[wp_tech_share]

We just published our 3Q20 reports where we looked at how the market performed during the three-month period from July to September.

During the third quarter, revenue was up 1% Y/Y, in contrast with our prior forecast of a 7% decline. The upside surprise was driven mostly by the non-Cloud segment (which includes Telco Service Providers (SPs) Large Enterprises, and Rest of the Enterprises). In the meantime, the Cloud segment was flat, in line with our expectations but showing uneven performance across various Tier 1 and Tier 2/3 Cloud SPs.

The recovery in the non-Cloud segment was propelled by increased demand from large enterprises (which include Fortune 2000 companies and comprise about half of the sales of the non-Cloud segment). We have been predicting this segment’s return to growth, as it is more economical for these enterprises with large scale to build and manage their own private data centers than to use the Public Cloud. However, to our surprise, the recovery occurred ahead of our predictions, as a result of improvement in macroeconomic conditions, easing of lookdown measures, and injection of government funding—which helped certain verticals, mainly in the public sector.

Additionally, we saw a recovery in the Telco SPs segment in the third quarter, driven by some 5G related projects, most notably in APAC regions. We also saw some growth in Telco SPs in Europe and North America (NA) as vendors such as Arista were finally able to add features to their products to tie back to legacy networks. Nevertheless, we expect demand from Telco SPs to remain lumpy. As for the Rest of Enterprise segment, we continued to see a decline, which is a trend that we do not expect to reverse due to ongoing migration to the Public Cloud.

Ongoing improvement expected for the remainder of the 2020 year

Looking at 4Q20, we expect the underlying growth drivers that helped the 3Q20 recovery in the Large Enterprise segment, will continue through 4Q20. However, we may have underestimated the benefit from backlog fulfillment that may have artificially boosted the performance of the Large Enterprise segment in 3Q20. Some of the manufacturers we interviewed confirmed benefiting from backlog fulfillment during the third quarter as supply challenges improved. Those benefits from backlog fulfillment may not repeat themselves to serve as a tailwind in 4Q20. If that’s the case, our 4Q20 forecast may be too high.

As for the Cloud segment, we expect to continue to see mixed performance among Tier 1 and Tier 2/3 Cloud SPs in 4Q20. In line with the 3Q20 performance, we expect network spending from some of the U.S.-based Tier 1 (for instance Facebook and Microsoft) as well as some Tier 2/3 Cloud SPs to remain soft through the remainder of the year, for reasons explained in more details in our report. We expect some of that softness to be offset by ongoing strength in spending from Chinese Cloud SPs, stimulated by an accelerated pace of digital transformation and untapped potential in the Chinese Cloud market.

For the full year of 2020, we now predict the market to decline only low-single-digit, compared to our prior forecast of mid-single-digit decline. We expect the Cloud segment to grow at a low single-digit rate, in line with our prior predictions. However, we adjusted our forecast for the non-Cloud segment from high single-digit to mid-single-digit decline, as the anticipated recovery in 2H2020 will offset some of the weakness in the first half of the year (The non-Cloud segment was down 8% Y/Y in 1H2020).

Data Center switch revenue forecast to grow high single-digit in 2021

Looking at 2021, we expect sales of data center switches to grow high single-digit. We expect the ongoing recovery in the non-Cloud segment, most of which is driven by large enterprises, will be propelled by improvement in macro-economic conditions. Recent GDP reports indicate that economists at world-leading banks are forecasting positive GDP growth in 2021, compared to negative growth in 2020. Additionally, the availability of a vaccine will boost business confidence, all helping to abate the uncertainty in the market. As for the Cloud segment, we predict a double-digit growth in network spending which will be triggered by an accelerated pace of 200/400 Gbps adoption outside of just Google and Amazon.

