You are currently browsing the tag archive for the ‘Intel’ tag.

Prof. Amnon Shashua: From Mobileye to COVID-19

By Shlomo Maital

Prof. Amnon Shashua

    Many years ago, Hebrew University Computer Science Professor Amnon Shashua attended a computer vision conference in Europe. Automobile executives there asked him, how many cameras are needed on a car, to warn of danger? The prevailing wisdom: at least two, because we need two eyes for depth perception (through ‘triangulation’). Shashua said, no, we need just one camera. It can measure depth by comparing data at two points in time…   The executives scoffed. Shashua came home to Israel, and launched Mobileye, which saves lives through its little camera and sophisticated software. Mobileye was acquired by Intel for $15 b. Shashua continues to head it.  

     With the outbreak of COVID-19, Prof. Shashua has tackled the issue of strategy.   His claim: Mathematics has the answer. In the online magazine Medium, he and Shai Shalev-Shwartz have published their mathematical analysis of three different strategies, and they recommend one of them. The title asks the key question: “Can we contain COVID-19 without locking down?”.   Here is a summary. *

     “We present an analysis of a risk-based selective quarantine model where the population is divided into low and high-risk groups. The high-risk group is quarantined until the low-risk group achieves herd-immunity. We tackle the question of whether this model is safe, in the sense that the health system can contain the number of low-risk people that require severe ICU care (such as life support systems).

   “ One could consider three models for handling the spread of Covid-19.

*   Risk-based selective Quarantine: Divide the population into two groups, low-risk and high-risk. Quarantine the high-risk and gradually release the low-risk population to achieve a managed herd immunity of that population.

*   Containment-based selective quarantine: Find all the positive cases and put them in quarantine. This requires an estimation of the “contagious time interval” per age group, then given this time interval one could recursively isolate all the individuals at risk from a person that is carrying the virus using “contact tracing”. Another tool is predictive testing using contact-tracing to identify people with many contacts with other people and perform tests on them.

* Countrywide (or region-wide) lock-down until the spread of the virus is under control. The lock-down could take anywhere from weeks to months. This is the safest route but does not prevent a “second wave” from occurring.

     “In the event a risk-based quarantine approach would be contemplated by decision-makers, the purpose of this document is to provide decision-makers a formal and tight bounds to investigate whether the health system can cope with the number of severe cases that would reach ICU. Embedded in the reasoning is the idea of selective quarantine (based on age groups and existing pre-conditions, but could be any other criteria) where the ”high-risk” group (the one we suspect will have a high rate of severe cases) is quarantined and the other is allowed to spread the virus under certain distancing protocols. The underlying premise is that a full population-wide quarantine is not a solution in itself — it is merely a step to buy time followed by a more managed (non brute-force) approach. The managed phase underlying our thinking is to create herd immunity of the low-risk group in a controlled manner while keeping the economy going. It is all about keeping the health system in check and not overwhelming its capacity to handle severe cases. The question we ask in this document is whether we can estimate in advance, through sampling, that the number of severe cases arising from the low-risk group would not overwhelm the system?

   “…When the high-risk group is released from isolation they would be facing a largely immune population thus naturally facing a very slow spread of infection with a good chance to whither the storm until a cure or vaccine is available. In all other selective quarantine models the high and low risk are equally susceptible to be infected so that even if the health system is not overwhelmed still the mortality of the high-risk group is likely to be higher than the risk-based model.”

     This model has been proposed before by Nobel Laureate Paul Krugman (see my blog on his proposal, April 2).   Shashua serves on an advisory board in Israel, advising Health Ministry officials. I believe his ideas are being implemented, though cautiously.

     Warning: the article whose URL is given below can be dangerous to your health; it is highly mathematical.



Beyond Moore’s Law: Vacuum Tubes?

By Shlomo Maital

  vacuum tube

Vacuum tube

   Sometimes, you can innovate by going back to the future. Take, for instance, the transistor. They are getting ever smaller, and more and more of them are packed into a microprocessor.   Semiconductor companies like Intel now work in 10 to 20 nanometer dimensions (a DNA strand is about 2.5 nanometers).   Below 10 nanometers, who knows how silicon will behave?  At those dimensions, it starts to emit light and becomes very flexible.

