Autonomous cars: technology for the future

In a not so distant future we may still be sharing the road but not driving our own cars. Autonomous vehicles are becoming less like something out of a science fiction novel, and more likely to become a real world phenomenon in the form of a consumer electronics product.

Various tech companies are working toward autonomous cars, and it is also a concept in which auto manufacturers are displaying interest.

For this to happen, cars need to be outfitted with a wide array of sensors and semiconductors. Luckily, the sensors are mostly available already and the technology exists, appearing in semi-autonomous vehicles that are already on the market.

For a fully autonomous vehicle, lots of technology is needed, from sensors to radars and camera systems, as well as long range cameras, laser radar, and various other sensors and trackers.

This means that your average family car will no longer just be a car — it will be a consumer electronics product. Outfitted with a complete set of sensors, these vehicles will be virtually unrecognizable from their original counterparts — this is not your great-grandfather’s car!

All of this technology is needed, however, to keep drivers, passengers, and pedestrians safe. The technology that is being developed is being heavily researched so that autonomous vehicles can be safely introduced onto the market.

Manufacturers will have to contend with a sometimes wary public, meaning the push for finely tuned sensors, semiconductors, and other technology makes sense from a commercial standpoint. To ride in an autonomous car, you have to trust that it will work just as well as it is supposed to.

Semiconductor technology is growing and changing just as quickly as the interest in autonomous vehicles is growing. From technology that makes optical isolation easier, to quartz wafers and other materials that can work with or replace some traditional substrates, the research and development going on in the industry goes hand in hand with technological advancements in other industries, like the automotive industry.

As high tech products like autonomous vehicles become commonplace in our world, do not be surprised to see other automated, connected products enter onto the market as well. The Internet of Things is intriguing to many people, both consumers and manufacturers, and automated vehicles are just one of the products that may become a consumer electronics product. Technology is always improving, and the products we use every day are improving alongside it.

Atomic Layer Deposition (ALD) can be described as a process that deposits materials at the atomic level, layer by layer, and that enables thin and conformal films on devices. Applied Materials is a new entrant to the ALD market and hopes to make headway within a competitive and fragmented environment.

Though ALD has been in use since the 1970s, it became popular at 45nm, when Intel used the process to deposit hafnium, a high-k material, for a transistor’s gate stack. High-k replaced silicon dioxide, which in turn enabled chipmakers to scale their devices, keeping the industry on Moore’s Law, which notes that the number of transistors in a dense integrated circuit has doubled approximately every two years.

While the deposition of high-k materials for logic remains an important application for ALD, this is only a small part of the ALD market. Whether the process is thermal or plasma enhanced, ASMI estimates that the total ALD tool market for semiconductor applications will reach $1.2 billion over the next three or four years, up from $600 million in 2014.

Despite the strong market competition, Applied Materials believes that there is room for more if it maintains its focus on such things as patterning, capacitor dielectrics and 3D NAND. Applied Materials is selling a spatial-based ALD tool, in addition to furnace/batch and single-wafer tools. The new spatial tool eliminates the purging process, thereby boosting ALD productivity. Other emerging ALD markets include interconnects, fin doping, ReRAMs and selective deposition.

ASM International, Lam Research, Tokyo Electron (TEL), Ultratech and others are also increasing their ALD efforts, because the applications for ALD are rapidly expanding. The industry is in a race to develop selective deposition, which combines novel chemistries with ALD and involves a process of depositing materials and films in exact places.

Other big applications for ALD are 3D NAND and multi-patterning for logic. Cutting-edge 3D NAND chips have 48 layers and as 3D NAND scales above 48 layers, the specs will become accordingly tighter and more precise. ALD, therefore, must improve in order to meet the challenge.

ALD is becoming more important for chip production and shows great potential, because it is a good methodology for achieving increasingly complicated, thin and conformal films while retaining quality.

