Download RTC June Industry Insider The Search Is on for the Successor to

RTC June Industry Insider
The Search Is on for the Successor to Flash Storage
The search for a memory solution to replace flash storage is on, with U.S. memory provider Data
Memory Systems tracking the updates and developments in the hope that they will be the first to
deliver the new flash storage solution to their customers. With semiconductor manufacturers
developing myriad solid-state technologies, the fight to see which solution will succeed
conventional flash memory is on.
NAND manufacturing process sizes are reaching their limits, which is why there has been a steep
downward trend in the prices of flash memory. Recently, however, famed supplier Samsung
announced that they would start producing NAND flash chips with a 3D internal architecture.
This sparked a new trend in flash memory, and other manufacturers set out to follow Samsung’s
lead. But when 3D NAND flash reaches its limits, which will happen in around three or four
years, what next?
The first potential candidate to take the place of flash memory is Resistive RAM—also known as
ReRAM or RRAM. This method of storage saves data by flipping resistors between two stable
states. It has a number of advantages over conventional NAND flash such as a longer life span,
lower power consumption and significantly higher performance levels. However, this method
may not hit the market until NAND flash has fully come to the end of the road, and it may not be
fast enough to compete with the other alternatives—SRAM and DRAM.
MRAM is also among the contenders, and it’s already being used as an SRAM alternative. It has
high performance levels and a far longer life, but it’s set to be more expensive than all other
memory options on the market so far. One type of MRAM has recently entered volume
production, and this may actually see prices come down eventually, making it a real contender.
Lastly, there’s phase-change memory, also known as PRAM or PCM. Although it’s not currently
in production, a number of big names in the memory industry, including Micron, IBM, Intel and
Samsung, have been monitoring and investigating the memory solution for a number of years. It
can be produced at small process sizes, making it great for regular consumers, and it heavily
outperforms NAND flash memory in every sense. It could provide everything from standard
consumer memory to “storage class” memory that, if produced cost-effectively, could eventually
replace SRAM, DRAM and NAND altogether.
The fight is on to see which method will win out. Many experts are making no predictions just
yet. There have been many rumors and false starts around the development of these solutions,
and none of them are willing to put their neck on the line and make the call on the future of flash
memory. Data Memory Systems will continue to monitor the situation with the goal of being the
first to provide the most up-to-date computer memory solutions when they hit the market.
Adlink and VadaTech Partner to Offer Enhanced AdvancedTCA Solutions
VadaTech and Adlink Technology have announced a partnership to offer AdvancedTCA-based
embedded computing solutions. VadaTech will utilize Adlink’s line of processor boards to
complement VadaTech’s switch, chassis, RTM and specialty products in the AdvancedTCA
architecture. The boards will include Adlink’s line of 10G and 40G processor and packet
processing blades based on the Intel Xeon E-series chipset. Adlink intends to capitalize on
VadaTech’s integration capability and diverse ecosystem of chassis and switches to complement
its core processor offering. Together the companies provide an unmatched array of
AdvancedTCA solutions, from boards up to integrated computing systems.
“We see a great synergy between our embedded platforms and the integrated solutions provided
by VadaTech,” said Daniel Yang, GM of Adlink North America. “This affiliation helps both of
our companies expand our reach in the AdvancedTCA market.”
“VadaTech is excited to work with Adlink and provide their versatile array of processor boards
based on the AdvancedTCA architecture,” said Saeed Karamooz, CEO of VadaTech. “This
partnership will expand the VadaTech offering and better serve the integrated solution needs of
our customers.”
AMD and Mentor Graphics Join Yocto Project as New Advisory Board Members
The Yocto Project, a Linux Foundation Collaboration Project, has announced that AMD and
Mentor Graphics are increasing their investments in the embedded Linux project. Both
companies are becoming Gold-level members and will sit on the Yocto Project Advisory Board.
The Yocto Project reduces fragmentation in the embedded market by providing developers with
greater consistency in the software and tools they’re using across multiple architectures for
embedded Linux development. It is a collaborative, open source project that provides templates,
tools and methods to help developers create custom, embedded, Linux-based systems, regardless
of hardware architecture.
AMD and Mentor Graphics will maximize their investments in the Yocto Project and deepen
their commitment to the embedded Linux development community with this latest move. The
two companies recently announced a multi-year partnership to deliver open source embedded
Linux software tools for AMD Embedded processors that comprise multicore and heterogeneous
systems targeting data storage and networking, industrial automation, Internet of Things (IoT)
and visual applications. The Yocto Project technical leadership and governance is merit-based;
maintainers and technical leaders are selected based on the quality and quantity of their code
contributions to the project.
Other Advisory Board members of the Yocto Project include Gold members Freescale, Intel,
Juniper Networks, LSI, OpenEmbedded, Sakoman, Inc., Texas Instruments and Wind River
Systems. Silver members of the Advisory Board include Dell, Enea AB, Huawei, LG,
MontaVista Software, OS Systems and Renesas.
Cadence Gobbles Up Jasper
2012 was the year that everyone remembers Synopsys going on an acquisition binge, but 2014
will go down as the year that Cadence Design Systems decided that EDA was worth investing in.
Rather than placing investment bets outside of its core competence, Cadence bought Forte in
February and now adds Jasper Design Automation to its fold.
