JB's Circuit

Regardless of system-design approach, painful tradeoffs are still needed–usually during integration.

Earlier this month, Steve Leibson shared his “prognostications from the ICCAD panel” concerning the shape of things to come for the EDA and chip design industry.

The part of this blog that caught my attention was the comments made by Patrick Groeneveld, Magma’s Chief Technologist and the General Chair for DAC 2012. Groeneveld acknowledged two paths to handling chip design complexity: partitioning and reuse. But he believed that both of the paths were evil since they introduce inefficiencies in the overall design.

Leibson disagreed; pointing out that the divide-and-conquer method was a tried and ture approach, dating back to theRoman Empire. “…it’s an approach that seems to have withstood the test of time. However, a divide-and-conquer strategy does indeed lead to suboptimal design in terms of efficient resource use. I just don’t know of any engineering discipline that avoids such inefficiencies when tackling projects of comparable complexity. Is it hubris to think that electrical engineering and chip design are somehow different?

Both Groeneveld and Leibson offer classic arguments to the age-old problem of dealing with complexity. There are no new solutions to this dilemma, only a re-shifting of unpleasant trade-offs. In a broader sense, this re-shifting can be thought of as maintaining the “Conservation of Design Pain.” I use the word “design” for brevity and rhythm. To be correct, I should have used “development” since the pain is spread across the full system/product life-cycle effects, from design through manufacturing.

This law of “pain” acknowledges the shifting of difficult decisions to different parts of the development cycle, depending upon the methodology. For example, both partitioning and reuse are useful techniques that overcome certain design complexities by increasing the design pain in other areas, namely, in integration.

Centuries of systems engineering confirm that most systems work best when they have low coupling and high cohesion between subsystems. This is a golden rule in the partitioning between (and within) hardware and software systems. Reuse follows the same rule, with the added advantage of functionally verified blocks of design.

By reducing complexity, both partitioning and reuse simplify the work of design engineers. For example, by utilizing code or hardware reuse, engineers don’t have to design everything, which affords them more time to concentrate on designing in there area of  expertise. This leads to greater specialization, which is can be good.

But it also leads to a greater need for reintegration and often increases the complexity of interfaces. This effectively shifts the “pain” from the module to the interface subsystem.

Shifting pain from one part of the development cycle is the result of dealing with complexity. If recent trends are any indication, then the integration engineers are in for a world of hurt.

Gamification is moving from social media networks to technical sites for both motivational and generational reasons. Engineers will have to play to win.

Let’s play a little game. First, start with a noun. Let’s choose the noun, “game.” Now, add the suffix “ification” to the noun. The result is a new word (no longer a noun) that seems to add something more to the original meaning of the word. Wasn’t that fun? No? Well, it was at least engaging.

Why should semiconductor intellectual property (IP) professionals care about the growth of gamification systems? The reason is that their careers may depend upon it. I’ll explain what I mean shortly, but first I need to briefly cover this emerging field, starting with its use in popular social media applications.

Gamification is the act of changing a traditional non-game activity into a game. Wikipedia offers this definition: “Gamification is the use of game design techniques and mechanics to solve problems and engage audiences.” Note the last phrase, i.e., “engage audiences.” Gamification is not about making every activity or process fun, but rather making these experiences engaging and motivating.

What are some examples of a good gamification experience? Here is a list of the recent applications:

• Foursquare: A location-based mobile platform that makes cities easier to use and more interesting to explore. Users check-in via a smarphone application of SMS, sharing their location with friends while collecting points and badges. The company claims a membership of over 10 million people worldwide.
• Farmville: A simulation social network game that involves farm management activities like plowing, planting, growing, etc. Players invite friends to be neighbors, who may then share gifts and supplies with one another. The Facebook site claims 30 million users monthly.
• EpicWin: A solo play scenario with this twist – players create personalized role-playing game (PRG) characters based on their real life to-do list. From the site: “Make being organized as much fun as gaming with EpicWin the to-do list app with an RPG setting.”

Before you shrug off gamification as a time consuming activity of questionable real-world value, consider one more example. Stack Overflow is a very popular programming question-and-answer site. It is a free site (no registration) where users ask and answer questions, gaining “karma and winning valuable flair that will appear next to (their) name.”

Mention gamification to many hardware engineers and they will envision their last massively multiplayer online (MMO) game adventure at a LANfest (or equivalent) competition.

Is there a similar site for hardware engineers? Not that I know of, but there should be. Why not have a site called “Black Box I/O,” where hardware IP designers and verification engineers could go to ask questions and get answers about integrating common types of IP?

 Black Box I/O – The place where semiconductor IP developers engage and earn valuable integration tips. Earn the respect of your core colleague with distinctive SoC badges, Facebook “Likes” and even frequent flier miles.