Semiconductor sales revenue is expected to reach US$354 billion this year, a 4 percent increase over 2014.  Currency shifts, excess inventory and sluggish growth in the PC upgrade cycle have forced a downward revision from the previous quarter’s forecast of 5.4 growth.

Recent and rapid depreciation in the value of global currencies relative to the US dollar coupled with excess inventories in the semiconductor and electronics supply chains and the end of a PC upgrade cycle drove the downward revision.  Europe and Japan are predicted to show declines in large part due to the exchange rate between the US dollar, the Euro and the Yen.

The strong dollar is prompting system suppliers and system buyers to re-evaluate their strategies.  System suppliers are raising prices in select regions like Europe to keep their margins in tact along with de-featuring some products in order to maintain current price points.  System buyers are pushing out purchases in selected regions, extending product life cycles and buying down the price curve.  This maneuvering naturally leads to reduced semi-conductor growth in 2015.

Smartphones, solid-state drives (SSDs) and ultramobiles are seeing the largest semiconductor growth while the traditional PC segment experiences the biggest decline.

DRAM, from a device standpoint, continues to be a primary growth driver of the overall industry.  DRAM revenue is expected to increase 7.9 percent in 2015, down from a 32 percent increase in 2014.  Some industry predictors say that DRAM growth will fall off dramatically in 2016 and 2017 with revenues declining 20.2 percent and 8.4 percent respectively.  Optoelectronics and Analogue product sectors are expected to grow by 8.3 percent and 5.6 percent respectively.

Overall growth in the semiconductor market is predicted for 2016 and 2017, with forecasts that the market will be worth US$370 billion by 2017.  The strongest growth region is forecasted to be the Asian Pacific which is expected to be worth US$216 billion by 2016 which will represent 60 percent of the total worldwide market.

The above figures mirror those released by the World Semiconductor Trade Statistics (WSTS) Organization which forecast sales of chips to be US$347 billion this year.  But meeting these numbers after the worst sequential growth since 2009 was reported in the first quarter of this year will require a strong second-quarter. In order to achieve 4 percent annual growth after a 7 percent first quarter decline will require a strong second quarter bounce.

Today’s cyber landscape is complex and getting more complex by the day.  Challenges to networks and software are ongoing but what about hardware? So far the main threats identified have been against networks and many organization still think that putting up a firewall or installing software-based cyber solutions is sufficient.

A significant increase in the number of attacks is bad, what’s worse is the attack profile is changing. In 2014, most DDoS attacks were short-duration, high-bandwidth occurrences. In the first quarter of 2015, the typical DDoS attack was less than 10 Gbps, and lasted much longer — more than 24 hours in most cases.

Right now technology is moving faster than the means to secure it and many companies are falling behind in their abilities to stay ahead of the burgeoning threat landscape.  Progress is being made but the need for incorporating security whenever date is generated, transported or otherwise handled is challenging traditional solutions.  The need to integrate security handling in chips and applications is opening up other areas to possible compromise . . . hardware and firmware.

On the hardware side there are cryptography chips that provide the most hardened solutions.  Layered onto that is secure code and IP and security software stacks.  These provide good cybersecurity but as threats intensify and diversify, hardware becomes the logical place to introduce new embedded threats.  Relentless attacks on financial, retail, transit, telecommunications, and identity platforms are giving embedded systems traction.

The past two decades have seen the semiconductor and IP industries breakup into small, highly focused companies that individually target challenging problems.  Specialization has helped to reduce development costs and cut time to market, but that comes at the price of security.  Being sure that what gets built into semiconductors or electronics by these highly diversified chip vendors has become difficult.  Too many pieces to the supply chain make it an increasing challenge to ensure that chips do what they were designed to do and nothing else.  Diversification means that no one can say for certain where a particular part came from without removing it, grinding off the package, inserting probes, and examining it under the scanning electron microscope.

Security has to catch up with technology and the chip industry is at the cutting edge with hardware solutions, but with these solutions also come with new opportunities for embedding problems.