Growth Stocks: Arista Networks

09:24 27 June 2023

We return to analyze other leading companies in the AI sector, this time it is the turn of Arista Networks (ANET.US). Technology company specializing in high-performance network and switch solutions. Founded in 2004, it is today an important benchmark in the network infrastructure market.

The company focuses on providing high-performance data network solutions for data centers and cloud computing environments. Its main objective is to provide scalable, secure and efficient networks that can meet the demands of the most demanding customers. Arista Networks stands out for its focus on product innovation, quality and reliability.

Start investing today or test a free demo

Open account Try demo Download mobile app Download mobile app

Arista's core business is the manufacturing and sale of high-performance Ethernet switches. Its products are designed to deliver superior speeds and capacities, enabling businesses to handle large volumes of data and ensure optimal performance in high-demand environments. Arista switches are also designed to be highly programmable, making it easy to implement custom solutions and automate network tasks.

Arista Networks has a relationship with Artificial Intelligence (AI) through the implementation of intelligent and AI-enabled network solutions. The growing demand for highly scalable and performance-optimized networks in AI and machine learning environments has led Arista to develop products and services that can support these data-intensive workloads.

One of the key areas in which Arista Networks has applied AI is in network management and operation. Its Arista Cognitive Networking platform uses AI and machine learning techniques to streamline network management, automate configuration tasks, and more efficiently troubleshoot. This allows organizations to reduce the complexity of network administration and improve the performance and availability of their infrastructures.

In addition, it has integrated AI capabilities into its switching products. By leveraging machine learning algorithms, Arista switches are able to analyze traffic patterns and data in real time to adapt their behavior and improve network efficiency. This helps optimize routing, balance load, and improve network security.

Arista has also developed specific solutions for the deployment of AI and machine learning applications in data centers. Its products and services are designed to provide high-performance, low-latency connectivity, which is essential for the rapid transfer of large volumes of data between servers running AI models and specialized processing units such as GPUs.

In summary, Arista Networks has applied Artificial Intelligence in its network solutions to improve infrastructure management and operation, optimize network performance and efficiency, and provide high-performance connectivity for AI and machine learning applications. Through these implementations, Arista seeks to provide its customers with the ability to build and manage networks that support the growing demands of AI in enterprise environments.

In recent years Arista has gained significant market share in the high-speed data center switch market at the expense of Cisco (CSCO.US) in recent years. Where the growing megatrend of AI Networking and AI Spine would equip Arista with another avenue for future growth.

 

Key Growth Drivers

Gaining High-Speed Data Center Switching Market Share: In recent years, Arista has gained notable market shares in the 100-gig data center switching market and, more recently, in the key data center market. 400 gig data. Where Cisco has consistently lost market shares to Arista. The chart below illustrates the market share landscape in the 10GbE and above switch market, excluding blade switches.

Crehan Research Datacenter Switch Market Share Report: Q4'22

Arista has experienced multiple cycles in the 400 gig switch market and their exceptional execution has allowed them to maintain their position as the market leader. Arista is currently actively investing in its 800 gig switching technology. Where the introduction of the 800 gig data center ethernet switch could contribute significantly to Arista's growth over the next decade.

Dell'Oro January 2023: Long-Term Ethernet Switch Forecast

The AI Networking and AI Spine Megatrend: Arista primarily operates in two key markets: data centers and campuses. The data center switching market has seen significant growth due to increased spending on cloud computing and investments by large enterprises. Over the next decade, both cloud computing and enterprise data centers will require AI networking and AI backbone technologies, including cognitive adjacencies and cognitive networking.

In recent earnings presentations, Arista revealed that AI networks are still in their early stages. Currently, in high-performance computing cases, there is about a 50-50 split between Ethernet and InfiniBand, which are pre-AI technologies. However, Arista believes that AI will become critical and integrated into all business processes.

In terms of the AI portfolio share, it is estimated that 80-90% of the portfolio share will be dedicated to processors. If data centers and large enterprise customers are fully vertically integrated into IT, it may not present significant growth opportunities for Arista. However, if these customers choose a best-of-breed, horizontal approach, Arista will become highly relevant. Considering the complexity of AI networks, it is argued that most customers would choose horizontally integrated technology for their AI implementations.