   Almost half a century ago, Intel founder Gordon Moore stated his famous law, that the number of transistors that could be etched into silicon wafers would double every 18 months or so. This has held true, remarkably. But it seems we are approaching the limit of Moore’s Law. The smaller transistors get, the more they leak electrons. This causes wasted power (up to half the power consumed by microprocessors is lost when electrons leak), and generates a lot of heat, which in turn requires massive cooling.

     Dr. Axel Scherer, a Caltech scientist, is working on a solution. He and two students have gone back to the vacuum tube. Vacuum tubes (“valves” in Britain) are devices that control electric currents between electrodes in an evacuated container;   electrons are emitted from a hot filament or a cathode heated by the filament. They are big, clunky and creating the vacuum is costly. But Scherer creates tiny tubes of metal, able to turn flows of electrons on and off between four very tiny probes. What is neat about this is that you do not need to use silicon, and the very ‘leakage’ of electrons that bedevils tiny transistors actually is the basis of the nano-vacuum tube substitute for transistors.    It reminds me of Dov Frohman’s invention of flash memory.  He was asked to solve a problem of stray electrons on the surface of silicon microprocessors, solved it, but realized you could make use of those electrons, as a way to store information.  Flash memory is now ubiquitous. 

   Could these tiny vacuum tubes help us keep Moore’s Law in business? Stay tuned. Meanwhile, think about other old technologies that can be adapted to create massive value in new ways.

   Source:   “Shrinking computer chips, thanks to grandma’s radio tubes”. NYT, John Markoff, Monday June 6, 2016

The End of Moore’s Law?  Why It Matters

By Shlomo Maital  


In 1965 the (later) co-founder of Intel, Gordon Moore, published a scientific paper, in which he made the following claim:   “Over the history of computing hardware, the number of transistors in a dense integrated circuit doubles approximately every two years.”   This was quite amazing, because integrated circuits were very new and there was not much history of ‘computing hardware’.   And how right he proved. (See the diagram).  In 1971,  around when Intel was born, there were 2,300 transistors on a single chip.  Today?  With 10-core microprocessors,  there are some 2.6 billion!   That is almost exactly 21 doublings, in the 43 years since 1971.  How in the world did Gordon Moore know???

   This powerful exponential curve,  which presumes a 42 per cent annual compounded rate of growth (of the number of transistors on a chip), has completely changed our lives,  placing a cell phone device in our hands that has the power of a major computer.  It has placed chips into cars, refrigerators, and nearly everything.  It has greatly reduced the price of electronic devices, because the same exponential curve that expands the number of transistors per microprocessor, also lowers exponentially (negative exponential) the cost, very rapidly.

     In “Tapeout”, the magazine of the Israeli semiconductor industry, the latest issue asks whether Moore’s Law is about to “exit” (R.I.P., die, stop)?    In a sense it already has.  Intel’s 10 core microprocessor is impressive (basically 10 microprocessors in one) but the truth is, many of those ‘cores’ do not operate at any given moment.  Intel simply shifted its marketing from “hey, count the megahertz” to “hey, count the number of cores”. 

   However, if microprocessor technology truly does ‘hit the wall’, like a marathon runner, and a limit to packing transistors onto a chip is reached,  then an entire industry will be in crisis.  It will have to find a new technology, to restore the exponential growth, on which profits and revenue are built.  And shifting to an entirely new silicon technology – perhaps chips based on cell biology? —  will be expensive, risky, difficult and will lead to a major shakeout in the industry, as small disruptive startups rise to the fore to replace big lumbering established firms. 

    In “Tape-out”, the experts are divided about the alleged end of Moore’s Law.  After all, its ‘end’, demise, has been predicted now for decades.  But as the technology approaches and passes 10 nanometers, it may be true that a physical limit is being reached.  

   Watch the semiconductor industry closely.  Repeal of Moore’s Law will affect all of us, in so many ways. 

