Multicore Mix

A look back: Two years of Multicore Mix

Lauren Reed1 — Mon, 20 May 2013 16:09:21 GMT

A big thank you to everyone who participated in our contest last week! We loved hearing about how our blog posts have inspired you. Congratulations to David Hoelscher for submitting the winning response: “I found the three-part series about Multicore Navigator Tips 'n Tricks very helpful in understanding the various ways it can be used. In particular, using the Queue Manager for Notification purposes as a means to avoid moving data and improve efficiency was very appealing.”

We have had a great two years on the Multicore Mix. We thought it would be an appropriate time to take a look back at a few of our top blog posts and the accompanying comments over the years.

Do we really need ‘Transactional Memory’? (By one and zero) explained what Transactional Memory is, how it works and asked the question do we need it?
Anybody got tools? (By Tom Flanagan) helped familiarize our Multicore Mix readers with the next steps after selecting a snappy new multicore device!
Happy Anniversary, Multicore Mix! Celebrate by entering a MONTH of contests! Everyone loves a contest! Last year, we celebrated our first anniversary by giving away a prize bag to a fan of the Multicore Mix. We chose our favorite answer from readers who wrote why they were a Multicore Mix fan in 50 words or less.
Share your love of MULTICORE – and win BIG! This contest was all about our LOVE for the readers of the Multicore Mix. In celebration of Valentine’s Day, we asked readers to tell us why they loved multicore for a chance to win a $100 Visa gift card and a box of chocolates!
The answer is “42.” What was the question? (By one and zero) discussed programming language and how to know which is best suited for multicore architecture.

If you didn’t enter the contest last week, don’t worry! Since it’s our SECOND anniversary here at the Multicore Mix, we are going to have TWO contests this month. This time we want to know what you would like to see covered in our third year. Do you want to see more about tradeshows, design tips, tutorials or even trends in vertical markets? We want to know!

To enter this contest, comment on this post to let us know what you want to see in the next year, and our bloggers will select their favorite suggestion based on which one they feel is most relevant and interesting to our multicore audience. You have until midnight CST on Sunday, May 26, to enter! The winning comment will get a $100 Visa gift card and a multicore T-shirt. The winner will be announced on Tuesday, May 28.

CONTEST DETAILS: This contest is open to all residents of the United States. TI employees and their family members cannot enter the contest. Prizes will be shipped via Fedex Ground and cannot be shipped to P.O. Box addresses. Entries will be judged based on what our Multicore Mix bloggers consider to be the best comment on what they want to see in the Multicore Mix this next year. The selected winner must respond to the email notification within 72 hours or they will forfeit the $100 Visa gift card and multicore T-shirt. If the selected winner does not respond within 72 hours, an alternate winner will be chosen based on originality and content of response by Multicore Mix bloggers.

It’s our second anniversary, but you get the present!

Lindsey Bare — Mon, 13 May 2013 16:27:55 GMT

It’s hard to believe it’s already been two years since we started “Multicore Mix!” Time flies when you’re having fun, though, right? We’ve heard from so many experts, from TIers to partners to customers, who have shared their knowledge and tips & tricks. It’s our goal to give you the information you want and need as you’re developing new products and considering TI as your multicore processor supplier.

To celebrate our second anniversary, we want to reward you for reading our blog. Over the next week in the comments section below, we want to hear how posts on “Multicore Mix” have helped you. Did a post help you solve a design challenge? Did a post inspire new innovation? Did you discover new resources on TI.com that you previously didn’t know about? You have until midnight CST on Sunday, May 19 to enter.

One winner will be chosen on Monday, May 20, and will receive a $100 Visa gift card in addition to a multicore tshirt. (In case you didn’t know, the two year anniversary gift is cotton!)

So, tell us how we’re helping, and you can win!

CONTEST DETAILS: This contest is open to all residents of the United States. TI employees and their family members cannot enter the contest. Prizes will be shipped via Fedex Ground and cannot be shipped to P.O. Box addresses. Entries will be judged based on what our Multicore Mix bloggers consider to be the best comment on how the Multicore Mix has helped you. The selected winner must respond to the email notification within 72 hours or they will forfeit the $100 Visa gift card and multicore shirt. If the selected winner does not respond within 72 hours, an alternate winner will be chosen based on originality and content of response by Multicore Mix bloggers.

