Showing posts with label ARM events. Show all posts
ARM is beefing up its safety technology for the ARM Cortex-R52, a processor designed for self-driving vehicles.
The
Cambridge, England-based company was recently acquired by Japan’s SoftBank for
$31 billion. And now it is expanding its chip designs to include a processor
with the robust, real-time performance needed for autonomous cars.
The
new chip will simplify the path for certification of automotive applications,
industrial robots, and medical operations. The processor — which ARM will
license to other chip manufacturers — must comply with tough safety standards
such as ISO 26262 ASIL D and IEC 61508 SIL 3.
It
will enable applications as diverse as surgical automation, safety management
and automotive powertrain control. STMicroelectronics is the first ARM chip
manufacturing partner to announce it has licensed the high-performance
processor to enable it to create highly integrated system-on-chips (SoCs) for
the automotive market.
“If
these systems go wrong in any way, they can affect life,” said Richard York,
worldwide marketing and business development manager at ARM, in an interview
with VentureBeat. “The R-52 will make it much easier to do increasingly complex
software.”
The
Cortex-R52 offers hardware-enforced separation of software tasks to ensure
safety-critical code is fully isolated. So when one part of the system goes
down, it’s easier to decipher what happened. This allows the hardware to be
managed by a software hypervisor policing the execution and resourcing of
tasks. By enabling the precise and robust separation of software, the
Cortex-R52 decreases the amount of code that must be safety-certified, so
speeding up development as software integration, maintenance and validation is
easier. The processor also deals with increased software complexity while
delivering the determinism and fast context switching that real-time systems
demand.
“The
Cortex-R52 supports our Smart Driving vision by enabling a new range of
high-performance, power-efficient SoCs for any in-vehicle application demanding
real-time operation and the highest levels of functional safety, including
powertrain, chassis and ADAS,” said Fabio Marchiò, Automotive & Discrete
group vice president and Automotive Digital Division general manager at
STMicroelectronics, in a statement. “The Cortex-R52’s ability to
compartmentalize software provides our users with the best solution for safety
without loss of determinism. Its virtualization support simplifies the consolidation
of applications and functions into a single processor, delivering a shorter
integration time.”
Denso,
a leading global supplier of advanced automotive technology, is supporting the
launch.The availability of ARM Fast Models and Cycle Models enables software
partners to develop solutions for the processor. They further speed the path to
market as software developers will get access to the Cortex-R52 early in the
design process.
The
Cortex-R52 is 35 percent faster than the previous generation Cortex-R5, which
is already deployed in a range of safety applications. It has achieved a score
of 1.36 Automark/MHz on the EEMBC AutoBench, the highest in its class, using
the Green Hills Compiler 2017.
The
processor has protections against various kinds of random errors, design
errors, and software errors. If it senses a system problem, the processor has
to handle cores such as shutting down a vehicle and bringing it to a safe halt.
So it has to be able to sense, perceive and analyze, make a decision, and acutate
(or execute on that decision).
James
Scobie, product manager for the Cortex-R52, said chips based on the design will
likely be available in 2018.
Intel's second-in-command Venkata Renduchintala is feeling at home with his new company after he switched over from Qualcomm
Venkata Renduchintala is president of Intel's Client and Internet of Things (IoT) businesses and Systems Architecture Group.
Intel is now more than just a PC company. At industry events, the company's keynotes feature drones flying around, robots walking on stage and musicians creating tunes from wearables. The chip maker is helping BMW build an autonomous car, will sell modems to Apple, and is leading the development of next-generation 5G cellular networks. For all these new markets, it will provide chip and data-center technologies.
The transformation is happening partly under the leadership of Venkata Renduchintala, president of the Client and Internet of Things (IoT) Businesses and Systems Architecture Group at Intel. As Intel's second-in-command, he helped cut struggling products like mobile CPUs and sharpened the company's focus on IoT, servers, and connectivity.
Hired from rival Qualcomm late last year, he's an outsider trying to rid Intel of its historical resistance to change. He's also bringing fresh ideas and wholesale changes to Intel, which promises to bring a new dynamic to the Silicon Valley institution.
