Categories
Covid-19 Electronic Components Supply Chain

Electronic component supply chain efficiency. Will we see another increase in supply and demand due to COVID-19 this year?

Supply chain efficiency: Will there be another demand increase due to Covid-19 this year?

In 2020, the electronics components industry saw both increases and decreases in supply and demand depending on where you look.

For example, demand for semiconductors that enable servers, connectivity and cloud usage skyrocketed due to stay at home workforces. Meanwhile, demand for semiconductors used in the automotive industry declined as car sales fell.

In other words, the supply and demand for electronic components was different across various sectors. Now that 2020 is behind us, 2021 is looking to follow much the same path as we continue to contend with COVID-19.

However, there will be one big difference – most of the sectors that had reduced demand for components in 2020 will ramp up their purchase orders in 2021. This is the result of economies opening up and companies getting back to operations.

Supply and demand in 2021

We believe the electronic component industry will witness a significant increase in supply and demand in 2021. There are a few reasons for this. The first is that most industries hampered by the COVID-19 pandemic will open up. Car manufacturing is the big one. This will fuel a surge in demand for semiconductors and sensors.

2021 will also play host to cyclical sectors and several tailwinds. 5G, Wi-Fi 6, AI, robotics, cloud, communications, edge computing and AR / VR are the big ones. These technologies will fuel demand for new electronic components.

Supply constraints will persist

Factories will have to ramp up production to meet demand. 2019 was a bumper year for electronics and a lot of infrastructure was built to meet demand. 2020 stuck a fork in the road, placing higher demand on certain components. In 2021, demand will return to a form of previous normality, increasing supply constraints.

We expect supply constraints of components to grow in 2021. Manufacturers will struggle to get a hold of the parts they need.

This will increase the need for partnerships with electronic component distributors like us who are ingrained into the fabric of the industry.

Things will get better over time

With the global rollout of the coronavirus vaccine in place and manufacturing sectors protected from Government shutdowns in most countries, 2021 should be a year where we see supply constraints reduce over time.

Supply and demand will get back to 80% normality toward the end of 2021. 2022 should be much better. This assumes we get to grips with this horrible virus.

In the meantime, tailwinds will continue to fuel demand for electronic components in sectors like AI and edge computing. COVID-19 has only accelerated digital transformation in most sectors. This is a powerful tailwind.

Ultimately, the demand for passive and active components will increase in 2021. You can make sure you have access to the components you need by partnering with us. We specialise in the procurement and delivery of electronic components and parts for a wide variety of industries from the world’s leading manufacturers.

Categories
Electronic Components

What does 2021 hold for the electronic components industry?

What does 2021 hold for the electronic components industry?

The coronavirus pandemic hit the electronic components industry like a freight train, knocking supply and demand for six. Now that 2021 is upon us, economies are starting to open up with pinned hopes on vaccines. This could be a banana skin, but 2021 should be a calmer year overall. The world should get back to business.

2021 in a nutshell

The avenues shut down for raw materials and shipments of electronic components will begin to open back up in 2021. This will create a healthier supply and demand market than 2020. Some issues will remain. Component shortages are likely, and this is especially true of newer parts that are found in connected devices.

Semiconductors will lead demand

The semiconductor industry saw a significant increase in global chip demand in 2020 and this will only continue in 2021. Cyclicity driven by 5G and Wi-Fi 6 upgrades and tailwinds like edge computing, AI and AR / VR will fuel demand.

Who will benefit most? Our money is on Broadcom, Arm, Qualcomm, Intel, AMD, Nvidia and Skyworks with TSMC winning on the foundry side.

DRAM will follow the path of semis

Dynamic random access memory (DRAM) is as essential to connected technologies as semiconductors. 2020 saw a sharp increase in recovery from the first quarter, and 2021 will exhibit a similarly healthy supply and demand situation.

Who will benefit most? Samsung, Micron and SK Hynix, who are the first, second and third largest manufacturers of DRAM respectively.

Shutdowns will continue

The risk of shutdowns of component production because of the coronavirus will remain in 2021. This will create extended lead times and supply issues. Governments may be forced to shutdown factories in localised areas.

The good news is this will become less common as the year goes on. The pandemic’s impact on production will reduce over time.

Tailwinds will fuel more demand than cyclicity

When evaluating electronic component demand, cyclicity and tailwinds are often pitched against each other. In 2021, we expect tailwinds like AI, edge computing, robotics and VR / AR to fuel greater growth than cyclical upgrades.

This is a sign of the times. The world is getting more connected and smarter. Innovation will fuel tailwinds and create booming tech sectors.

Counterfeiters will grow more prolific

One of the sad realities of electronic components is counterfeit components. They are becoming more sophisticated. As manufacturers clamber to get stock in this year, they are at a high risk of being targeted by counterfeiters.

Companies should rigorously control purchase sources, conduct quality inspections and use a trusted distribution partner to combat these risks.

Looking to the future

In 2020, the electronic components industry handled the coronavirus pandemic in an efficient and calculated manner. Supply and demand issues hit the industry, but they were solved for the most part in good time.

2021 will be calmer for several reasons: 1) We now have a lived experience of the coronavirus and know how to manage shutdowns efficiently, and 2) There is an increased need for us to get back to work and get on with our lives.