200/400 Gbps refresh cycle expected to accelerate in 2021

Most of the 400 Gbps adoption in the market has been so far driven by Google and Amazon. The availability of high-volume low-cost 400 Gbps optics has been the major constraint for the adoption. Additionally, adequate switch chips with the right density, power, and buffer size were needed for specific use cases at some of these large Cloud SPs to enable this transition. According to our interviews with the major chip suppliers as well as the optical transceiver vendors, we predict the 200/400 Gbps refresh cycle to start to materialize at Facebook in early 2021 and at Microsoft in the second half of 2021.

Despite our optimism, a lot of uncertainty remains in the market

Despite our optimistic outlook, we wouldn’t be surprised if the market continues to show some volatility next year. The level of uncertainty in the market remains relatively high due to elevated levels of unemployment rates and bankruptcy, which may impact business confidence and suppress network spending from large enterprises and even some Tier 2/3 Cloud SPs.

Additionally, it bears mentioning that at the time of this writing, the rate of COVID-19 infections is accelerating, prompting many countries to slow business reopening, while invoking curfews and lockdowns. This may potentially suppress market growth but we believe that the impact will not be as significant as what the market has experienced with the first wave of infections and lockdowns. We believe that many vendors, as well as their customers, have figured out ways of doing business during the unusual circumstances in these unprecedented times.

All that said, although we believe the worst of the pandemic is behind us, we are all watching how fast and smooth the distribution of the vaccine will be. Until then, and until the vast majority of the population is vaccinated and builds immunity to the virus, a lot remains uncertain about the outcome of this pandemic.

[wp_tech_share]

 

As we arrive at the end of 2020, it’s a good time to look at the forthcoming year. We assume that 2021 won’t surprise like this year, and we’ll be able to start returning to pre-pandemic routines. On a global scale, most things will return to the way they were, but some may not. The same is true for the network security market, where we predict two new trends and a return of a historical one:

1 – From Work at Home to Work Anywhere Securely (New)

Before the pandemic, few business leaders imagined running their entire business with a remote workforce, let alone plan or invest in one. Fortunately, the bottom didn’t fall out when their employees had to stay home to help stem the spread of COVID-19. Many businesses found that productivity didn’t suffer, and in some cases, improved. While some employees may never adjust to a life where the line between personal and work life is fuzzy, for many it has been empowering change, and they have no desire to go back into an office.

Enterprises are taking notice. Progressive ones, like Nationwide Insurance and Twitter, have already announced that most, if not all, employees never need to come back. Most businesses now see that remote work isn’t a liability but an asset.

In 2020, many enterprises were forced to cobble together a work from home solution. These solutions weren’t meant to be permanent. In 2021, we see enterprises making “work anywhere” a formal business goal and begin making long-term changes and investments.

Security vendors have already been responding with new cloud-based solutions that bring together networking and security in what is being billed as SASE (Secure Access Service Edge) solutions. Early adoption of SASE has started, and 2021 may be the first year it becomes mainstream.

2 – Cloud-Centric Security Will Continue to Grow Faster than Overall Market (New)

During our 2020 interviews with state VARs, we heard how digital, multi-cloud, and mobile user-friendly enterprises weathered the pandemic better than those that were not. Because of the pandemic, enterprises that were already investing in IT transformation accelerated, while those that had not started finally began their journey. The numbers we track bear this out.

We track the network security market not just by product segments but also by form factor. We follow three form factors, Physical Appliances, Virtual Appliances, and Software-as-a-Service (SaaS). We consider Virtual Appliances and SaaS to be the two form factors to be most cloud-friendly and at the heart of cloud-centric security deployments.

While we expect the overall network security market to return to high single-digit growth in 2021, we expect both Virtual Appliance- and SaaS-based network security to grow north of 20 % in 2021, or more than twice as fast as the overall market.

3 – Firewall Revenue Will Rebound (Historical)

Historically, Firewall revenue has been growing on average nine to ten % every year. In 2020, Firewalls are on a path to squeeze out just 2 to 3 % growth. Considering the Firewall segment is a $9 B market and nearly four times bigger than any other security segment we track, any sluggishness quickly impacts overall market numbers.

