Page 22 - EETEurope FlipBook February
P. 22
22 EE|Times EUROPE — CES 2021
NXP BlueBox 3.0 Lets Carmakers Build Safety, Portfolio of Apps
make the ‘server on the wheels’ concept
more of a reality by deploying [BlueBox
3.0] in the automotive realm, where
safety and security are fundamental
pillars of NXP’s portfolio.”
SOFTWARE ORCHESTRATION
All of the hardware building blocks for
BlueBox 3.0 are ready, according to NXP,
but it will take until the end of the first
quarter for the company to finalize all
the software elements that sit above the
basic SDK layer. Those software elements
are essential for BlueBox 3.0 to orches-
trate data movement among multiple
devices on the platform, said Örs.
For example, the data stream input for
the cameras goes to the Kalray accelerator
(Source: NXP Semiconductors)
for object identification and classification.
That data must then be fed into a planner
BLUEBOX 3.0 compute module running on a large CPU device (i.e., LX2160A).
NXP, however, believes the new BlueBox has the computing horse- “That data path orchestration, though the data is moving on a PCIe,
power to serve the needs of future vehicles. Örs noted that “the must be carried out efficiently,” said Örs. “You need to make sure that
LX2168A has 16 high-performance ARM Cortex-A72 cores. We have the timing of each data packet coming in makes sense. Then it’s S32G
S32G274 for the current-generation gateway. It is somewhat flexible, that validates that those actuation commands are sane.”
since we can upgrade on that slot in the future without changing the
BlueBox itself.” He added that S32G allows for ASIL-D compliance, ECOSYSTEM
which can act as both a safety co-processor and a gateway device NXP takes pride in creating a “close knit” partner ecosystem to provide
for automotive. “a quick ramp-up” and “an example that system designers can work
NXP is launching BlueBox 3.0 with Kalray as the accelerator part- from,” said Örs. In addition to Kalray, partners backing BlueBox 3.0
ner. “Kalray can have up to 160 CPU accelerator cores for neural net include dSPACE, Embotech, Edge Case Research (ECR), eProsima, Green
and math acceleration as needed,” said Örs. Hills Software (GHS), Intempora, Micron Technology, MicroSys, Real-
But what makes BlueBox 3.0 stand out is its doubling of PCIe slots Time Innovations (RTI), and Teraki. ■
to six, from the three offered with the BlueBox 2.0.
Leveraging the terminology “high-performance compute” (HPC) Junko Yoshida is global editor-in-chief of AspenCore. This article was
often used in the server domain, Örs explained that NXP is “trying to originally published on EE Times.
BOOK REVIEW
Sensors in Automotive: Book Does More Than Explain Tech
By Ed Niedermeyer
Automated-vehicle technology is for anyone in need of a one-stop education Sensors in Automotive” remain valuable
extremely complex, making it an intrinsi- on automotive sensors. even after the technology has begun to
cally challenging and intimidating topic for Though technology changes fast, this book move on. ■
educators and the media. Compounding this not only provides an overview at a pivotal
challenge is the lack of credible, independent moment, when autonomous-vehicle devel- Ed Niedermeyer is the commu-
reference material from which educators can opment appears to be maturing into actual nications director for Partners
draw trustworthy information. driverless services, but explains the trends for Automated Vehicle Education
With “AspenCore Guide to Sensors in that are shaping that change as well. (PAVE), a nonprofit backed by the
Automotive: Making Cars See and Think Critically for a book in this space, it goes auto and technology sectors that
Ahead,” AspenCore global editor-in-chief beyond just explaining individual technol- works to educate the public, media, and policy-
Junko Yoshida and her top-notch band of ogies and provides the broader context of makers. Niedermeyer is also the co-host of The
collaborators have stepped up to fill this gap how they work within automated driving Autonocast mobility technology podcast.
in one of the most important aspects of the systems, how they complement and compare
technology. Leveraging the impeccable cred- with each other, and why their strengths and “AspenCore Guide to Sensors in Automotive:
ibility, deep connections in the industry, and weaknesses matter. Making Cars See and Think Ahead” is
knack for explaining challenging technical This context, which explains percep- available now at the EE Times bookstore.
topics that she is known for, Yoshida has tion and automated driving systems more Visit www.eetimes.com/shop for more
brought together an indispensable resource broadly, will help “AspenCore Guide to information and to purchase.
FEBRUARY 2021 | www.eetimes.eu