Limited time offer: Save $100 on Keystone-based EVM!

tscheck — Tue, 30 Apr 2013 20:50:06 GMT

Have you been thinking about ordering a TI Keystone-based EVM and exploring the highest performance DSP core on the market? Now through May 31 (offer recently extended), you can save 25 percent on TI’s TMS320C6657LS EVM, an easy-to-use, cost-efficient development platform for designs using TI's KeyStone-based C6657, C6655 or C6654 multicore DSPs.

For only $299, this EVM:

Includes an on-board, single C6657 processor (dual-core) with robust connectivity options
Supports multiple boot interface options, including NAND, NOR, Ethernet, Serial RapidIO and PCIe
Supplies on-board 512MB DDR3 memory
Provides an AMC form factor card that can be used in uTCA systems – or as a standalone development platform
Comes with the Multicore Software Development Kit (MCSDK), providing the needed software libraries and platform software building blocks for your application
Comes with Code Composer Studio™ integrated development environment and needed license for EVM programming development and debugging
Features XDS200 emulation capability via standard USB interface

The C665x devices were designed to deliver real-time high performance at low power coupled with smaller form factors so developers could better meet the requirements of markets such as industrial automation, mission critical, imaging, medical, audio and video infrastructure and testers. The C6657/55 platform devices are pin-to-pin compatible, offering an easy migration path and scalable solution. They are also fully compatible with all fixed- and floating-point C6000 devices.

When partnered with the MCSDK, you can quickly begin programming on this EVM, starting with the multiple demos and examples wrapped around our low-level drivers and software libraries. Find support and large knowledge base for our MCSDK software and EVM on the TI E2E™ Community. Our goal is to make the transition from EVM development to your custom board as seamless as possible by re-targeting only the platform-specific code and maintaining the general code framework and structure.

So, if you want to get your hands on a very powerful and flexible evaluation module for TI’s Keystone family, now is the time to act!

Imagine the impact…TI’s KeyStone SoC + HP Moonshot

Sanjay35057 — Fri, 19 Apr 2013 14:28:40 GMT

Last week, market leader Hewlett Packard announced a huge change in the server landscape with its recent Moonshot announcement. HP responded to the new challenges being driven by the diverse requirements of mobility, cloud, social and Big Data by enabling application-optimized servers rather than the traditional, one-size-fits-all servers. TI is very excited to be part of such a monumental industry shift as we strongly believe that a better way to cloud is driven by such purpose-driven server architectures.

TI’s participation in HP’s Pathfinder Innovation Ecosystem is the first step towards arming HP’s customers with optimized server systems that are ideally suited for workloads such as oil and gas exploration, Cloud Radio Access Networks (C-RAN), voice over LTE and video transcoding. This collaboration between TI and HP is a bold step forward, enabling flexible, optimized servers to bring differentiated technologies, such as TI’s DSPs, to a broader set of application providers. TI’s KeyStone II-based SoCs, which integrate fixed- and floating- point DSP cores with multiple ARM® Cortex™A-15 MPCore processors, packet and security processing, and high speed interconnect, give customers the performance, scalability and programmability needed to build software-defined servers. HP’s Moonshot system integrates storage, networking and compute cards with a flexible interconnect, allowing customers to choose the optimized ratio enabling the industry’s first software-defined server platform. Bringing TI’s KeyStone II-based SoCs into HP’s Moonshot system opens up several tantalizing possibilities for the future. Let’s look at a few examples:

Think about the number of voice conversations happening over mobile devices every day. These conversations are independent of each other, and each will need transcoding from one voice format to another as voice travels from one mobile device, through the network infrastructure and to the other mobile device. The sheer number of such conversations demand that the servers used for voice transcoding be optimized for this function. Voice is just one example. Now think about video and music, and you can imagine the vast amount of processing required. Using TI’s KeyStone II-based SoCs with DSP technology provides optimized server architecture for these applications because our SoCs are specifically tuned for signal processing workloads.