Most firewall sales – in particular, physical firewalls, which account for 95% of the Firewall segment by revenue – are part of a broader network infrastructure purchase together with network switches and routers. Early on in the course of the pandemic, many enterprises halted network expansion projects, which naturally led to the deceleration in Firewalls we saw in 2020.

In 2021, we expect the Firewall segment to rebound and return closer to historical growth rates for the following two reasons:

  • Need for greater capacity – 5G networks and increasing popularity of 100G/400GbE over 10G/40GbE or 25G/100GbE with cloud and telco service providers drive the need for faster, more scalable Firewalls. High-End Firewalls will stand to benefit.
  • New features and use cases – Over time, Firewall vendors have proven adept in maintaining customer relevancy with support for new features and use cases. We see this continuing in 2021 with an expansion of Low End and Midrange Firewalls into SASE.

Across all three predictions, we hold a generally optimistic view towards 2021. We feel our bias is well warranted considering the high priority security investments have held in enterprise budgets due to the threat landscape remaining as thorny as ever.

[wp_tech_share]

We just published our 3Q20 reports where we looked at how the market performed during the three-month period from July to September.

Surprisingly, the campus switch market performed well above our predictions in 3Q20. Revenue was up 3% Y/Y, compared to our prior forecast of a 12% decline, which would have been in line with prior recessions and in correlation with prior GDP forecasts.

The upside surprise was driven mainly by a stronger performance in North America as a result of strong government funding (CARES Act and E-Rate). Some of the manufacturers we interviewed expect government funding, mostly related to E-Rate, to flow into the next couple of quarters.

For 4Q20, we predict the market to grow at 3% Y/Y in 4Q20, in line with the 3Q20 performance.

 

 

We believe that the underlying growth drivers that helped the 3Q20 recovery will continue through the remainder of the year. We believe ongoing government funding, as well as improving macroeconomic conditions—namely in the U.S. and APAC—will encourage the large and even some mid-size enterprises in those regions to resume spending on infrastructure.

We would like to caution, however, that we may have underestimated the benefit from backlog fulfillment that may have artificially boosted the market performance in 3Q20. Some of the manufacturers we interviewed confirmed benefiting from backlog fulfillment during the third quarter as supply challenges improved. Those benefits from backlog fulfillment may not repeat themselves to serve as a tailwind in 4Q20. If that’s the case, our 4Q20 forecast may be too high.

For the full year of 2020, we currently predict the market to decline only 2%, as the anticipated recovery in 2H2020 will offset some of the weakness in the first half of the year (1H2020 was down 8% Y/Y).

 

Campus switch revenue forecast to return to growth in 2021

Looking at 2021, we predict the positive momentum to continue and drive 2% growth in the market.  This projected growth will be propelled by improvement in macro-economic conditions as recent GDP reports indicate that economists at world-leading banks are forecasting positive GDP growth in 2021, compared to negative growth in 2020. Additionally, the availability of a vaccine will help abate the uncertainty in the market and boost business confidence.

 

This pandemic-induced recession looking better than the two prior ones

The question now is how this projected performance in the market compares with the performance during the two prior recessions (in 2000 and 2008). In the 2000 recession, sales of enterprise switches declined in the mid-teens in the first year of the recession and it took the market three years before returning to growth. In the 2008 recession, the market declined in the mid-teens in 2009 as well, but returned to growth in the following year, thanks to deep government stimulus.  We believe a few major tailwinds would help the market perform better during this pandemic-induced recession.

 

 

First, government spending and stimulus helped not only the government sector but also the lower-education vertical and even manufacturing as governments in some regions of the world, such as Asia Pacific, provided incentives to manufacturers to modernize their processes.

Second, China, which recovered relatively fast and has been providing some uplift to the market since 2Q20, is now a significant portion of the total sales. (China was less than 10% of the market in 2008 and prior vs. about 20% in 2019).

