Showing posts with label internet of things. Show all posts
2017
promises to be an exciting year for servers and the competitiveness of
compute offerings. This year will see this scope of impact not only
include enterprise datacenters and the public cloud, but extend to the
emergence of "edge computing". Edge computing is defined as compute
required to deal with data at or near the point of creation. Among other
things, these devices will include the “ocean” of remote, smart
sensors, commonly included in internet of things (IoT) discussions.
Here
is a list of a few things to we’ll see concerning specific CPUs.
It
should come as no surprise that Intel INTC -0.16% continues to dominate
(>99%) the server market but is under enormous pressure on all fronts. Xeon
and its evolution continue to be their compute vanguard. Xeon-Phi (and now the
addition of Nervana) make up their engines for high-performance computing /
machine learning. Phi has seen some success, but it isn’t clear yet how Nervana
offerings will materialize.
Advanced
Micro Devices AMD -0.57% (AMD) has their best shot in years for fielding an
Intel competitor that just about everyone (except perhaps Intel) is eager to
see. If the AMD Zen server CPU is simply good enough (meaning, it shows up,
works and has at least some performance value), it will take market share
simply by being an x86 competitor. AMD is encouraged by early indicators. They
also have their ATI GPGPU technology which will provide additional
opportunities.
ARM
Holdings will continue to dominate the mobile and embedded device space, but
the fight is hard in these segments. The more likely opportunity for ARM
expansion will be at the "edge" and not so much in the server space.
The death of Vulcan by Avago, the acquisition of Applied Micro Circuits (APM)
and their plan to find a place for X-Gene leaves the Cavium CAVM +0.69%
ThunderX, the "yet to be launched" Qualcomm QCOM -0.02% Centriq CPU
and a few other very focused ARM initiatives still standing. After years of
"This is the Year for ARM Servers", the outlook could be better, and
if AMD produces a plausible Intel competitor (capable of running x86 software),
it will put extreme pressure on whole ARM server CPU initiative.
OpenPOWER
seems on the other hand to have a lot of momentum but to date has not
significantly impacted the x86 server market. 2017 may end on a different note.
OpenPOWER‘s (IBM IBM -1.27%) willingness to embrace NVIDIA NVDA -0.74% (the
darling of the machine learning segment) and embed an NV-Link interface is
going to play well with much of AI and HPC communities. By the end of the year,
we will have seen some interesting OpenPOWER offerings emerge based on advanced
silicon process technology from a variety of sources, and 2018 may see a whole
different story. Especially if an embrace from Google GOOGL -0.14%, who has
been flirting with OpenPOWER for a while now, materializes and creates a
tipping point.
The
real challenge to all CPUs is the way they do work. Their philosophy is built
on the principle that data must come into the chip, be operated on by the chip,
with results or even new data being pushed out of the chip. This whole process
creates a natural bottleneck that we've flirted with for decades. As the
magnitude and scope of data increases, something has got to give, and a
favorite candidate is more parallelism. So far, this has favored GPGPUs or
accelerators.
At
the bigger-picture business level for datacenters and the public cloud, the
real question is not so much which CPU (in fact, the business folks probably
couldn't care less), but the economics of private, public or hybrid solutions.
It is safe to say enterprise computing will not disappear any time soon, and
while there is much activity, the implementations and economics of hybrid
solutions have proven to be difficult. According to Gartner, by 2020 more
compute power will have been sold by IaaS and PaaS cloud providers than sold
and deployed into enterprise datacenters. The fact that companies (especially
smaller ones) are either being born in or moving to the cloud at a rapid pace
is undeniable. However, NOT all are seeing the expected saving materialize from
this move. 2017 will certainly see some careful thinking and maybe even some
rethinking of strategy.
The
explosion of data at the edge is simply going to change data processing as we
know it and will create a variety of computing problems that are difficult to
do in the cloud (even though the results may end up there). However, they may
not be in the enterprise datacenters as we know them either, and we may find
them “stuck” all over the place. For more than sixty years, we have seen
compute follow the data. First from the original mainframe datacenter to the
desktop, to departmental servers, into enterprise datacenters, and now
significantly into the cloud. It is my opinion, that If you plan to put just
your data into the cloud, economics (the cost of network usage) will drive your
compute there sooner or later. You want to consider this carefully based on
your actual needs and usage. There might be a better overall business outcome,
depending on your size and ability to operate, in your own datacenter.