Another example can be with C-RAN. We have seen a huge push for mobile operators to move most of the mobile radio processing to the data center. There are several approaches to achieve this goal, and each has pros and cons associated with them. But one thing is certain - each approach has to do wireless symbol processing to achieve optimum 3G or 4G communications with smart mobile devices. TI’s KeyStone II-based SoCs are leading the wireless communication infrastructure market and combine key accelerators such as BCP (Bit Rate Co-Processor), VCP (Viturbi Co-Processor) and others to enable 3G/4G standards compliant for wireless processing. These key accelerators offload standard-based wireless processing from the ARM and/or DSP cores, freeing the cores for value-added processing. The combination of ARM/DSP with these accelerators provides an optimum SoC for 3G/4G wireless processing. By combining TI’s KeyStone II-based SoC with HP’s Moonshot system, operators and network equipment providers can now build customized servers for C-RAN to achieve higher performance systems at lower cost and ultimately provide better experiences to their customers.

We are very excited to be part of HP’s Pathfinder Innovation Ecosystem because the collaboration between TI and HP engineers are opening a new world of possibilities for so many different markets. Stay tuned for more examples of the impact that TI’s KeyStone II-based SoC with HP Moonshot will create.

For more information about HP Project Moonshot, visit http://www.hp.com/go/moonshot/

TVWS: A key component to Internet Everywhere

PeterFlynn — Mon, 01 Apr 2013 13:37:51 GMT

Given the exploding demand for wireless internet access, there simply is not enough ‘clear’ spectrum which can provide the capacity required in time to meet this demand. The lack of sufficient wireless data capacity is predicted to be a constraint on the economic growth across several markets.

Think for a minute on why we return to our office desk to perform tasks which are available from our mobile devices? Most people would point to the order of magnitude greater internet speed on our land line connection over mobile hotspot connection as the number one issue; close second behind that is the lack of availability and reliability of mobile internet access solutions.

It doesn’t take long before our desire for immediate fast access exceeds our lust for mobility and we head back to the office environment.

The availability of the internet access and the capacity to do what we need is the new ‘glass ceiling’ for mobile devices. Breaking through this glass ceiling means embracing the concept of Internet Everywhere and building an infrastructure where 10Mbps is the minimal acceptable throughput for all users.

This will certainly take a combination of technologies to achieve this, including new standards to improve performance as well as range, more spectrum dedicated to wireless data services and a more cooperative model for spectrum sharing.

This is what TVWS is all about; opportunistic use of spectrum which is underutilized by licensed users; optimized protocols that reach out 30km or greater; increasing data capacity beyond 22Mbps/channel; providing remote backhaul in rural areas for WiFi, LTE, WiMax and other services.

The technology and standards exist today to build this infrastructure. And best of all, we don’t need to wait for the next spectrum auctions to get started!

Multicore Navigator Tips ‘n Tricks, pt. 3

db_woodall — Tue, 26 Mar 2013 19:16:39 GMT

Part 1 of this blog series introduced using Multicore Navigator for synchronization purposes. This uses the queue manager by itself to perform a core-to-core sync. Part 2 discussed using Navigator to perform notification jobs, which uses the queue manager plus the descriptors that it manages. This third installment will discuss messaging, which utilizes the queue manager plus descriptors, plus the next tool in the “Swiss Army knife”, the packet DMA (pktDMA).

Using Navigator for messaging

Messaging is a form of IPC where a data buffer is actually copied from one memory location to another – or in some cases from one device to another (device to device transfers can be performed using Navigator with hyperlink). Reception of the data (or packet, in Navigator terms) is the stimulus that causes the receiver to begin consuming the buffer. This master/slave (transmitter/receiver) model is the standard mode of operation for the pktDMA.

The pktDMA can copy packets between L2, MSMC and DDR memories, and from L2 to L2 of different processors. It can also perform scatter/gather operations by linking multiple descriptors together in a single operation. All this is done in the background while the host processors are doing other things.

The pktDMA comes in two configurations. In the first, each [Tx, Rx] side of the pktDMA is connected to a peripheral, such as an FFTC or AIF2. In this case, the pktDMA is usually dedicated to sending and receiving data from the peripheral. In the second configuration, the Tx and Rx sides are connected together to perform memory to memory copies. This is called the infrastructure configuration. One thing most people don’t know is that sometimes peripheral pktDMAs can be used to do infrastructure copies!