Third and more importantly, this pandemic has actually amplified the importance of the network and is accelerating the digital transformation and network upgrade cycles. Our interviews with end-users as well as system integrators revealed that some of the digital transformation projects have been pulled in by about one to two years. This accelerated pace of digital transformation is offsetting some of the impact from work from home and cannibalization from WLAN, which we are planning to discuss in more details in our upcoming five-year forecast report.

Despite our optimism, a lot of uncertainty remains in the market

Despite our optimistic outlook, we wouldn’t be surprised if the market continues to show some volatility next year. We believe the level of uncertainty in the market remains relatively high due to elevated levels of unemployment rate and bankruptcy. Additionally, our interviews as well as commercial real estate reports are showing decreased demand and high level of excess capacity which may potentially impact demand for network equipment. We question whether the true level of private demand is currently camouflaged by massive government funding.

Finally, it bears mentioning that at the time of this writing, the rate of infections is accelerating, prompting many countries to slow business reopening, while invoking curfews and lockdowns. This may potentially suppress market growth but we believe that the impact will not be as significant as what the market has experienced with the first wave of infections and lockdowns. We believe that many vendors, as well as their customers, have figured out ways of doing business during the unusual circumstances in these unprecedented times.

All that said, although we believe the worst of the pandemic is behind us, we are all watching how fast and smooth the distribution of the vaccine will be. Until then, and until the vast majority of the population is vaccinated, a lot remains uncertain about the outcome of this pandemic.

[wp_tech_share]

AI and Network Virtualization Drive Overall Market Changes

The quickening pace of technological innovation across a growing number of industries will drive continued growth in the semiconductor industry. In the communications and consumer electronics verticals, the global deployment of 5G mobile and fiber-based broadband networks along with the phones and other devices used to access those networks and services, will be significant drivers of new semiconductor designs. Additionally, the proliferation of AI (Artificial Intelligence) and machine learning throughout service provider, cloud hyperscaler, enterprise, and industrial networks, will also drive demand for chips with embedded processing capabilities.

From a regional perspective, the Asia-Pacific region will continue to provide the largest source of revenue for the overall semiconductor industry, as China will remain the world’s largest importer and purchaser of components. China is estimated to purchase roughly 40% of worldwide semiconductor shipments, with an estimated 80% of semiconductors used in communications and consumer electronics product designs estimated to be imported from abroad. However, the domestic Chinese semiconductor manufacturing industry is estimated to be capable of meeting at most 30% of total demand.

This large discrepancy has resulted in a massive trade and technology deficit, which the Chinese Government is attempting to balance through a combination of subsidization, private equity, and the lowering of barriers to entry for foreign participants. The primary goal of these efforts is to advance the overall semiconductor industry to increase self-reliance and reduce the uncertainty that has arisen due to ongoing trade tensions with the US and other Western countries.

In 2014, China’s State Council published the “National Integrated Circuit Industry Development Guidelines,” which proposed to set up a special industry investment fund to back domestic semiconductor startups, particularly around 14nm finFETs, memory, and packaging. The “Big Fund,” as it is called, has gone through two rounds of funding, most recently raising around $29 billion in 2019.

The coordinated efforts have resulted in some notable advances, including SMIC’s (Semiconductor Manufacturing International Corporation) capability of shipping 14nm finFETs with 7nm in R&D. This is an advance over just one year ago, when SMIC’s most advanced process was a 28nm planar technology. Additionally, China will spin up its first 28nm lithography machine in either 2021 or 2022, which will help Chinese companies manufacture advanced 28nm chips, possibly within 1-2 years. That would be a significant step forward for the domestic industry and provide a foundation for more domestic foundries to begin more advanced design and manufacturing for 14nm and 7nm-based processors.

Ramping up 28nm chip production is an important milestone for the Chinese industry, as there will remain a large market for trailing-edge chips as AI features and functionality are embedded in more consumer electronics, automobiles, robots, smart electric meters, smart traffic lights, etc. The AI chips used in these applications will require more leading-edge chipset design, as opposed to leading-edge fabrication. Thus, the short-term goal of achieving scale at 28nm is a very meaningful step in the long process of developing a more complete, domestic IC ecosystem.