Arista's flagship AI platform is called the 7800, and the company has done extensive simulations of how it works with clusters of GPUs and different types of NICs. Where the company believes that the next 2-3 years will be crucial to experience technological advances and achieve mass production.

Growing Software and Service: Software support and service renewals currently contribute approximately 13.5% of Arista's revenue and offer high margins. Software-as-a-Service (SaaS) offerings can deliver zero-trust networks, improve deployment designs, and improve visibility and observability. The trend is that services and software will grow at a faster rate than the rest of the hardware business, leading to a higher margin profile for Arista over time.

 

Can Arista continue to grow its 800 gig market share?

We've always wondered why big tech companies can't sustain their technology for the long haul. And it seems that the main reason lies in the software/hardware architecture. When designing a product with embedded software, the software architecture must be carefully designed and produced. This architecture determines how software can evolve in the future as other technologies advance. While features can be added and upgraded to the software, the architecture itself cannot be easily changed in the future, because of its higher cost. This explains why Oracle (ORCL.US) is losing market share to cloud database providers, why IBM (IBM.US) keeps fighting to adapt to the cloud, and why Intel CPUs (INTC.US) can't compete with GPUs.

Arista initially built its product spanning both hardware and software, with a focus on the cloud and data centers. Its 400G and 800G switches are easy to deploy in cloud environments and can be easily scaled to meet the growing demands of cloud computing. By contrast, Cisco (CSCO.US) switches dominated the traditional local world. They are expensive and complex to implement, requiring Cisco-certified technicians to maintain and design the network. While this structure has worked in the past, it is not suitable for the age of cloud and AI.

While Arista continues to expand its portfolio with 100G and 400G products, it also introduced its first 25 terabyte single rack unit product this year, with 800 gig ports that can be split into 2x400 gig. These products have significant use cases and are ideal for high speed applications such as AI. Arista's operating system, EOS, is a high-quality, resilient network data lake-based system that has matured over time and now supports cloud-scale operations with multiple copies of the Internet routing tables. In my opinion, Arista is way ahead of the game in the new age of AI.

 

Analysis of prospects and short-term risks

Weak Data Center Spending: The current macro environment is raising uncertainties regarding capital spending among large enterprises, particularly in the data center sector. Arista operates in both the hyperscale and cloud segments of the business, which can be more cyclical compared to other sectors.

During their latest earnings presentation, Arista acknowledged the potential uncertainties surrounding cloud spending in 2024, and they hope to gain more visibility on this matter later this year. However, despite these challenges, Arista remains committed to targeting 20% growth from FY20 to FY25.

Cloud giants including Microsoft (MSFT.US), Meta (META.US) and Amazon (AMZN.US) account for around 40% of Arista group sales. And there are no signs that these companies have started to reduce their capital spending. Even if there is a macroeconomic downturn in the near future, Arista's ability to continue to gain share in the 100G and 400G switching markets may allow them to escape a slower demand environment. Furthermore, a significant slowdown in cloud migration is unlikely, as it is a megatrend, and cloud workflows actually reduce expenses for businesses over time. Many companies need to invest in the cloud, AI, and data analytics to maintain a competitive advantage. Taking all of this into consideration, short-term risk is manageable for a company like Arista.

Regarding the near-term outlook, Arista supported the consensus of 26% annual growth, with the goal of reaching approximately $5.5 billion in revenue by fiscal year 23. In addition, they anticipate that the growth of the biggests of the clud sector will moderate compared to the triple-digit growth experienced in FY22.

Cisco Nexus 9800 Competition: Cisco has designed the Nexus 9800 for core/DC switching solutions, making it a direct competitor to Arista's 7800 series. The Nexus 9800 features 400G readiness and 800G capacities, along with a low-power distributed architecture. However, I think the Nexus 9800 can be quite expensive to implement and customers may find it difficult to expand its capabilities. By contrast, Arista's products are more affordable and easier to implement. Arista's Extensible software is very convenient and easy to use, allowing for a smooth start in my opinion.

In addition, Arista has a strong margin profile, with approximately 60% gross margin and 40% operating margin. As of the first quarter of fiscal year 23, they had $3.3 billion in cash on their balance sheet. These strong financials position them well to withstand any recessionary macroeconomic environment.