Navigator messaging control is shown in this diagram:

Described in outline form, this is what happens:

Host software starts the transfer by performing these steps:
1. Pop a descriptor from the Tx free queue.
2. Write the buffer address or data to be transmitted into the buffer provided by the descriptor.
3. Push the descriptor into one of the pktDMA’s dedicated Tx queues.
The QM triggers the Tx side of the pktDMA to take control. It performs these steps:
1. Pop the descriptor from the Tx queue.
2. Copy the data from the buffer and pass it across an interface to the Rx side of the pktDMA.
3. Push the descriptor back to the Tx free queue (Tx recycling is automatic).
The Rx side of the pktDMA then takes over:
1. Pop a descriptor from an Rx free queue, which gives the pktDMA a buffer to write to.
2. Copy the received data into the buffer.
3. Push the descriptor to the Rx queue.
Host software on the Rx side finally performs these steps:
1. Pop the descriptor from the Rx queue.
2. Consume the packet buffer.
3. Push (recycle) the descriptor back to an Rx free queue.

To put more simply, the host software starts the transfer by pushing a descriptor to a Tx queue. The Tx side of the pktDMA copies it to the Rx side and recycles the descriptor. The Rx side then obtains a buffer by popping an Rx free queue, writes the data into it and pushes it to the Rx (destination) queue.

That’s a wrap! I hope this series on Multicore Navigator has been informative, and has shown you a few of the ways that Navigator can be used. Can you think of other uses? Practically anything involving communication or data movement can be done - a multicore token ring is just one example. If you’re ready to dive into the details, a user guide can be downloaded from www.ti.com. Just search for SPRUGR9.

The FCC proposes a lesson in sharing

PeterFlynn — Wed, 20 Mar 2013 22:57:23 GMT

In a workshop on Wednesday, March 13, wireless industry leaders met with the FCC to discuss opening up 100 MHz of government-owned spectrum in the 3550-3650 MHz band for use in small cell deployments . This band is generally used for coastal radar systems but is heavily underutilized across the country, making it an ideal candidate for sharing wireless data services solutions.

The challenge, as discussed with the FCC and representatives from Alcatel-Lucent, Nokia Systems, White Space Alliance, NTIA, President's Council of Advisors on Science and Technology, Telcordia and others, is that cellular service has always been provided in licensed bands with dedicated spectrum allocated for use. To take advantage of this spectrum, a new model for LTE services will be needed to ensure that primary licensees (the government) will be protected and that peer users (other cognitive radio users) can be given fair access.

Fortunately, we have already developed such a model for sharing spectrum in the TV bands - TV white space. This solution was demonstrated by Spectrum Bridge and Telcordia as part of the workshop, showing that the current database-driven approach for spectrum sharing in the TV bands can work equally well in the 3550MHz band.

To make use of this government-owned spectrum as seamless as possible, , policies will need to be defined for all users, including alignment on RF characteristics, such as power level and spectrum mask. With government and industry cooperation, this new spectrum could be made available within a few years.

Members of the workshop generally agreed that the technology exists to utilize this spectrum today, and all were very appreciative of the FCC for its leadership in proposing this direction and hosting the workshop.

How does embedded processing enable mission critical systems?

Hector Rivera — Tue, 05 Mar 2013 18:44:00 GMT

Our March application of the month is focused on mission critical. Mission critical applications include radar, image processing for radar, software defined radios and unmanned aerial vehicles (UAV). In this blog I will discuss how multicore ties into mission critical applications and how it all plays into embedded computing.

Today’s advanced military systems depend on embedded computing. The reasoning is because many of the systems’ functionality is implemented on software running on a single or multiple processing boards, instead of on an ASIC or hard wired electronics. As the software defined systems are the trend for mission critical applications.

Even as defense budgets in the US shrink, the embedded processing market is seeing positive trends. For example, developers of UAVs are currently rolling out new systems with the proper size, weight and power (SWaP) requirements. As the trend toward consolidation is impacting radar, imaging processing and communication capabilities of mission critical systems by allowing more functionality in the same space of the systems they replace.