Categories
Electronic Components Technology Transistors

The multimodal transistor (MMT) is a new design philosophy for electronics

Researchers from the University of Surrey and University of Rennes have developed a technology called the multimodal transistor (MMT), which could revolutionise electronics by simplifying circuits and increasing design freedom.

The multimodal transistor is a thin-film transistor that performs the same job as more complex circuits. The MMT sandwiches metals, insulators and semiconductors together in a package that’s considerably thinner than a normal circuit.

However, the key breakthrough with the MMT is its immunity to parasitic effects (unwanted oscillations). The MMT allows consistent, repeatable signals, increasing a transistor’s performance. This is necessary for precision circuits to function as intended and is especially useful for next-gen tech like AI and robotics.

How it works

In the image below, we can see the design of the MMT. CG1 provides the means to control the quantity of charge, while CG2 is the channel control gate. CG1 controls the current level and CG2 controls the on/off state.

This is a massive shift in transistor design because it enables far greater engineering freedom. It is a simple and elegant design, yet it is so useful. It has numerous applications in analogue computation and hardware learning.

Digital-to-analogue conversion

MOSFET transistors are one of the building blocks of modern electronics, but they are non-linear and inefficient.

In a conventional circuit, gate electrodes are used to control a transistor’s ability to pass current. The MMT works differently. Instead of using gate electrodes, it controls on/off switching independently from the amount of current that passes through. This allows the MMT to operate at a higher speed with a linear dependence between input and output. This is useful for digital-to-analogue conversion.

The breakthrough in all its glory

The MMT transforms the humble transistor into a linear device that delivers a linear dependence between input and output. It separates charge injection from conduction, a new design that achieves independent current on/off switching.

There is a profound increase in switching speed as a result of this technology, enabling engineers to develop faster electronics. Researchers estimate that the switching speed is as much as 10 times faster. Also, fewer transistors are needed, increasing the yield rate and reducing the cost to manufacture the circuit.  

Just how revolutionary the MMT will be remains to be seen. After all, this is a technology without commercialisation. It could find its way into the electronics we use on a daily basis, like our phones. The potential is for the MMT to be printable, allowing for mass production and integration into billions of electrical devices.

With devices getting smarter and digital transformation advancing at a rapid rate, the electronics industry is booming. Semiconductor foundries are at peek capacity and more electrical devices are being sold than ever. The MMT is a unique solution to a problem, and it could make manufacturing electronics cheaper.   

With this, comes a great opportunity for the MMT to replace MOSFET transistors. We can think of few other design philosophies with such wicked potential.

Categories
Electronic Components Technology

How “Chiplets” May Help the Future of Semiconductor Technology

How 'Chiplets' may help the future of Semiconductor Technology

The global demand for semiconductors is accelerating faster than a speeding bullet, with integrated device manufacturers, systems companies, and foundries like Taiwan Semiconductor Manufacturing Company making a killing.

This accelerating demand is largely fuelled by the rollout of 5G infrastructure and the increasingly connected devices we use on a daily basis. From semi-autonomous driving aids to the connected home, semiconductors power our digital lives. They are the brains of every smart electronics operation.

In the semiconductor industry, advancements come fast. Some companies have been painfully slow to react to change. Intel is a good example – they have fluffed the development of their 7nm chips and are stuck at 12nm, while AMD already has 7nm chips and is on course to deliver a 5nm chip. Nvidia is even further ahead.  

Chiplets

Chiplets are a proven (but niche) way for semiconductor developers to make semiconductors more efficient and easier to produce.

As semiconductors get more advanced, they get smaller. At a sub 10nm scale, foundries have to be spotlessly clean. This brings with it manufacturing complexities. Also, the smaller transistors get, the more likely they are to fail.

You can increase the yield of dies with small transistors by reducing overall size. But as you reduce the size of the die, you have less space for the transistors.

So, one solution is Chiplets. Chiplets are smaller functional dies that integrate multiple chiplets into a single semiconductor. By giving functions their own circuits (sub-circuits) we can remove design complexity and focus on efficiency.

Maximising yield reduces cost

Using chiplets maximises the yield of dies and reduces design complexity, which in turn reduces manufacturing cost. To give you an idea of by how much, AMD says chiplet designs can cut costs by more than half. 50%! That’s an astonishing saving and worth the effort if it also means keeping up with technological change.

(For what it’s worth, AMD uses chiplet design in its Zen 2 and Ryzen chips. The idea being that taking smaller dies and putting them together improves yield).

Intel is also a fan of chiplet design, and they have a vision for advancing it further, where instead of multiple dies, each IP has its own building block. This creates a more modular and flexible configuration. Here’s an illustration:

chiplet internal image

This is an exciting technology because the chiplets with IP/SOC are considerably smaller than the chiplets used in multiple dies. The benefit of this is you can configure the chiplets in more ways and maintain a common architecture.

Chiplets – the future, or not?

Chiplet design is already being used by AMD, and Nvidia has said they will go chiplet when it’s economically viable to do so. This means two of the three biggest CPU and GPU companies on the planet are on the chiplet train. As for Intel, they are too – but it looks like they will go their own way to build the chiplet model they want.

Clearly, chiplets are here to stay. Scaling chips with monolithic dies will always be a thing, but it gets expensive with advanced nodes. Chiplets are necessary to break up the cost and deliver the massive number of chips our connected world needs.