Jasper has been the premier player in the formal verification market, making all other players
look like second-class citizens. Many had said that so much money had been invested in Jasper
over the years that a buyout was unlikely, but Kranen and the investors always had faith that they
would get the returns they expected. Today that belief turned into a reality when Cadence offered
approximately $170 million in cash. Jasper had approximately $24 million of cash, cash
equivalents and short-term investments as of December 31, 2013. That makes Jasper one of the
highest valued companies in the EDA industry. Over the years, Jasper has received
approximately $27M in funding, with the latest round being $2.13M in 2012.
In addition, the latest new app is JasperGold Sequential Equivalence Checking App. This new
app enables designers to exhaustively verify the sequential functional equivalence of RTL
implementations, ensuring that they function identically at sequential design points—and 10x
faster than competing tools. This ties in nicely with Cadence’s Forte acquisition, because one of
the technologies necessary to make high-level synthesis successful is sequential equivalence
checking. Calypto is the only other company that has these two pieces in the same company, and
because Calypto is majority owned by Mentor Graphics, this would potentially be a hole in the
Cadence portfolio.
Earthquake Simulation Tops One Quadrillion FLOPS
A team of computer scientists, mathematicians and geophysicists at Technische Universitaet
Muenchen (TUM) and Ludwig-Maximillians Universitaet Muenchen (LMU) have—with the
support of the Leibniz Supercomputing Center of the Bavarian Academy of Sciences and
Humanities (LRZ)—optimized the SeisSol earthquake simulation software on the SuperMUC
high-performance computer at the LRZ to push its performance beyond the “magical” one
petaflop/s mark one quadrillion floating point operations per second.
Geophysicists use the SeisSol earthquake simulation software to investigate rupture processes
and seismic waves beneath the Earth’s surface. Their goal is to simulate earthquakes as
accurately as possible to be better prepared for future events and to better understand the
fundamental underlying mechanisms. However, the calculations involved in this kind of
simulation are so complex that they push even super computers to their limits.
In a collaborative effort, the workgroups led by Dr. Christian Pelties at the Department of Geo
and Environmental Sciences at LMU and Professor Michael Bader at the Department of
Informatics at TUM, have optimized the SeisSol program for the parallel architecture of the
Garching supercomputer “SuperMUC,” thereby speeding up calculations by a factor of five.
Using a virtual experiment they achieved a new record on the SuperMUC: To simulate the
vibrations inside the geometrically complex Merapi volcano on the island of Java, the
supercomputer executed 1.09 quadrillion floating point operations per second. SeisSol
maintained this unusually high performance level throughout the entire three hour simulation run
using all of SuperMUC’s 147,456 processor cores.
Stephen Hawking Warns of Danger of Artificial Intelligence
In a commentary in the British daily, The Independent, famed Physicist Stephen Hawking has
issued a warning about the potential danger involved in perfecting artificial intelligence. Citing
such efforts as self-driving cars, the personal assistant Siri and Google Now, Hawking says that
these early successes will pale against what the ambitious pursuit of AI is likely to achieve in
coming decades. Mostly, we tend to think of the potential benefits of the pervasive spread of AI,
and Hawking notes that there seem to be no theoretical limits to what could be achieved with
intelligent machines continually working on improving their own design. This could eventually
take them out of the realm of science fiction as portrayed in the latest Hollywood blockbuster,
Transcendence, starring Johnny Depp, since, as Hawking states, “there are no fundamental limits
to what can be achieved.”
The “dark side” of this vision is potentially that with the world’s militaries looking at
autonomous weapons systems that choose and attack their own targets, the possible actions of
such systems could get out of control. Could they, for instance, outsmart financial markets,
develop weapons that humans cannot even understand, or outstrategize the leaders trying to
combat them? “The short-term impact,” says Hawking, “depends on who controls it; the longterm impact depends on whether it can be controlled at all.”
We are currently and quite optimistically building out the Internet of Things. This interconnected
universe is bound to be host to an increasing amount of artificial intelligence distributed
throughout the world. The idea that this could harbor potential danger is something that could
easily be disregarded in our efforts to advance the technology. The fact that a person with the
stature of Stephen Hawking is raising a flag of caution should at least be worthy of some serious
attention.
Google’s ATAP Group Selects Lattice FPGAs for its Project Ara Modular Smartphone Prototype
Lattice Semiconductor Corp. has announced that Google’s Advanced Technology and Projects
group has selected Lattice FPGAs for its ambitious Project Ara initiative that aims to deliver the
world’s first modular smartphones for consumers to configure from a variety of modules. Made
available to developers last week, and the subject of the recent Project Ara Modular Developers
Conference, the Module Developers Kit (MDK) incorporates Lattice FPGAs for critical
connectivity between reference implementations of removable modules and the Project Ara
endoskeleton.
In addition to enabling companies to rapidly develop prototypes of Project Ara modules, the low
power and small size of Lattice FPGAs meet the system requirements of a thermally constrained
environment, as well as provide the flexibility to support the MIPI UniPro network protocol that
will be used for connectivity between modules. Finally, Lattice FPGAs are a proven solution for
mobile consumer products, making them ideal for production modules as well. Developers can
go from prototype to production, reducing the product development effort and accelerating the
time-to-market. The advantages of Lattice FPGAs have already been proven in millions of
smartphones currently used by consumers worldwide.