There are system consolidation trends which impact the radar, imaging processing, and communications capabilities of military systems by allowing more functionality in the same space. This has caused a huge increase in data collection capacity. The data processing capability presents major system design challenges for developers of military systems that need to meet the SWaP requirements while maintaining high speed and performance.

TI’s Keystone-based multicore DSPs and SoCs are well suited to meet the military requirements for embedded processing. TI offers a portfolio of multicore DSP, DSP+ARM (A15) and ARM (A15) processors which provide one of the best performances per watt (GFLOP/W or GMAC/W) in the embedded processing market. Also, these processors have high bandwidth interfaces which are well suited for data intensive applications. TI’s multicore processors provide the designer with the flexibility of software compatibly as well as scalability. For example, TI offers a pin compatible DSP with 1, 2, 4, and 8 cores that scale across multiple applications.

I would like to hear your comments on where you think the military embedded processing market is going. Are the defense cuts going to hinder the growth of embedded systems? What are other considerations when designing such systems?

TI extends vision for a better way to cloud through Project Moonshot

Timothy Wesselman — Mon, 04 Mar 2013 16:18:04 GMT

The HP Moonshot Pathfinder Partner Program is pleased to announce our newest member, Texas Instruments (TI). The Pathfinder Partner Program is chartered with developing the Moonshot ecosystem from Silicon Partners through Solution Partners.

With the Moonshot architecture, HP has created an innovative platform architecture that enables Texas Instruments, and other Pathfinder Partners, to develop workload-optimized hardware and software solutions. Coupling TI’s new KeyStone II architecture with HP Moonshot enables large-scale, concurrent real-time processing of cloud and traditional telecommunications workloads by one integrated system optimized for high performance, power-efficient processing. We are thrilled to be collaborating with Texas Instruments to bring this unique solution to our customers.

TI recently announced its new product series based on the KeyStone II architecture that enables “a better way to cloud” for applications in high performance computing, gaming and media processing. TI’s KeyStone II System-on-Chip (SoC) family, including the K2H and K2E devices, provides industries first implementation of quad ARM Cortex-A15 MPCore processors in an infrastructure-class embedded SoC. This unique SoC creates an unmatched combination of Cortex-A15 processors, C66x DSPs, packet processing and Ethernet switching.

Specifically, TI’s 66AK2Hx family and its integrated c66x multicore DSPs are applicable for workloads ranging from high performance computing, media processing, video conferencing, off-line image processing & analytics, video recorders (DVR/NVR), gaming, virtual desktop infrastructure and medical imaging.

While it is too soon to disclose any product plans, we are excited to welcome Texas Instruments into the Pathfinder Partner Program and look forward to ongoing technology collaboration to bring the KeyStone II SoC family into Moonshot platforms.

Tim Wesselman

Sr Director, Ecosystem Strategy, HyperScale Business Unit, Hewlett-Packard

Five for Friday – Hot topics to expect at Mobile World Congress

Debbie Greenstreet — Fri, 22 Feb 2013 18:33:00 GMT

The annual Mobile World Congress starts next week in Barcelona and here are five topics I would expect to be discussed! Are you attending? Let us know what answers you found, what other topics you found hot! In the meantime, check out our latest whitepaper on KeyStone software.

WiFi

Where in the world is WiFi? The demand for wifi access on the MWC conference venue will be high; to what extent will wireless operators deploy it?

BYOD

Bring Your Own Device - Not unlike ‘bring your own drink’, how will the limit of a single personal preference compare to the temptation for a wider selection?

Small cell demos

Is my small cell better than your small cell? How many small cell demonstrations can you count on the show floor?

Spectrum

Where’s Waldo? How will the market solve the need for bandwidth; spectrum re-farming, carrier aggregation, new licensing, all of the above?

The Cloud

How mobile is The cloud?

Wi-Fi , Wi-Max, Wi-FAR, why now?

PeterFlynn — Wed, 20 Feb 2013 16:49:10 GMT

Want to discuss Wi-Fi? We’ve got you covered. The next generation of “Wi-Fi” is all over the news and the ‘alphabet soup’ of 801.11 standards can be a bit dizzying. In the past few weeks, there have been articles published by the Washington Post, ZDNet, Forbes , Venture Beat and many others discussing the future of Wi-Fi. These articles also referenced Wi-Fi differently, including new names like Super Wi-Fi, Wi-Far, TV white space and cognitive radio. With so much different terminology and recent interest in Wi-Fi, I felt like this was a perfect time for me to share my thoughts on the industry, as well as some of my experiences from the Super Wi-Fi Summit I attended earlier this month in Miami.