Some might argue against the need for such a gamification approach, noting that semiconductor IP suppliers already provide specific interface information and models. Past columns have addressed some of the shortcomings with these models, (see, “IP Characterization Moves from The Backroom,”)

If reuse is to grow as quickly as most analyst predict, then the need to integrate disparate blocks of IP will also grow. Some of reuse might even come from providers of open source cores, like OpenCores.org  Both of these trends will increase the need for web sites where developers can post questions and get answers. Gamification techniques would provide the engagement and motivation to keep such a Q&A site going with a minimum of outside support.

Before too long, some smart and enterprising professional inSilicon Valleywill apply the techniques of gamification to IP design.

I can hear my hardware colleagues now: “That’s crazy. Our designs are too complex. Besides, we are a(n) BLANK house – (ARM, Intel, etc.). We use proprietary interfaces like BLANK (AMBA, ISOF, etc) to interconnect most of our cores. If we have questions, we ask the IP supplier directly.”

It might well turn out that the first successful semiconductor IP gamification site will be launched by one of the large ecosystem leaders, like ARM, Intel or Chipestimate – rather than from a single person or small team with VC backing. The large players could easily augment their touted user group communities with gamification techniques. It will happen.

When it does, the next generation of engineers will be ready. They are already being gamified. The proliferation of the video games, social media and engagement activities on the Internet means that the next generation of engineers and managers will be easily affected by gamification techniques. Today’s kids are easily bored if not engaged. Gamification will be one way to increase productivity and pass on learning, while dealing with the increase need for stimulation and shrinking attention spans of future technologists.

(BTW: For an insightful discussion as to the dangers of such an activity – especially to non-technical business and marketing types – check out this article: “Gamification… or is that exploitification?”)

Gabe Zichermann on gamification.

Special Event Notice: Gabe Zichermann, CEO of Gamification, Inc., and author of “Gamification by Design” (O’Reilly, 2011), will speak this Thursday ( 11/17/11) at the Arlene Schnitzer Concert Hall as part of the ISEPP lecture series.

(First published on Chipestimate.com under “IP Insider.”)

Call it what you will, commoditization is not a welcomed word in the EDA, IP or electronic hardware communities.

An abstract in an EETimes subscription publication (which I don’t subscribe to) used the catchy phrase “commodity ghetto” to describe the slide of a company’s products into highly marginalized profits. Formerly, the phase was associated with traditional commodity issues, most noticeably world food shortages or imbalances.

While electronics hardware is not considered a traditional commodity, it still can suffer from commoditization.

In the EDA and electronics markets, avoiding the commodity ghetto means becoming a “platform provider.” If you are an EDA tool vendor, IP supplier or chip manufacturer, avoiding the commodity ghetto is typically accomplished through acquisition of a software company. Software acquisitions by predominantly hardware or semiconductor IP companies mean that those companies can now claim to be “systems or platform providers.”

Do these changes work? Sometimes they do, but not often. The gulf between hardware and software worlds is too wide – in terms of engineering, sales, marketing, and even leadership.

What alternative is open to companies that feel the slide toward the commoditization ghetto? Maybe a rich activist hedge-fund investor will suddenly appear on the door step? Few corporations would find this a comforting alternative.

So companies press on, trying to reinvent themselves into system houses. Those that succeed will do so only in so much as they actually become system oriented in their approach to engineering, sales and marketing. It’s not rocket science, but it isn’t easy.

Related Blogs:

• The Second Law of Technology
•  Consumers Forget the Real Cost of Commodified Technology

Alright – I admit it! Last night wasn’t spent working on story about silicon manufacturing variability below 20nm. Instead, I snuck out with my better half and attended … must I confess? A poetry reading!

But true to the season, it was a soul satisfying recitation of terrifying tales with unexpected endings. The readers, well versed in their craft, were actors from the “Bag and Baggage” company. All told, it was memorable evening made all the better with the right company and a few goblets of Renaissance wine at Orenco Station.  What better way to prepare for tonight’s festivities on the eve before Hallowmas?

Although my time is timeless on this earth,
I still recall when days of gold turned cold
and how we all believed this new year’s birth
was eve that spirits roamed; bad ones grew bold!

Grisly Reminder

“When you do things right, people won’t be sure you’ve done anything at all.” –  God Entity

Two deaths this week reinforce America’s media obsession with marketing over engineering. One death is heralded as the passing of a demigod (Steve Jobs), while the other received obscurity (Dennis Ritchie). The former was the marketing genius behind Apple success. The later was the co-inventor of Unix and, more importantly, the creator of C. Without Ritchie’s innovation and hard work, there would be no technological foundation for a company like Apple.