 

Valuation

Analyzing Arista through cash flow discount, we will try to point out the fair value of its share price. Assuming Arista's revenue growth rate will be 20% over the next four years, then moderate to 15% between FY27 and FY32. Its operating margin is estimated to expand by 0 .5% pa driven by operating leverage.

In addition, the model uses a WACC (Weighted Average Cost of Capital) of 10% and a 4% terminal growth rate (g).

Discount Cash flow projections


In the model, the present value of free cash flow to the firm (FCFF) over the next 10 years and the terminal value are $15 billion and $38 billion, respectively. As such, the total value of the company is estimated at $53 billion. Adjusting for gross debt and cash, the fair value of Arista's stock price is $187, based on this analysis.

Fair value of Arista shares, by DCF


Technical analysis

The recent behavior of the markets has left us with an important support area in the share price of Arista Networks (ANET.US). This area coincides with relative maximum levels at the end of 2021 that indicate the end of the corrections.

Source: xStation

The recent movement and upward trend in the stock since the end of 2022 indicate that the target posed by the discounted flow analysis, at $187 per share, has a high probability, as long as it exceeds the recent highs.


Conclusion

Arista projects an interesting market share growth rate. In the high-speed data center switching market, its technology is well positioned for the 800 Gbps and 1.6 Tbps switching market in the next decade. Additionally, financially, Arista is poised to benefit from the AI networking and AI Spine megatrend, and the growing mix of software and services has the potential to increase its margins. It is important to highlight that in the AI market there are still uncertainties, especially in the short term that may affect the cloud computing sector. Which however looks very consolidated for the long term. Because unlike other AI companies, Arista Networks has been in action much longer and has the world's largest AI technology companies as its main customers.

 

Darío García, EFA
XTB Spain

This content has been created by XTB S.A. This service is provided by XTB S.A., with its registered office in Warsaw, at Prosta 67, 00-838 Warsaw, Poland, entered in the register of entrepreneurs of the National Court Register (Krajowy Rejestr Sądowy) conducted by District Court for the Capital City of Warsaw, XII Commercial Division of the National Court Register under KRS number 0000217580, REGON number 015803782 and Tax Identification Number (NIP) 527-24-43-955, with the fully paid up share capital in the amount of PLN 5.869.181,75. XTB S.A. conducts brokerage activities on the basis of the license granted by Polish Securities and Exchange Commission on 8th November 2005 No. DDM-M-4021-57-1/2005 and is supervised by Polish Supervision Authority.

Back

Join over 1 Million investors from around the world

We use cookies

By clicking “Accept All”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts.

This group contains cookies that are necessary for our websites to work. They take part in functionalities like language preferences, traffic distribution or keeping user session. They cannot be disabled.

Cookie name
Description
SERVERID
userBranchSymbol Expiration date 17 October 2024
adobe_unique_id Expiration date 16 October 2025
test_cookie Expiration date 1 March 2024
SESSID Expiration date 9 September 2022
__hssc Expiration date 16 October 2024
__cf_bm Expiration date 16 October 2024
intercom-id-iojaybix Expiration date 13 July 2025
intercom-session-iojaybix Expiration date 23 October 2024
xtbCookiesSettings Expiration date 16 October 2025
xtbLanguageSettings Expiration date 16 October 2025
TS5b68a4e1027
countryIsoCode
userPreviousBranchSymbol Expiration date 16 October 2025
TS5b68a4e1027
_cfuvid
intercom-device-id-iojaybix Expiration date 13 July 2025
__cfruid
__cf_bm Expiration date 16 October 2024
__cf_bm Expiration date 16 October 2024
_cfuvid
adobe_unique_id Expiration date 16 October 2025
TS5b68a4e1027
_cfuvid
xtbCookiesSettings Expiration date 16 October 2025
SERVERID
TS5b68a4e1027
__hssc Expiration date 16 October 2024
test_cookie Expiration date 1 March 2024
intercom-id-iojaybix Expiration date 13 July 2025
intercom-session-iojaybix Expiration date 23 October 2024
intercom-device-id-iojaybix Expiration date 13 July 2025
UserMatchHistory Expiration date 31 March 2024
__cf_bm Expiration date 16 October 2024
__cf_bm Expiration date 16 October 2024
__cf_bm Expiration date 16 October 2024

We use tools that let us analyze the usage of our page. Such data lets us improve the user experience of our web service.