At the Summit, I spoke on three different panels with other industry leaders surrounding the hot topic of Wi-Fi. The first panel focused on the wide-range of views and significant hurdles facing the FCC in order to meet the demands of Congress to maximize revenue from underutilized spectrum. During the second panel, we had an open and frank discussion on the importance of improving receivers and the economic impact to do so. And finally, the third panel helped to underscore the complex relationships between the different spectrum users and interrelated applications. Three over-all insights emerged from all of these panels which beg further discussion:

The huge demand for wireless data capacity likely cannot wait the four to seven years it will take to complete the Incentive Auctions, much less the additional three to five years for deployment after that!

Finding more spectrum for unlicensed use is a bit like drilling for oil; it’s a limited resource, but you know there is plenty of it out there. So it will take better ‘drills’, greater investments and more concern for the existing ‘environment’ to get to it. We just have to “Drill Baby Drill” for that spectrum!

All industries and government agencies need to develop a broader view of “their” spectrum. It can no longer be about locking down spectrum as a protection strategy; it needs to move toward finding the right technology to maximize spectrum utilization without compromising existing uses. This will often mean ‘sharing’ things nicely, just like mommy taught us…. ;-)

I plan to do a follow-up post so please let me know which Wi-Fi insight you would you like to see discussed in my next blog post.

In the meantime, if you want more information on Wi-Fi, you can read all about it in my new white paper which goes into details on the realities and potentials of this space.

Share your love of MULTICORE – and win BIG!

Lauren Reed1 — Thu, 07 Feb 2013 14:52:00 GMT

We have so much love for our Multicore Mix readers that we are celebrating YOUR love of multicore with a contest in the spirit of Valentines Day!

To enter the contest, leave a comment on this post and tell us what you LOVE about multicore. It’s as simple as that! Our Multicore Mix bloggers will select their favorite comments and we will feature them in a blog post. The #1 FAVORITE comment will win a $100 Visa gift card AND a box of chocolates! You have until midnight CST on Wednesday, Feb. 13 to enter. We will announce the lucky winner on Valentines Day, Feb. 14!

And before you start commenting on your multicore love, read what some of our Multicore Mix bloggers love about multicore!

Mark Nadeski said, “What I love about multicore is the ability to add high performance processing at the edge and allow the creation of new world-changing products. An example is the high definition smart surveillance camera that will help protect people and their possessions.”

Zhihong Lin loves multicore because, “Multicore processors are the most sophisticated chips one can ever build, it is like building a space ship, and you get a sense of pride to be part of the creation process.”

Debbie Greenstreet said, “Well, I LOVE working with my TI multicore counterparts. Together we worked on our most recent device, the 66AK2H12, which has 2.7 BILLION transistors, is a mere 215 mm² and the first samples work!”

CONTEST DETAILS: This contest is open to all residents of the United States. TI employees and their family members cannot enter the contest. Prizes will be shipped via Fedex Ground and cannot be shipped to P.O. Box addresses. Entries will be judged based on what our Multicore Mix bloggers deem is the most creative comment on why they love multicore. The selected winner must respond to the email notification within 72 hours or they will forfeit the $100 Visa gift card and box of chocolates. If the selected winner does not respond within 72 hours, an alternate winner will be chosen based on originality and content of response by Multicore Mix bloggers.

The Week of the Ws: White Space, Wi-Fi and Weather!

Lauren Reed1 — Wed, 30 Jan 2013 14:50:00 GMT

Although it's rainy and cloudy in a usually Sunny Miami, there is something to be excited about! Need a hint? What’s new, exciting and at a convention center near you this week? Why it’s the Super Wi-Fi Summit, where all things White Space are to be discussed, demonstrated and explored.