But such myopic coverage is nothing new in theUS. It is the norm. Unlike other countries, the actual architects of technology of rarely recognized in our country. Instead, the focus is on consumption, on the shiny surface of things.

But this is the way of things. If we, as engineers, do our job well, no one will really be sure that we did anything at all. Unlike Jobs, engineers are not demigods. But our creations allow our business leaders and market visionaries (like Jobs) to shine – at least for a little while.

Robot Bender (not shown) talks with the God Entity from Futuroma’s Godfellas episode.

By John Blyler
Will neutrinos from a supernova in the Tarantula Nebula save Einstein endangered postulate? Does all research have to be immediately profitable? Will collisions at Fermilab’s Tevatron slow the speed of declining scientific prominence in the United States?

These questions evolved from a short interview conducted by LPE with Lisa Randall, renowned physics professor at Harvard University. Her insightful responses gave clarity to some of the complex issues facing science and engineering. Next Tuesday, Randall will launch a new season of speakers for the Institute for Science, Engineering and Public Policy (ISEPP) lecture series in Portland, Ore. Her presentation will draw heavily from her latest book, “Knocking on Heaven’s Door.”

Lisa Randall (photo courtesy of Christopher Kim)

Read the full story at: http://chipdesignmag.com/lpd/blog/2011/10/06/when-a-physicist-comes-knocking/

All of these very cool technologies – showcased by Intel’s ECA partners at IDF2011 – provide a clear direction for future trends in medical, consumer and automotive electronics.

Let’s start with the cluster controller and backplane technology.

Designers that require blazingly fast backplane buses are happy to see the development of PCI Express, Generation 3 products.  The latest version of the popular interface will provide an impressive eight gigabits per second (Gbits/s) per lane and 128 Gbit/s in designs using x16 port widths. Such performance will be welcomed in the enterprise computing, storage and communications spaces.

IDT demonstrated its latest high-performance PCIe switches alongside  re-timing devices for longer distance application. Ken Curt, Sr. Product Marketing Manager of Enterprise Computing Division at IDT, gave me the one-minute demonstration tour:

“Here are our newly announced Gen3 packet-switch devices. In this demonstration (see Figure 1), we are using a Gen2 server since we can not get a Gen3 server. The Gen2 signal comes out over cable to go into our packet switch which does a rates conversion from 5Gbits to 8Gbits per sec. The 8Gbit/sec signal – 8 lanes in parallel – is sent to a Gen3 Sata-SAS controller card from LSI logic.”

“Also, we are tapping off to a LeCroy bus analyzer (not shown) to verify that we are running 8Gb/s across 8 lanes. Further, we are showing our Eye-diagram capability to see the waveform inside of our chip and optimize the signal configuration.

“For longer traces and longer cables, we also provide PCI Express Gen3 re-timing devices, which will easily extend across 30 inches of trace or backplane.”

Short, sweet and too the point. Great demo, Ken!  

Figure 1: Ken Curt from IDT demonstrations a PCI Express Gen3 (converted from Gen2) data storage application at IDF2011.

Turning from data storage and cloud computing backplane technology, let’s now take a brief look at the medical space.

Embedded and mobile software vendor Wind River introduced a new platform for medical device development at the show. The platform, built on the company’s real-time operating system (RTOS), includes a collection of embedded software development tools, networking and middleware run-time technologies, such as IPsec, SSL, IPv6 and USB.

Having a platform is great, but experiencing the end-product is even better (see Figure 2). Automated “cuff” blood pressuring measuring devices are nothing new. What is new is having such automated devices meet stringent vendor qualification summary (VQS) processes – which is part of the company’s development platform.

Equally important to accurate monitoring of ones’ blood pressure is displaying the information in a user-friendly format (see Figure 3). This is accomplished through a collection of development tools known as the Tilcon Graphics Suite. Products such as these are sure to find a place in the booming home-care market, as well as in hospitals and the like.

Figure 2: The Wind River folks have a great bedside manner.

Figure 3: Apparently, I’m a somewhat overweight woman with higher than normal blood pressure. Well, that’s good to know.

Moving on – Let’s look at the world of intelligent displays.

Emerson Networks had a great demonstration of facial recognition applications for intelligent kiosks. I believe this kiosk was running the KR8-315, a fanless embedded computer based on the Atom E640 processor running at 1.0 GHz with 1GB DDR2 and a 64GB Solid State Drive.

Figure 4: Note the “Viewer Count” and “Majority Gender” in the bottom part of the display panel. The next figure shows how these numbers are derived via facial recognition technology.

Figure 5: Facial recognition is used to determine “Viewer Count” and “Majority Gender.” That is Connie Schultejans from Emerson in the background.

Changing direction – Let’s now move to the automotive market.