The
major emerging source of data is at the edge and will drive the need for much
compute there. By the way, all the CPUs mentions here should be able do the
edge reasonably well so … Game on again!
Disclosure:
My firm, Moor Insights & Strategy, like all research and analyst firms,
provides or has provided research, analysis, advising, and/or consulting to
many high-tech companies in the industry including Advanced Micro Devices,
Applied Micro Circuits, ARM Holdings, IBM, Intel, NVIDIA and Qualcomm. I do not
hold any equity positions with any companies cited in this column.
A sprinkling of Internet of Things ... and ... it's alive: Startup
Igneous Systems has re-discovered and re-imagined the idea of customers renting
an externally managed system on their premises, giving it an Internet of Things
(IoT) and public cloud make-over.
Customers
start with a 2x1 system of two dataRouters and one dataBox at a usable capacity
of 212TB. The dataRouters and dataBoxes can scale independently.
The
new angles are that IoT devices can generate vast amounts of data which is
difficult to send to a set of on-premises servers or the public cloud because
transmission is too slow and/or the data is sensitive and needs to be retained
on site.
Igneous's
answer is to store it locally in a public cloud-like, scale-out,
hyper-converged system that can run some compute processes on the data, and be
managed through the cloud. Example compute workloads are auto-tagging, metadata
extraction, image processing, and text analysis.
Igneous
owns and operates the on-premises equipment and the customer has a
capacity-based subscription pricing scheme. The systems are remotely managed
and monitored through a cloud service, and there is automated cloud management
processes for upgrades, troubleshooting, and failure management. Igneous claims
that these functions can be performed on a mirror of incoming and outgoing data
streams without slowing down low-level system functions.
The
Igneous architecture utilises a microservices approach and incorporates stream
processing, an event-driven framework, and container services.
The
equipment is based on ARM-powered nano-servers, with each being a disk drive,
ARM processor and Ethernet link. Igneous calls this a JBOND - Just a Bunch of
Networked Drives. There is a SoC (System On Chip) with a Linux-running Marvell
32-bit Armada 370 processor, with two Cortex-A9 cores running at up to 1 GHz.
Networking is via two 1 GbitE ports, and the nanoservers operate in a kind of
mesh.
It
reminds El Reg of Seagate's Ethernet-addressed Kinetic disk drives, except that
Igneous has had its own nanoservers built and the JBOND is accessed as an S3
data store. Any processing capabilities to operate on the stored data need
adding. Data is stored using erasure coding and failed nanoserver disks have
data rebuilt using other nanoservers.
We
also think 32-bit ARM CPUs look a bit light in compute power terms and 64-bit
ones will need to be used if any serious compute is going to be done on the
data.
Access
is via Amazon's S3 API across a customer's 10Gbit Ethernet infrastructure.
Other public cloud service access protocols will probably be added in the
future.
A
dataBox has 60 drives (nano-servers with 6TB disks) in a 4U enclosure and is a
capacity store. A dataRouter is a stateless 1U server used for protocol
endpoints and keeping track of data in the nanoserver mesh.
There
is some existing third-party software support. For example, Commvault Virtual
Server Protection (VSP) enables users to back up to an Igneous System and
Infinite IO works with Igneous.
This
Igneous nanoserver kit is an interesting start in what we might ultimately call
micro or nano hyper-converged systems. Getting involved with it means
leading/bleeding edge work for Linux and S3 buffs.
Get
a datasheet here. The Igneous Data Service is available immediately to
customers in North America. Customers can purchase annual subscriptions in
increments of 212TB of usable capacity, starting at under $40,000 (equivalent
to 1.5 cents/GB/month). This is claimed to be lower than the cost of S3 (3
cents/GB/month). Volume discounts are available, based on installed capacity or
contract duration. ®
Will a fight break out over who’s responsible for securing data? Maybe. Will companies start taking security seriously? Not sure. Will design engineers need to address security before corporate management?