Cookie name
Description
_gid Expiration date 9 September 2022
_gat_UA-22576382-1 Expiration date 8 September 2022
_gat_UA-121192761-1 Expiration date 8 September 2022
_ga_CBPL72L2EC Expiration date 16 October 2026
_ga Expiration date 16 October 2026
AnalyticsSyncHistory Expiration date 8 October 2022
af_id Expiration date 31 March 2025
afUserId Expiration date 1 March 2026
af_id Expiration date 1 March 2026
AF_SYNC Expiration date 8 March 2024
__hstc Expiration date 14 April 2025
__hssrc
_vwo_uuid_v2 Expiration date 17 October 2025
_ga_TC79BEJ20L Expiration date 16 October 2026
_vwo_uuid Expiration date 16 October 2025
_vwo_ds Expiration date 15 November 2024
_vwo_sn Expiration date 16 October 2024
_vis_opt_s Expiration date 24 January 2025
_vis_opt_test_cookie
_ga Expiration date 16 October 2026
_ga_CBPL72L2EC Expiration date 16 October 2026
__hstc Expiration date 14 April 2025
__hssrc
_ga_TC79BEJ20L Expiration date 16 October 2026
af_id Expiration date 31 March 2025
afUserId Expiration date 1 March 2026
af_id Expiration date 1 March 2026
AF_SYNC Expiration date 8 March 2024
_gcl_au Expiration date 14 January 2025
AnalyticsSyncHistory Expiration date 31 March 2024
_gcl_au Expiration date 14 January 2025

This group of cookies is used to show you ads of topics that you are interested in. It also lets us monitor our marketing activities, it helps to measure the performance of our ads.

Cookie name
Description
MUID Expiration date 10 November 2025
_omappvp Expiration date 28 September 2035
_omappvs Expiration date 16 October 2024
_uetsid Expiration date 17 October 2024
_uetvid Expiration date 10 November 2025
_fbp Expiration date 14 January 2025
fr Expiration date 7 December 2022
muc_ads Expiration date 16 October 2026
lang
_ttp Expiration date 10 November 2025
_tt_enable_cookie Expiration date 10 November 2025
_ttp Expiration date 10 November 2025
hubspotutk Expiration date 14 April 2025
YSC
VISITOR_INFO1_LIVE Expiration date 14 April 2025
hubspotutk Expiration date 14 April 2025
_uetsid Expiration date 17 October 2024
_uetvid Expiration date 10 November 2025
_ttp Expiration date 10 November 2025
MUID Expiration date 10 November 2025
_fbp Expiration date 14 January 2025
_tt_enable_cookie Expiration date 10 November 2025
_ttp Expiration date 10 November 2025
li_sugr Expiration date 30 May 2024
guest_id_marketing Expiration date 16 October 2026
guest_id_ads Expiration date 16 October 2026
guest_id Expiration date 16 October 2026
MSPTC Expiration date 10 November 2025
IDE Expiration date 10 November 2025
VISITOR_PRIVACY_METADATA Expiration date 14 April 2025
guest_id_marketing Expiration date 16 October 2026
guest_id_ads Expiration date 16 October 2026
guest_id Expiration date 16 October 2026
muc_ads Expiration date 16 October 2026
MSPTC Expiration date 10 November 2025
IDE Expiration date 10 November 2025

Cookies from this group store your preferences you gave while using the site, so that they will already be here when you visit the page after some time.

Cookie name
Description
personalization_id Expiration date 16 October 2026
UserMatchHistory Expiration date 8 October 2022
bcookie Expiration date 16 October 2025
lidc Expiration date 17 October 2024
lang
bscookie Expiration date 8 September 2023
li_gc Expiration date 14 April 2025
bcookie Expiration date 16 October 2025
lidc Expiration date 17 October 2024
bscookie Expiration date 1 March 2025
li_gc Expiration date 14 April 2025
personalization_id Expiration date 16 October 2026

This page uses cookies. Cookies are files stored in your browser and are used by most websites to help personalise your web experience. For more information see our Privacy Policy You can manage cookies by clicking "Settings". If you agree to our use of cookies, click "Accept all".

Change region and language
Country of residence
Language