IDG News Service spoke with him on a range of topics including VR headsets, IoT, autonomous cars, competitors and the decision to cut products. This is an edited version of the discussion.
IDGNS: How have you settled into your new job? What drew you to Intel?
It's been a really interesting process of acclimation. It's a great mixture of feeling, like an organization where I think my experience and my interests can really help the journey [CEO] Brian [Krzanich] wants to undertake with the company. The scale at which Intel can play is probably going to be very difficult for others to match if you look across, client, networking and the data center groups. The goal is to be able to think as one Intel.
IDGNS: There have been questions on how you would fit into Intel, which has a closed culture and history of promoting executives internally. Many people hired from external companies haven't worked out.
One thing that's really important to understand is that Intel is a company of tremendous heritage. I'm not coming in to fix anything. I'm coming in hopefully to add another dimension and an important ingredient to the management team that Brian has at his disposal. It requires me to respect what Intel has been able to achieve and the caliber of the management team and the brands assimilated. I don't think Brian hired me to maintain the status quo. I think what he wanted was a strong ingredient of outside-in thinking complementing the original thoughts. I'm feeling very comfortable now in being able to feel like I've got a good bunch of colleagues who know where I'm coming from; we can speak straight to each other and we can actually have really good discussions of meritocracy.
IDGNS: You had to make some decisions on cutting products Intel has worked on for years as the company's priorities were reset. How tough was it?
When you come into a company you have a degree of objectivity that isn't tainted by your attachments to the genesis of certain projects. For me it was a fairly structured, objective discussion where you make decisions in a transparent and open manner. As long as you can walk people through your thinking, you can take what was very controversial and make it very logical. I'm passionate about technology but I'm also passionate about profitability and how the two are married in a seamlessly reinforcing way.
IDGNS: What's the reasoning behind cutting mobile processors to focus on modems?
First of all, we rationalized what we were spending our R&D on. We had a couple of mobile SoC products that I don't think were worthy to continue to conclusion. That doesn't mean to say we're no longer doing mobile platforms. On the mobile platform side, my commitment is to talk less and do more. When we have something to say we'll talk about it.
On the modem side, it's a fundamental technology and this is where I think it comes down to being as indelible for us as our competence in CPU or GPU. We've set ourselves up with a very interesting road-map, but more importantly, we've established a degree of credibility, relevance and importance as a key technology partner with a number of key players in the industry that I think is really important.
IDGNS: What are your top priorities and goals?
I have three uber-level goals. One is to continue to drive our client computing business to a position of stable profitability in the face of a slowly declining [market]. I think we're doing well in that area. The second is to grow and scale our IoT business from something that's very interesting to something that's really substantial in the longer term. The IoT business for us is a microcosm of the entire company coming together -- we're creating a type of all-for-one, one-for-all mentality. The third is to maintain a degree of vibrancy in the technology leadership of our entire systems architecture organization. It's developing all the core technologies that really moves the competitive needle forward.
IDGNS: Intel's untethered mixed-reality headset called Project Alloy was big news at IDF. What are the expectations from Alloy and how are things going?
The whole point of having tetherless VR is a big deal. Everything we're doing in Alloy we're going to open-source. We can take VR and evolve it from the very rudimentary definitions today of [VR] in a smart phone that you clip into some kind of visor. You can move it to a capable, embedded PC that's driving two to three teraflops of computing and generate a really immersive experience. That was really it -- taking ideas out from the lab, productizing them, solving all those problems of integration, figuring out how RealSense and depth camera fits into all of that, figuring out how to do merged reality, and saying "now go scale the ecosystem."
IDGNS: Is the VR headset the new PC?
I think it's another very interesting growth opportunity for the PC. I think it can generate a specific class of products in its own right. It will generate different segmentation points and probably a custom piece of silicon built on the PC platform that amplify the use case. So we're very excited about the whole VR space.
IDGNS: Intel hasn't given up on Moore's Law, though many believe it is reaching its end. How is Intel preparing for a future when manufacturing reaches atomic scale, and how will chips look beyond Kaby Lake?