A
key component of the Internet of Things (IoT) and the Industrial Internet of
Things (IIoT) is the cloud; that group of services residing nearly anywhere
that will house all the data collected. Despite all the buzz about the IoT and
its variations, most actual implementations are in the very beginning stages of
development. Now is a good time for users and designers of equipment that will
link to the cloud to look into just how they will secure all of the data.
Recent
surveys and studies indicate, though, that companies are not as focused on data
security as they should be. For example, according to findings from “The 2016
Global Cloud Data Security Study” study from Ponemon Institute, organizations
and companies are not adopting appropriate control and security measures to
protect sensitive data they store in the cloud. The study surveyed more than
3,400 IT and IT security practitioners worldwide to gain a better understanding
of trends in data collection and security practices for cloud-based services.
They
found that:
•
Half of all cloud services and corporate data stored in cloud are not
controlled by IT departments.
•
Only a third of sensitive data stored in cloud-based applications are
encrypted.
•
More than half of companies do not have a proactive approach for compliance
with privacy and security regulations for data in cloud environments.
“Cloud
security continues to be a challenge for companies, especially in dealing with
the complexity of privacy and data protection regulations,” said Dr. Larry Ponemon,
chairman and founder, Ponemon Institute. “To ensure compliance, it is important
for companies to consider deploying such technologies as encryption,
tokenization or other cryptographic solutions to secure sensitive data
transferred and stored in the cloud.”
Agreed
Jason Hart, Vice President and Chief Technology Officer for Data Protection at
Gemalto, a leader in digital security, “It’s quite obvious security measures
are not keeping pace because the cloud challenges traditional approaches of
protecting data when it was just stored on the network. It is an issue that can
only be solved with a data-centric approach in which IT organizations can
uniformly protect customer and corporate information across the dozens of
cloud-based services their employees and internal departments rely every day.”
The
state of IoT security today
Thus,
working with IT departments will be key to securing cloud data. But, the study
found that nearly half (49%) of cloud services are deployed by departments
other than corporate IT, and an average of 47% of corporate data stored in
cloud environments are not managed or controlled by the IT department. Until
such time as individual companies come up with a policy, engineers may have to
take a proactive approach and initiate conversations with customer IT
departments early in the design phase.
Just
what kind of security measures are needed? 54% of survey respondents felt it
was more difficult to protect confidential or sensitive information when using
cloud services. 53% of respondents report difficulty in controlling or
restricting end-user access. The other major challenges include the inability
to apply conventional information security in cloud environments (70% of
respondents) and the inability to directly inspect cloud providers for security
compliance (69% of respondents).
Customer
information stored in the cloud is most at risk. According to the survey,
customer information, emails, consumer data, employee records and payment
information are the types of data most often stored in the cloud. Since 2014,
cloud storage of this information has increased from 53% in 2014 to 62% today.
53% considered customer information data to be the most at risk in the cloud.
The
majority of respondents (64%) said their organizations do not have a policy
that requires use of security safeguards, such as encryption, as a condition to
using certain cloud computing applications. This situation challenges designers
during product design.
72%
of respondents said the ability to encrypt or tokenize sensitive or
confidential data is important, with 86% saying it will become more important
over the next two years, up from 79% in 2014.
Yet,
passwords and similar conventional security measures are no longer adequate.
67% of respondents said the management of user identities is more difficult in
the cloud than on-premises. However, organizations are not adopting measures
that are easy to implement and could increase cloud security. About half (45%)
of companies are not using multi-factor authentication to secure employee and
third-party access to applications and data in the cloud, which means many
companies are still relying on just user names and passwords to validate
identities. This puts more data at risk because 58% of respondents say their
organizations have third-party users accessing their data and information in
the cloud.
Easier
security solutions on the way
In
some cases, communication developers are adding features that are easy for
design engineers to incorporate into their designs, helping improve security.