Mentor Graphics is a member of the GENIVI alliance , a non-profit industry alliance for the adoption of an In-Vehicle Infotainment (IVI) reference platform. After the recent relationship cool-off with mobile phone giant Nokia, Intel has repositioned (or re-emphasized) it MeeGo operating system platform within GENIVI. (see, “ATOM Leader Leaves Intel”)

In addition to MeeGo, Mentor also offers a complete Android-platform development environment. All of these operating systems, including Mentor’s Embedded Linux, run on Atom processors – among others. A tool suite known as Inflexion is used to create the impressive user interfaces (see Figure 6).

Figure 6: Supporting the GENINI In-Vehicle Infotainment market, Mentor Graphics offers user interface development tools that operate on MeeGo, Android and Linux system running on Intel Atom processors.

Does a drop in residential power usage – thanks in part to lower power semiconductor devices – really mean that the world is becoming energy efficient or does it hide manufacturing energy costs?

 A recent report from the Electric Power Research Institute  – a nonprofit group funded by the utility industry – concluded that the growth rate for residential power would decline over the next 10 years by about 0.5 percent a year.

Many factors contribute to this decline, including the use of more efficient lighting, energy usage improvements in newer and older homes, and cost conscious consumers in a sagging economy. But one factor, which will come as no surprise to the semiconductor industry, is the continuing evolution of lower power fixed and mobile electronic devices.

Consumer electronic devices consumer less power, but at what cost?

But does the power savings of new semiconductor devices really translate to a system-wide decrease in energy consumption? To answer that question, one would need to know the total power (energy) costs that go into manufacturing all of these electronic devices. With new semiconductor foundries costing billions and billions of dollars, these manufacturing costs are staggering. If not for the sheer volume of devices sold, no consumer would be able to afford a mobile device of any kind.

But including the manufacturing cost of each device would still not be enough to determine the total power consumption. Design costs would also be needed. Most of today’s designs involve teams spread all over the world, meaning that communication costs must be included, e.g., office lighting, and desktop-server-network power requirements.

Would the decline in residential power usage cited by the EPRI report really be enough to offset the design and manufacturing power costs for today’s “low-power” devices? As we continue to move toward a global community of consumers, this is the question that should be addressed to gain a true evaluation of energy trends.

I just finished reading “The Lost Symbol,” by Dan Brown. It was about the misunderstood meaning and purpose of the Masonic Order. A similar story could be written about the engineering profession.

The ring is given in a ceremony developed with the assistance of Rudyard Kipling and known as "The Ritual of the Calling of an Engineer."

Let me share with you two recent activities that emphasis the problems facing engineering – at least in the US. The first one comes from a Portland State University panel event with Intel’s CEO Paul Otellini and Energy Secretary Steven Chu. Here is a portion of that discussion, reported by Mike Rogoway of the Oregonian:

 

 “Panelists broadly agreed Wednesday that engineering has gotten a bad rap in the U.S., lamenting with laughs that that there aren’t many TV shows that glorify the job. And they said that teaching methods need to improve, noting that large numbers of students transfer out of engineering programs.

PSU engineering dean Renjeng Su sounded a cautionary note on the forum’s goal Wednesday, warning that teaching 10,000 new engineering grads to truly innovate costs more than simply certifying that they’ve taken classes or passed a test.

Randy L. Rasmussen/The OregonianPanelists spoke to an audience of Portland business and civic leaders at Wednesday’s forum in Portland State University’s Hoffman Hall.

“I don’t think that can be done cheaply,” Su said. “There is an intense cost to be worked out.”

For his part, Otellini said cost is less of an issue than attracting and retaining engineering students. Foreign students are filling seats in domestic engineering programs, he said, because American students aren’t pursuing them.

“We have plenty of spots,” he said. “We just need to get more of the population into those spots.”

The second relevant data point in this quandry facing the engineering profession came from a conversation that I had earlier this summer with Terry Bristol, President of the Institute for Science, Engineering and Public Policy. Terry, a long time advocate for the engineering profession, called my attention to this article: “Public Understanding of Engineering: Consequences and Solutions.”  The gist of this story is that the decline of the number of engineering and technology graduates throughout the last decade must be addressed to avoid serious problems in the future. This paper called for specific changes in the way engineering is taught, as well as a serious campaign to improve public awareness.

The recent PSU talks and the paper on public understanding caused me to reconsider the plight of the engineering profession. Why have engineers lost their respect and appreciation in the public eye? The marvel of semiconductor technology created by engineers at ever decreasing cost to the consumer seems to have done our profession more harm then good.

Perhaps a massive public relations effort is needed to remind the average American that without engineering, science is of little worth. It is time for all of us to rediscover the lost “importance” of the least appreciated profession in the US.