SMIC is also on the verge of building out a $7.6 billion plant in Beijing that will produce 12-inch wafers with the intention of fabricating 28nm chips. This factory, along with the expected buildout of other plants, could help to solve one of the Chinese industry’s biggest hurdles to the global competition: production capacity.

Additionally, SMIC and other manufacturers are also in the process of adding both foreign and domestic technical talent with the necessary years of experience to design and manufacture high-quality chips with consistent performance at price points that are competitive. These efforts will ultimately benefit the overall industry and supply chain, though the results will take time. Currently, SMIC’s top wafer production is at 14nm, while others are at 7nm and already pursuing 5nm and 3nm processes. Though improving and evolving its production knowledge and facilities are important goals, the company must still balance being the primary supplier of chips that don’t necessarily require the latest nodes. That balance is just as important to the overall growth of the semiconductor industry in China as is the ultimate evolution to 14nm and 7nm production capabilities.

When it comes to AI chips, specifically—including GPUs (Graphics Processing Units) and FPGAs (Field Programmable Gate Arrays)—Chinese companies are still expanding their knowledge and capabilities to compete effectively in what is expected to be a massive market over the next decade. These are the chips that are the most heavily in-demand for communications networks, especially as these networks are transformed and processing capabilities are distributed to the edge of the network and away from centralized data centers and central offices. The result will be smaller platforms supporting and processing the data traffic coming from billions of connected devices.

Currently, Chinese FPGA makers and network equipment providers license cores from Western companies, such as Intel and ARM. These companies also rely on EDA (Electronic Design Automation) software from Western companies, such as Cadence. Despite recent trade tensions, Chinese firms need these partnerships to continue to deliver their products to the market. These Western vendors also depend heavily on the China market for their revenues.

Although China is investing heavily in building out its semiconductor capacity, the innovation capacity advantage enjoyed by US and Western countries means that Chinese companies will continue to need access to US and Western technology for core components, software, design, and systems integration. For Western companies, this means that new market opportunities have opened up for them, provided that concerns around intellectual property, forced technology transfer, and cybersecurity are understood and that these Western firms continue to remain ahead on the innovation curve.

The opportunities for cooperation are there but will require effort to ensure both sides have their concerns around competition and information security are acknowledged and addressed. There is no question that Chinese firms will continue to move down a path towards more self-sufficiency when it comes to the design and manufacturing of leading-edge semiconductors. The investments that have already been made and will continue to be made by both existing semiconductor companies, as well as government and private investment, will ultimately result in a more self-sufficient ecosystem in China. It will take a combination of industry maturity, trial, and error, along with a focus on mass production and scale. Given the size of the investments being made coupled with geopolitical uncertainty that is accelerating the drive towards self-sufficiency, the Chinese semiconductor ecosystem could potentially close the gap faster than expected.

[wp_tech_share]

 

Last week, Marvell announced another giant deal with the acquisition of Inphi. We had an opportunity to talk with members of Marvell’s executive team to garner insights on the acquisition, which we share in this blog.

We view this acquisition as strategic and complementary, rather than opportunistic, given Inphi’s strong positioning in electro-optics interconnect both inside data centers as well as connecting data centers. Inphi also has a strong and growing position in 5G backhaul, mid-haul, and front-haul. Its technology and addressable market complement Marvell’s storage, networking, processor, and security portfolio and accelerate its growth in Cloud and 5G infrastructure. Note that more than 70% of Inphi’s fiscal 2020 year-to-date revenue is derived from Cloud and 5G. Marvell’s executive team expects this acquisition to help expand its footprint with large Cloud Service Provider (SP) customers. Four additional networking Cloud customers with greater than $100 M in annual revenue will be created by the combined company.