Nobody inside Intel is coming anywhere near the kind-of-like fatalistic conclusions about where Moore's Law is. Intel has had a stellar track record in delivering node generation like clockwork. Maybe we've moved from a two-year to a two-and-a-half-year cadence, but we already see light at the end of the tunnel. We will continue to drive process technology and nobody is calling timeout on anything. We're working hard on 7-nanometer, we're talking about pathfinding for 5-nanometer. All of that is in the throes. We made a great announcement on Kaby Lake -- that's using an evolution of 14-nanometer transistor geometry that gave a substantially improved user experience compared to Skylake. We're going to continue to do more of that as we continue to drive process leadership.
IDGNS: Are you happy with your current chip line-up -- Kaby Lake for PCs, and Atom for IoT?
We have a competent portfolio of products. I'm in no way shape or form concluding they are complete and aren't going to be benefited from augmentation. For me I think it's really wanting to understand the use cases a lot more. I don't see an IoT strategy for Intel being one where everything is delivered by Intel. It's integrating a number of different technologies that could be indigenous to Intel, or could be created by other companies, but managed in a way where people could look at Intel as somebody providing the overarching framework of integration.
IDGNS: IoT is a big part of Intel's future. What's the strategy for that market?
That's a significant business. I think we're just starting. As you see the advent of autonomous driving vehicles, you see robots and drones start to ship in scale: those are very high value opportunities for us. We characterize our IoT interests into three verticals: industrial, transportation and retail -- all of them have an end-to-end dimension where we're providing a client environment, the networking infrastructure and the data analytics platform that drives all of that through industry partnerships.
IDGNS: Would in any way the ARM foundry deal help Intel achieve its goals in IoT and other areas? Would you be open to the idea of taking an ARM CPU license, as an example?
Open to? Yes. My view is fairly straightforward -- that Intel's IoT plan has to not only be able to harmoniously integrate Intel-based microprocessors and MCUs, it has to be able to aggregate and harmoniously integrate a plethora of different types of MCUs, whether it be ARM-based, MIPS-based, or proprietary MCUs. All of them have the ability to monitor, sense data that they want to get on to an information highway of some kind. Our ability to [support] many different client environments is going to be a necessity in any vertical IoT strategy we have. There are many areas in the ARM ecosystem where Intel can pragmatically play in for its own benefit. I'm a big believer in paying respect to established ecosystems.
IDGNS: Self-driving cars are a big deal for Intel. Could you talk about projects in the pipeline?
Our goal is to provide the type of computing power that dwarfs anything that exists in a car today, but basically make it mainstream. What we're doing on our Xeon Phi processor for machine learning and deep learning, what we're doing in computer vision and also supplemented by radar and lidar. Being able to aggregate that data, generate intelligence, make decisions on it with assistance from machine and deep learning algorithms -- that's all happening as we speak.
IDGNS: How do you see the autonomous car market evolving?
I see the first explosive area to be in the urban transportation environment where services like Uber and Lyft will evolve and develop. There's going to be a lot of experimentation and path-finding to do in addition to technology creation. We're probably talking about a decade away. Stamina to invest is going to be really important; those that have the stamina to stay the course are going to win big.
IDGNS: Nvidia is approaching the automotive markets aggressively with its GPUs, how will you compete?
I have a great deal of respect for Nvidia. But every time I think of Nvidia, I think about Californian wine where they can make great wine but it contains only one grape -- great Cabernet Sauvignon or a great Chardonnay. I love French wines and French wines are blends where you need to be great at growing Cabernet, great at growing Merlot, great at growing Cabernet Franc. The art is in the mixture. That's the benefit Intel has. We have GPU, we have CPU, we have custom silicon, we have embedded storage, we have FPGA. Nvidia's going to basically say "I've got GPUs and I've got GPUs and I've got GPUs." Great strategy, but it doesn't give anywhere near the extensibility, flexibility and scalability that Intel is able to offer.
IDGNS: How will 5G influence changes in the way devices are made and work?