   “Moore’s law” is the observation that, over the. The observation is named after Gordon E. Moore, co-founder of the Intel Corporation, who described the trend in his 1965 paper. His prediction has proven to be accurate, in part because the law now is used in the semiconductor industry to guide long-term planning and to set targets for research and development.[4] The capabilities of many digital electronic devices are strongly linked to Moore’s law: quality-adjusted microprocessor prices,[5] memory capacity, sensors and even the number and size of pixels in digital cameras.[6] All of these are improving at roughly exponential rates as well

Too Big to Succeed? Carve it up.

By Shlomo  Maital


  Over the years, in working with big companies, I’ve learned how difficult (impossible?) it is for huge organizations to sustain creativity and innovation. In a recent magazine column, I wrote about Intel, and how a young rather junior Israeli engineer kept Intel from abandoning its CISC technology, leading to the highly successful Pentium.  This occurred only because Andy Grove, then CEO, was willing to listen to those below him.  Many CEO’s of huge MNC’s simply are not able or willing.  Creative people get lost in the swamp of organizational bureaucracies.

   A new fashion is developing to grapple with this problem.  Split huge companies, elephants, into smaller pieces, rabbits.  Like on Thanksgiving (always the 4th Thursday in November – Nov. 27, this year, in America),  big companies are being carved up like turkeys, in the hope the pieces will be tastier than the whole bird.

    eBay is divesting PayPal.   Now, HP is splitting into two. HP stock soared on the news. Shareholders are delighted.  It’s an act of creation – making something out of nothng.

    I am very doubtful.   Many industries have seen a wave of ‘consolidation’ – mergers.  A merger is when two sick companies merge, to create one really BIG sick or sicker company.  This is what happened in the airline industry.

   Now this is being reversed.  Reverse mergers.  Very very profitable for Wall St. investment banks that shepherd the process, for a huge fee.   HP is a company that lost its way, under very poor management, until Meg Whitman.   But it will not solve its problems by splitting them into small pieces.  You cannot make a company healthy by combining it with another;  nor can you make a company healthy by carving it up like a turkey.   The pieces are still turkey.

   Long ago, management educators taught that ‘structure is not strategy’.  The way you structure the pieces of a company is NOT a strategy.  Companies that seek innovation by restructuring rarely succeed.   Because the DNA, the company culture, remains. 

  Let’s wish HP success.  But I’m very skeptical.

Is Intel Making a Big Mistake?

By Shlomo  Maital   


Israel’s President Shimon Peres (left) with Intel Israel President Muli Eden and CEO Maxine Fassberg

    Israelis love Intel, for two major reasons.   One, its fab in the south of Israel is exceptionally efficient and generates billions of dollars in exports for Israel, as well as creating thousands of well-paying jobs; its R&D center in Haifa, near where I live, is a collection of exceptionally creative engineers.   Two, an Israeli, David (Dadi) Perlmutter, Intel’s Executive VP, is Israel’s most senior executive working for a large global company.  The Pentium and the Centrino were his idea.

    This is why I read with distress this morning’s Financial Times report, that 52-year-old Brian Krzanich was appointed  Intel’s next CEO, succeeding Paul Otellini.  Krzanich, an engineer, comes from the production side of Intel, rising through the ranks to manage a fab.   Otellini came from Marketing.

      Here is FT’s take on the appointment:  “His appointment will increase speculation that Intel could focus on growing this side of its business (manufacturing) to become more of a foundry for outside companies, given its lead in the miniaturization of chips….”

   A financial analyst commented:   “One of the things that Intel really needs to look at is how it can tie up with some of the key fabless chip suppliers, which are Arm-based companies like Apple and Qualcomm, in order to leverage its manufacturing strengths, even if it doesn’t get to leverage its processor design strength.”

   To me, this means Intel is giving up on innovation and focusing on its operational excellence.  I think this is a wrong decision.  History shows that those who make products – the middle of the value chain – lose profit margins to those who design the products, and those who market them.  Why else would Nike forego having its own factories, and invest solely in back-end and front-end activities? 

   Intel chips have lost the battle to ARM chips.  This doesn’t mean, however, that Intel has lost the war.  The hasty retreat is premature and does not do justice to Intel’s history of bold innovation and risk-taking.  

Blog entries written by Prof. Shlomo Maital

Shlomo Maital