Our very own multicore expert, Peter Flynn will be participating in THREE sessions during the Super WiFi Summit. Wednesday, January 31 he will be a part of the panel: Incentive Auction: Post Comment -- Persepctives - White Spaces at 12:30 p.m. EST. On Friday, February 1. He will be speaking in the Receivers Standards- Out of Band Emissions session from 9:00 - 9:45 a.m. EST in room B122. Right after, Peter will be speaking at the session, Small Cells: Filling the Void with Spectrum Sharing and Priority Use, from 10:00 – 10:45 a.m. EST in room B122. You can find more information on these sessions here.

If you want to learn more about TI’s multicore solutions for the TV white space market, visit us in the main demo pavilion on the demonstration floor. We will be there with the White Space Alliance.

In addition to our activities at the Summit, we have a new white paper focused on what TV white space is and how to succeed in the market. We also have a new Ask the Experts video where Peter answers the question: What exactly is TV white space and how are TI’s multicore processors helping make it a reality?

Today marks the first day of the Summit, so if you’re attending, please stop by and visit us!

COWs come to Washington D.C. for the Inauguration

Debbie Greenstreet — Thu, 24 Jan 2013 20:17:00 GMT

The donkey and the elephant are usually the animal icons that one typically associates with the politics centered in Washington, D.C., so why is this story about COWs? Well, there was that incident several years back when a disgruntled farmer parked his tractor on the National Mall but he didn’t bring any animals with him.

Actually I am talking about Cells on Wheels (COWs), as highlighted in a recent Washington Post article on preparation for the presidential inauguration. As you might imagine, wireless communications were severely challenged at the 2009 inauguration with 1.8 million people in several square miles. The article points out that while this year’s attendance was anticipated to be approximately half that of the previous inauguration, this time around, attendees would be armed with much more powerful data eschewing devices. As such, an estimated nine temporary cell towers, COWs, were planned versus only several in use in 2009.

I always find it enjoyable to see our multicore technology in action! In this case, our multicore technology is in some sort of wireless base station so I couldn’t resist the chance to head over to the National Mall to check these COWs out! Approaching the city by car, these COWs were not readily visible, even with the leafless trees. On foot however, I quickly spotted three, approximately 160 meters apart from each other, in line along the edge of the Mall, each close to a National landmark, and each about 50 feet in height.

The Verizon unit was not cloaked and was the only one out of the three I spotted readily identifiable. While most of the system was indeed on wheels, a truck and flatbed, the other two appeared to be on platforms, along with adjacent temporary vans. My first reaction was that of indeed a cow plodding across a field; these COWs may be “mobile” but they certainly didn’t appear to be able to move very quickly. They weren’t running on this Saturday before the inauguration, despite the large crowds already present for the pre-event activities. Perhaps the generator capacity was sized for 24 hours?

Well it was fun to witness, even if I did get funny looks taking pictures that were NOT aimed at the monument and memorials! I wonder though how the small cells and heterogeneous networks may change the landscape in four years from now. Will network planning be such that large capacity demands are served with the systems in place? Will smaller, lower power units be commonplace?

I would love to hear from anyone who has had some interesting observations or experiences with COWs or otherwise temporary cell installations, perhaps at the Olympics, or the upcoming Super Bowl!

Video killed the radio star...

Arnon Friedmann — Mon, 21 Jan 2013 23:59:51 GMT

… and now it's coming for the cloud, too! There is a crazy statistic that over 72 hours of video are uploaded to YouTube every minute. We all know that video traffic is dominating all forms of network transmission throughout the world - but what exactly does this have to do with multicore? Turns out it has everything to do with multicore - every piece of video that gets sent, received, edited, etc. needs to be processed. At a minimum, this means compression and decompression but it can also include many other processing steps such as noise reduction, analytics, and transcoding. Layered on top of this is a multitude of coding formats like MPEG-2, MPEG-4, DivX, H.264, X.264 and HEVC, which support a multitude of formats like Standard Definition, 720p High Definition, 1080p High Definition in combination with frame rates like 24, 25, 30, 60 and higher. Put this together and you get a system that needs to be very flexible yet still provide enough processing power to manage the massive amount of video traffic in today’s networks. So, where can you find a device that provides this required flexibility with high performance? Well here at Texas Instruments of course!

This month we’re highlighting Video Infrastructure as our application of the month. We encourage you to take a look at how our multicore DSP and SoC devices are helping solve some of the challenges facing the video market today.