While the pandemic may have slowed investments in some areas of the network, we have highlighted in our reports that spending on Cloud data center infrastructure as well as 5G networks will remain robust and may even accelerate because of the pandemic.

Our Data Center Capex five-year forecast projects spending on Cloud data center infrastructure to grow at 11% CAGR, reaching $140 B by 2024. Additionally, Dell’Oro Group’s 5-Year RAN Forecast shows that 5G NR investments are anticipated to advance five to ten fold over the next five years.

Nevertheless, although growth in Cloud SPs as well as 5G networks presents attractive opportunities for many suppliers, it also carries margin challenges, technology challenges, and the rising need for custom solutions:

Margin challenges:

It is widely known that Cloud SPs have a relentless need to lower the cost of deploying and operating their data centers. Incremental savings on each piece of purchased equipment translates into significant improvement in their return on investment (RoI) as well as other profitability metrics. This is because of the scale at which they operate data centers with hundreds of thousands of server installed base. Our interviews with players throughout the supply chain revealed how vigorously suppliers are struggling to win Cloud SPs business (viewed as strategic) without taking a margin hit. This low-margin business is driving small companies to get acquired and large companies to try to grow even larger as scale is the only way to remain competitive. There are tremendous synergies when you can sell more products to a single customer. There are also synergies when you can partner with the same manufacturers. Additionally, the development costs of some advanced—and, in some cases, custom—products required by Cloud SPs is so high that they can be offset only through scale. Marvell’s executive team expects both Marvell and Inphi to benefit from the larger scale in R&D—in particular, in process technology. Note that Marvell recently announced its 5 nm platform and started its research in 3 nm. This will help Inphi drive higher performance, lower power, and potentially more custom products.

Technology challenges:

The transition to a digital world with artificial intelligence (AI), mobility, and distributed computing at the edge has accelerated the explosion of traffic and data movement inside Cloud SP networks, across data centers and at the edge. This means that networks must become faster to satisfy the insatiable demand for bandwidth. Our data center switch five-year forecast report shows that about 30% of data center switch ports shipped by 2024 will be at 400 Gbps speeds and higher. However, as network speeds continue to increase, they will create challenges. For example, pluggable optics (currently the form factor of choice for network connectivity inside the data center) will hit density and power issues. When this occurs, the industry will be forced to adopt alternative technologies, such as co-packaged optics (CPO), with optics co-packed and integrated on the switch chip. We expect this trend to favor vertically integrated companies, at least in the early stages of adoption, due to the lack of a standardization level that would allow for a more diverse ecosystem. I mentioned this trend at the beginning of the year in my blog (Ethernet data center switch trends in 2020 and beyond) and predicted that CPO will drive numerous acquisitions, consolidations, and partnerships among switch chip vendors, switch system vendors, and optical transceiver vendors. Note that Marvell has a switch chip business, which may be bolstered by Inphi’s strong optics technology.

Rising need for custom solutions:

The data and compute intensity of modern AI and machine learning workloads is putting tremendous pressure on the performance of Cloud data centers. While computing demand continues to surge, CPU performance improvements are slowing down, as Moore’s law is reaching its limits. We expect this to drive new ways to design modern data centers to become giant compute engines with hundreds of thousands of compute nodes. This, in turn, will drive innovations to interconnect the different nodes without compromising latency and performance. As Cloud SPs will try to differentiate themselves through these architectural innovations, their need for custom—and sometimes proprietary—solutions will increase, potentially forcing their suppliers to own different pieces of technologies that they can integrate. Additionally, this differentiation will require a large scale to drive synergies, while developing custom-made solutions.

Marvell’s acquisition of Inphi was the chip sector’s second acquisition last week, following Advanced Micro Devices’ acquisition of Xilinx. The deal followed on the heels of an already active year of giant tech acquisitions with Nvidia’s acquisition of ARM and Analog Devices’ acquisition of Maxim Integrated. We expect more acquisitions to follow as the only way to survive in this highly competitive market is to join forces and get “stronger and better together.”