One
example is the PAC Project 9.5, which provides updated firmware for Opto 22
SNAP PAC S-series and R-series controllers that enable a secure HTTPS server on
PAC controllers. Combined with a RESTful open and documented API, it allows
developers to write applications that access data on the PAC using the
developer’s programming language of choice with the JSON data format. This new
capability allows software and IoT application developers to eliminate layers
of middleware for secure Industrial Internet of Things (IIoT) applications.
Firmware
version 9.5 for SNAP PAC R-series and S-series controllers enables REST
endpoints for analog and digital I/O points as well as control program
variables including strings, floats, timers, integers, and tables. REST
endpoints are securely accessed using the RESTful API for SNAP PACs.
Client
data requests are returned in JavaScript Object Notation (JSON) format. PAC
controllers and I/O can be used with almost any software development language
with JSON support, including C, C++, C#, Java, JavaScript, node.js, Python,
PHP, Ruby, and many more. They can use the development environment and language
of their choosing to write new software, create web services, and build
Internet of Things applications.
The
addition of a secure RESTful server and an open, documented API to a
programmable automation controller (PAC) is a significant industry innovation,
because REST architecture and associated technology are intrinsic to the
Internet of Things and paramount to web and mobile-based application
development. Opto 22’s implementation of REST directly into a commercially
available, off-the-shelf industrial PAC places the company as one of the first
industrial automation and controls manufacturer to offer this industry-changing
technology.
More
IoT solutions
The
UNO-1251G is a DIN-rail mountable IoT Gateway from Advantech’s IIoT Automation
Group. It’s about the size of a micro PLC. For accessibility, the industrial
computer comes with a programmable OLED display, a wireless communication slot,
and built in CANbus protocol. It supports over 450 PLCs, controllers, and I/O
device protocols with WebAccess/HMI software.
This
gateway is suitable for networking intelligent I/O devices such as sensors and
actuators. To aid development of CANbus applications, the UNO-1251G includes
the Advantech CANopen protocol library, which provides a C application
programming interface (API) for configuring, starting, and monitoring CANopen
devices. (Know More)
Support
for 32-bit ARM® processors available immediately, with PowerPC and MIPS added
to the Zulu Embedded roadmap
- Azul
is now the largest commercial vendor delivering multi-platform, 100% open
source OpenJDK builds targeting embedded and IoT markets
- Fully-supported
Zulu Embedded builds of OpenJDK for 32-bit ARM processors are now
available immediately
- Zulu
Embedded builds supporting 32-bit PowerPC processors will be delivered in
Q3 2016, plus 64-bit ARM v8 support in Q1 2017
- Supported operating systems include multiple Linux distros plus Windows 10 IoT
SUNNYVALE, Calif.-- Azul Systems
(Azul), the award-winning leader in Java runtime solutions, today announced
Zulu Embedded with expanded support for multiple 32 and 64-bit embedded
processors. Zulu Embedded, Azul's 100% open source Java Virtual Machine (JVM)
based on OpenJDK, now includes support for ARM, Intel, PowerPC and MIPS
processors. With today’s announcement, Azul is highlighting its full embedded
Java SE roadmap, which spans a variety of processor architectures and operating
systems, offers exceptional performance via optimized JIT compilation, supports
various form factors and compact profiles, and provides a pure open source,
standards-compliant solution for embedded Java.
General Availability: Zulu Embedded for
ARM® V7 and V8
Zulu Embedded brings a variety of
choices to product teams and manufacturers who integrate Java for their
embedded and IoT-centric designs. Java is used virtually everywhere in the
embedded world, ranging from manufacturing and logistics to security,
networking, wireless, mobile and many other embedded applications.
Fully-supported Zulu Embedded builds of
OpenJDK for 32-bit ARM are now available immediately, with builds supporting
32-bit PowerPC processors to be delivered in Q3 2016, plus 64-bit ARM v8
support in Q1 2017. Azul is also working with Cavium on Zulu Embedded support
for 64-bit versions of the MIPS processor architecture. Zulu Embedded has
supported both 32 and 64-bit x86 processors since its initial launch in early
2015.
“Highly efficient 32-bit ARM processors
power a diverse range of embedded applications in automotive, TV, gateways,
payment systems, and more," said Dominic Pajak, Marketing Director,
Embedded Segments, ARM. "OpenJDK-based products like Zulu Embedded greatly
enhance the Java development experience and add to the wide choices in the ARM
ecosystem.”