5G is as much about the transformation of the network and the infrastructure as it is the client environment. [There is] going to be an even greater demand from mobile broadband bandwidth, people are going to want tens of gigabytes per second, if not hundreds of gigabytes per second. We're going to see much greater pervasiveness of client devices. If you talk about autonomous vehicles or delivering health services over a mobile network, you need to be able to make life or death decisions based on that. The network has to transform and the data center becomes a much higher order entity that's focused on massive data analytics that orchestrates that entire network.
The transformation is happening partly under the leadership of Venkata Renduchintala, president of the Client and Internet of Things (IoT) Businesses and Systems Architecture Group at Intel. As Intel's second-in-command, he helped cut struggling products like mobile CPUs and sharpened the company's focus on IoT, servers, and connectivity.
Hired from rival Qualcomm late last year, he's an outsider trying to rid Intel of its historical resistance to change. He's also bringing fresh ideas and wholesale changes to Intel, which promises to bring a new dynamic to the Silicon Valley institution.
IDG News Service spoke with him on a range of topics including VR headsets, IoT, autonomous cars, competitors and the decision to cut products. This is an edited version of the discussion.
IDGNS: How have you settled into your new job? What drew you to Intel?
It's been a really interesting process of acclimation. It's a great mixture of feeling, like an organization where I think my experience and my interests can really help the journey [CEO] Brian [Krzanich] wants to undertake with the company. The scale at which Intel can play is probably going to be very difficult for others to match if you look across, client, networking and the data center groups. The goal is to be able to think as one Intel.
IDGNS: There have been questions on how you would fit into Intel, which has a closed culture and history of promoting executives internally. Many people hired from external companies haven't worked out.
One thing that's really important to understand is that Intel is a company of tremendous heritage. I'm not coming in to fix anything. I'm coming in hopefully to add another dimension and an important ingredient to the management team that Brian has at his disposal. It requires me to respect what Intel has been able to achieve and the caliber of the management team and the brands assimilated. I don't think Brian hired me to maintain the status quo. I think what he wanted was a strong ingredient of outside-in thinking complementing the original thoughts. I'm feeling very comfortable now in being able to feel like I've got a good bunch of colleagues who know where I'm coming from; we can speak straight to each other and we can actually have really good discussions of meritocracy.
IDGNS: You had to make some decisions on cutting products Intel has worked on for years as the company's priorities were reset. How tough was it?
When you come into a company you have a degree of objectivity that isn't tainted by your attachments to the genesis of certain projects. For me it was a fairly structured, objective discussion where you make decisions in a transparent and open manner. As long as you can walk people through your thinking, you can take what was very controversial and make it very logical. I'm passionate about technology but I'm also passionate about profitability and how the two are married in a seamlessly reinforcing way.
IDGNS: What's the reasoning behind cutting mobile processors to focus on modems?
First of all, we rationalized what we were spending our R&D on. We had a couple of mobile SoC products that I don't think were worthy to continue to conclusion. That doesn't mean to say we're no longer doing mobile platforms. On the mobile platform side, my commitment is to talk less and do more. When we have something to say we'll talk about it.
On the modem side, it's a fundamental technology and this is where I think it comes down to being as indelible for us as our competence in CPU or GPU. We've set ourselves up with a very interesting road-map, but more importantly, we've established a degree of credibility, relevance and importance as a key technology partner with a number of key players in the industry that I think is really important.
IDGNS: What are your top priorities and goals?
I have three uber-level goals. One is to continue to drive our client computing business to a position of stable profitability in the face of a slowly declining [market]. I think we're doing well in that area. The second is to grow and scale our IoT business from something that's very interesting to something that's really substantial in the longer term. The IoT business for us is a microcosm of the entire company coming together -- we're creating a type of all-for-one, one-for-all mentality. The third is to maintain a degree of vibrancy in the technology leadership of our entire systems architecture organization. It's developing all the core technologies that really moves the competitive needle forward.
IDGNS: Intel's untethered mixed-reality headset called Project Alloy was big news at IDF. What are the expectations from Alloy and how are things going?
