongoing quest for faster access to data, Diablo Technologies has taken what could be a significant next step.
Diablo’s Memory Channel Storage (MCS) architecture, expected to show up in servers shipping later this year, allows flash storage components to plug into the super-fast channel now used to connect CPUs with memory. That will slash data-access delays even more than current flash caching products that use the PCI Express bus, according to Kevin Wagner, Diablo’s vice president of marketing.
The speed gains could be dramatic, according to Diablo, helping to give applications such as databases, big data analytics and virtual desktops much faster access to the data they need most. Diablo estimates that MCS can reduce latencies by more than 85 percent compared with PCI Express SSDs (solid-state disks). Alternatively, the flash components could be used as memory, making it affordable to equip servers terabytes of memory, Wagner said.
Other than on-chip cache, the memory channel is the fastest route to a CPU, Wagner said. Not only do bits fly faster over this link, there are also no bottlenecks under heavy use. The connection is designed to be used by many DIMMs (dual in-line memory modules) in parallel, so each component doesn’t have to relinquish the bus for another one to use it. That saves time, as well as CPU cycles that would otherwise be used managing the bus, Wagner said.
The parallel design of the memory bus also lets system makers scale up the amount of flash in a server without worrying about diminishing returns, he said. A second MCS flash card will truly double performance, where an added PCIe SSD could not, Wagner said.
Diablo, which has been selling memory controllers for about 10 years, has figured out a way to use the standard DDR-3 interface and protocols to connect flash instead of RAM to a server’s CPU. Flash is far less expensive than RAM, but also more compact. The MCS components, which come in 200GB and 400GB sizes, will fit into standard DIMM slots that typically accommodate just 32GB or so of memory. The only adaptation manufacturers will need to make is adding a few lines of code to the BIOS, Wagner said.
Enterprises are more likely to use MCS as high-capacity memory than as low-latency storage, said analyst Jim Handy of Objective Analysis.
“Having more RAM is something that a lot of people are going to get very excited about,” Handy said. His user surveys show most IT departments automatically get as much RAM as they can for their servers, because memory is where they can get the fastest access to data, Handy said.
“Basically, you’d like everything to be in the RAM,” Handy said. Virtualized data centers, where many servers need to share a large set of data, need a shared store of data. But in other applications, especially with databases and online transaction processing, storage is just a cheaper and more plentiful—but slower—alternative to memory. “Everything that’s on the storage is there just because it can’t fit on the RAM,” he said.
To implement the MCS architecture, Diablo developed software and a custom ASIC (application-specific integrated circuit), which it will sell to component vendors and makers of servers and storage platforms. Flash vendor Smart Storage Systems, which earlier this month agreed to be acquired by SanDisk, will be among the companies using the MCS technology, Wagner said. In addition, a tier-one server vendor is preparing about a dozen server models with the technology and will probably ship the first of them this year, Walker said.
For the most part, Diablo doesn’t expect consumers or small enterprises to install MCS flash on their own computers. However, Diablo may work directly with enterprises that have very large data centers they want to accelerate, he said.
Using MCS flash to supplement DRAM would dramatically reduce the per-gigabyte cost of memory but also would allow for further consolidation of the servers in a data center, Wagner said. A large social networking company with 25,000 servers analyzed the MCS technology and said it would make it possible to do the same amount of work with just 5,000 servers.
That’s because the current DRAM-only servers can be equipped with just 144GB of memory, but MCS would allow each server to have 16GB of DRAM and 800GB of flash. With that much memory, each server can do more work so fewer are needed, Wagner said. Fewer servers would mean savings of space and energy, which would translate into lower costs, he said.