“Azul’s Java expertise is well
recognized in the industry, and the commitment to bring Zulu JVM to 64-bit
ARMv8 and MIPS architectures will accelerate the efficiency and optimization of
Java for Cavium customers,” said Larry Wikelius, Vice President Software
Ecosystem and Solutions, Cavium. “Java based applications continue to be
critical to our OCTEON® and ThunderX® installed base and Zulu will provide the
commercial options that many end users require. Azul is a welcome addition to
Cavium’s thriving software ecosystem partner community.”
“Java is a key enabling technology for
the Internet of Things,” said Christian Renaud, Research Director, Internet of
Things, 451 Research. “Many IoT applications demand real-time performance, and
Zulu Embedded’s new support of 32-bit ARM architecture enables a number of
opportunities and new applications in smart grid, healthcare, transportation,
and many other applications.”
“Canonical and Azul are working
together to ensure that Java developers building IoT solutions for Ubuntu Core
can easily package their applications as snaps and incorporate Zulu Embedded
using the new snapcraft tooling,” said Maarten Ectors, VP of IoT at Canonical.
"The Eclipse Foundation's Internet
of Things Working Group has a number of Java-based open source IoT projects
that will benefit from running on Zulu Embedded for 32-bit ARM processors. Zulu
Embedded will be an important open source option for Java developers building
IoT solutions," said Mike Milinkovich, Executive Director of Eclipse
Foundation.
“Today’s announcement of the immediate
availability of Zulu Embedded with support for 32-bit ARM processors and our
extended embedded roadmap highlight Azul’s role as the industry’s leading
provider of multi-platform, open source Java solutions,” said Scott Sellers,
Azul Systems president and CEO. “The embedded and IoT markets are exploding
with new products and devices, and Zulu Embedded provides the industry’s only
open source, standards compliant solution for Java for these markets. By
offering a wide range of processor and operating system support, and delivering
customized builds to meet specific embedded requirements, Zulu Embedded is
quickly become the de facto standard for developing and commercializing
Java-based solutions in the embedded, maker and IoT markets.”
ARM microchips are used by millions of people to connect everyday
devices and appliances over the internet. Arm’s designs lie at the heart
of almost every smartphone sold today. Many cheaper phones use its
off-the-shelf chips but for more sophisticated and expensive handsets,
chipmakers often customise Arm’s technology to produce more distinctive
features. With Intel backing out of the mobile processor market earlier
this year to focus on servers, modems and the “internet of things”,
there are few viable alternatives to Arm.
Television
Many
modern televisions enable users to watch programmes and films through apps such
as Netflix, which are powered by ARM-based processors. The firm’s technology is
also frequently used in remote controls and set-top boxes.
Smartphones
and tablets
ARM
technology helps power both smartphones and iPads. E-readers, such as Kindles,
and digital cameras also use the products.
Home
‘smart’ systems
Smart
household appliances often use ARM-based chips to give homeowners greater
control over their functions – and costs. This can include internet-connected
thermostats, allowing users to control their heating via their smartphone,
fridges that alert users when they have run out of groceries, and electricity
metres that can help save money on energy bills.
Wearable
gadgets
Fitness
trackers and smart watches are becoming increasingly popular with consumers.
These gadgets have relied on ARM technology for years.
Internal
car systems
ARM-based
chips are frequently used in cars to show drivers maps, offer voice recognition
and to control music. They are also being used in prototypes for self-driving
cars to power systems that trigger automatic electronic braking, for example.
The technology could also be used in smart roads warning drivers about spots of
black ice ahead.
Drones
Drones
are becoming increasingly popular, be it families flying them in the park or
photographers capturing aerial shots. They rely on tiny computers called
microcontrollers and ARM estimates that a quarter of the chips made last year
used its technology.
Energy
efficient cities
Street
lamps that dim themselves and parking meters that detect when spaces are empty
are examples of sensors being used by cities to cut costs and help inhabitants.
ARM believes this sector holds great potential for its business.