The whole point of having tetherless VR is a big deal. Everything we're doing in Alloy we're going to open-source. We can take VR and evolve it from the very rudimentary definitions today of [VR] in a smart phone that you clip into some kind of visor. You can move it to a capable, embedded PC that's driving two to three teraflops of computing and generate a really immersive experience. That was really it -- taking ideas out from the lab, productizing them, solving all those problems of integration, figuring out how RealSense and depth camera fits into all of that, figuring out how to do merged reality, and saying "now go scale the ecosystem."
IDGNS: Is the VR headset the new PC?
I think it's another very interesting growth opportunity for the PC. I think it can generate a specific class of products in its own right. It will generate different segmentation points and probably a custom piece of silicon built on the PC platform that amplify the use case. So we're very excited about the whole VR space.
IDGNS: Intel hasn't given up on Moore's Law, though many believe it is reaching its end. How is Intel preparing for a future when manufacturing reaches atomic scale, and how will chips look beyond Kaby Lake?
Nobody inside Intel is coming anywhere near the kind-of-like fatalistic conclusions about where Moore's Law is. Intel has had a stellar track record in delivering node generation like clockwork. Maybe we've moved from a two-year to a two-and-a-half-year cadence, but we already see light at the end of the tunnel. We will continue to drive process technology and nobody is calling timeout on anything. We're working hard on 7-nanometer, we're talking about pathfinding for 5-nanometer. All of that is in the throes. We made a great announcement on Kaby Lake -- that's using an evolution of 14-nanometer transistor geometry that gave a substantially improved user experience compared to Skylake. We're going to continue to do more of that as we continue to drive process leadership.
IDGNS: Are you happy with your current chip line-up -- Kaby Lake for PCs, and Atom for IoT?
We have a competent portfolio of products. I'm in no way shape or form concluding they are complete and aren't going to be benefited from augmentation. For me I think it's really wanting to understand the use cases a lot more. I don't see an IoT strategy for Intel being one where everything is delivered by Intel. It's integrating a number of different technologies that could be indigenous to Intel, or could be created by other companies, but managed in a way where people could look at Intel as somebody providing the overarching framework of integration.
IDGNS: IoT is a big part of Intel's future. What's the strategy for that market?
That's a significant business. I think we're just starting. As you see the advent of autonomous driving vehicles, you see robots and drones start to ship in scale: those are very high value opportunities for us. We characterize our IoT interests into three verticals: industrial, transportation and retail -- all of them have an end-to-end dimension where we're providing a client environment, the networking infrastructure and the data analytics platform that drives all of that through industry partnerships.
IDGNS: Would in any way the ARM foundry deal help Intel achieve its goals in IoT and other areas? Would you be open to the idea of taking an ARM CPU license, as an example?
Open to? Yes. My view is fairly straightforward -- that Intel's IoT plan has to not only be able to harmoniously integrate Intel-based microprocessors and MCUs, it has to be able to aggregate and harmoniously integrate a plethora of different types of MCUs, whether it be ARM-based, MIPS-based, or proprietary MCUs. All of them have the ability to monitor, sense data that they want to get on to an information highway of some kind. Our ability to [support] many different client environments is going to be a necessity in any vertical IoT strategy we have. There are many areas in the ARM ecosystem where Intel can pragmatically play in for its own benefit. I'm a big believer in paying respect to established ecosystems.
IDGNS: Self-driving cars are a big deal for Intel. Could you talk about projects in the pipeline?
Our goal is to provide the type of computing power that dwarfs anything that exists in a car today, but basically make it mainstream. What we're doing on our Xeon Phi processor for machine learning and deep learning, what we're doing in computer vision and also supplemented by radar and lidar. Being able to aggregate that data, generate intelligence, make decisions on it with assistance from machine and deep learning algorithms -- that's all happening as we speak.
IDGNS: How do you see the autonomous car market evolving?
I see the first explosive area to be in the urban transportation environment where services like Uber and Lyft will evolve and develop. There's going to be a lot of experimentation and path-finding to do in addition to technology creation. We're probably talking about a decade away. Stamina to invest is going to be really important; those that have the stamina to stay the course are going to win big.
IDGNS: Nvidia is approaching the automotive markets aggressively with its GPUs, how will you compete?
I have a great deal of respect for Nvidia. But every time I think of Nvidia, I think about Californian wine where they can make great wine but it contains only one grape -- great Cabernet Sauvignon or a great Chardonnay. I love French wines and French wines are blends where you need to be great at growing Cabernet, great at growing Merlot, great at growing Cabernet Franc. The art is in the mixture. That's the benefit Intel has. We have GPU, we have CPU, we have custom silicon, we have embedded storage, we have FPGA. Nvidia's going to basically say "I've got GPUs and I've got GPUs and I've got GPUs." Great strategy, but it doesn't give anywhere near the extensibility, flexibility and scalability that Intel is able to offer.
IDGNS: How will 5G influence changes in the way devices are made and work?
5G is as much about the transformation of the network and the infrastructure as it is the client environment. [There is] going to be an even greater demand from mobile broadband bandwidth, people are going to want tens of gigabytes per second, if not hundreds of gigabytes per second. We're going to see much greater pervasiveness of client devices. If you talk about autonomous vehicles or delivering health services over a mobile network, you need to be able to make life or death decisions based on that. The network has to transform and the data center becomes a much higher order entity that's focused on massive data analytics that orchestrates that entire network.
Stealthy,
tricky to remove rootkit targets Linux systems on ARM and x86
Security
researchers have identified a new family of Linux rootkits that, despite
running from user mode, can be hard to detect and remove.
Called
Umbreon, after a Pokémon character that hides in the darkness, the rootkit has
been in development since early 2015 and is now being sold on the underground
markets. It targets Linux-based systems on the x86, x86-64 and ARM
architectures, including many embedded devices such as routers.
According
to malware researchers from antivirus firm Trend Micro, Umbreon is a so-called
ring 3 rootkit, meaning that it runs from user mode and doesn't need kernel
privileges. Despite this apparent limitation, it is quite capable of hiding
itself and persisting on the system.
The
rootkit uses a trick to hijack the standard C library (libc) functions without
actually installing any kernel objects. Libc provides system call functions
that other Linux programs can use for important operations like reading and
writing files, spawning processes or sending network packets.
Umbreon
hijacks these functions and forces other Linux executables to use its own
libc-like library. This puts the rootkit in a man-in-the-middle position,
capable of modifying system calls made by other programs and altering their
output.
The
rootkit also creates a hidden Linux account that can be accessed via any
authentication method supported by Linux, including SSH (Secure Shell). This
account does not appear in files like /etc/passwd because the rootkit can
modify the output of such files when read, the Trend Micro researchers said in
a blog post.
Umbreon
also has a backdoor component called Espeon, named after another Pokémon
character, that can establish a reverse shell to an attacker's machine when a
TCP packet with special field values are received on the monitored Ethernet
interface of an affected device. This means that attackers can open remote
shells by simply sending a specially crafted packet to the infected device over
the Internet.
It's
hard to detect Umbreon using standard Linux tools, because most of them are
written in C and rely on libc, whose output the rootkit hijacks, the Trend
Micro researchers said. "One way is to develop a small tool to list the
contents of the default Umbreon rootkit folder using Linux kernel syscalls
directly."
Removing
the rootkit from an infected system can also be tricky, especially for inexperienced
users and attempts to do so could render the system unusable, the researchers
said.
Trend
Micro provided indicators of compromise in the form of file names and hashes,
manual removal instructions and YARA detection rules for the new rootkit.
It
seems that the rootkit was designed for manual installation, which means that
attackers install it on systems manually after compromising them through other
vulnerabilities.
While
many desktop Linux systems receive automatic patches and are generally kept up
to date by users, embedded devices like consumer routers and IP-based cameras
are rarely updated.
As
a result, there are hundreds of thousands of embedded devices out there that
are vulnerable to known exploits and are routinely infected with malware. Just
last week, Web security firm Sucuri blocked a massive DDoS attack that
originated from two botnets, one made up of infected CCTV cameras and one made
up of hijacked home routers.
