Vertiv Holdings Co., Ltd. (New York Stock Exchange:VRT) specializes in the design, manufacture, and service of critical digital infrastructure technologies and lifecycle services for data centers, communications networks, and commercial and industrial environments.
The company's products include power management products, thermal management solutions, and integrated racks. The systems, software, and services that ensure the continuity and efficiency of the digital ecosystem, as shown in Figure 1.
This article's analysis shows that VRT is a powerful AI strategy and investment that believes some AI stocks, particularly Nvidia (NVDA), are overbought or at risk of spectacular stock declines. It shows that it is an alternative for home and traders. Soaring inventory prices.
I focus on the significant growth potential of VRT thermal management technology in data centers, a key segment of the fast-growing artificial intelligence semiconductor market.
chart 1
data center
As background and data, The center will serve as a dedicated physical facility to house computing machines and associated hardware components. Within its scope, you can find the essential infrastructure needed for your IT systems, such as servers, data storage devices, and networking equipment. Essentially, it serves as a central location where a company's digital data is stored and managed.
Hyperscalers are organizations that operate large-scale data centers and cloud infrastructure to support business operations. Table 1 shows the world's top hyperscalers. According to a report from The Information Network, these are some of the “Who's Who” companies that integrate data centers with his Nvidia artificial intelligence (“AI”), traditional AI, and generative AI. Featured ICs: Artificial Intelligence (“AI”), 5G, Automotive, and Memory Chip Market Analysis.
Why thermal management is necessary
The need for thermal management in data centers is highlighted by certain data-driven arguments that emphasize the importance of maintaining optimal operating conditions. Here's a more detailed breakdown:
- Hardware failure rates: Industry research data consistently shows a direct correlation between temperature and hardware failure rates. For example, research conducted by leading hardware manufacturers and data center operators shows that the failure rate of computing equipment doubles for every 10 degrees Celsius above the recommended threshold. This statistical relationship highlights the importance of precise thermal management to minimize the risk of costly hardware failures.
- Energy consumption and efficiency: Data center cooling systems account for a significant portion of overall energy consumption. Research to analyze energy use efficiency (You) Indicators across a variety of facilities show that inefficient cooling methods can lead to significant increases in energy consumption. By implementing advanced cooling technologies such as containment systems, liquid cooling solutions, and airflow optimization strategies, data centers can achieve significant reductions in energy usage while maintaining or increasing operational efficiency.
- Financial Impact: The financial impact of improper thermal management can be significant. Data center downtime caused by overheating incidents can result in significant revenue loss for businesses, with estimates suggesting that each hour of downtime costs thousands to millions of dollars. . Additionally, inefficient cooling systems drive up operating costs, leading to a higher total cost of ownership for data center operators. Detailed financial analysis based on real case studies provides compelling evidence that significant cost savings can be achieved through optimized thermal management practices.
As reported at the company's 2023 Investor Conference, Vertiv is the market leader in data center thermal management. According to Vertiv,
“Organizations are increasingly deploying cloud services and using artificial intelligence (“AI”) to power advanced analytics and automated decision-making, as well as blockchain and cryptocurrency applications.”
As shown in graph 2.
chart 2
AI infrastructure data centers require more power and liquid cooling
While general-purpose data centers serve a variety of computing tasks and applications, AI infrastructure data centers are specifically tailored for AI and ML workloads, leveraging specialized hardware accelerators such as GPUs and TPUs. I am.
The proliferation of AI will increase the demand for computing power, storage, and network bandwidth, driving data center growth. As AI applications expand, data centers must scale to accommodate increased data volumes. Requires specialized hardware such as GPUs and TPUs, facilitating investment in infrastructure. AI-powered insights improve operational efficiency and further accelerate data center expansion.
Graph 3 shows Vertiv for data center market estimates and CAGR growth from 2023 to 2028.
chart 3
In Figure 3 above, data centers can be categorized as follows:
- Cloud: Operated by service providers such as AWS, Azure, and GCP, which provide scalable resources.
- Colocation: Rent space for your own infrastructure with shared facilities for power and connectivity.
- Enterprise: Owned and operated internally to support business needs.
- Distributed IT: A distributed approach that uses resources across multiple locations.
financial indicators
For the fourth quarter of 2023, Vertiv reported the following financial metrics:
- Fourth quarter net sales were $1.865 billion, an increase of 13% compared to the fourth quarter of 2022
- Organic orders in Q4 increased 23% from Q4 2022. Backlog at end of Q4 2023 at record high of $5.5 billion
- Fourth quarter operating income was $285 million and adjusted operating income was $330 million, an increase of 57% from the fourth quarter of 2022.
Vertiv continues to see strong and accelerating trends in the data center market, including increased demand for critical digital infrastructure to support large-scale AI deployments. Vertiv expects first-quarter sales to increase by about 5% in real terms, with the Americas in his high single digits, APAC in his mid-single digits, and EMEA relatively flat. .
Adjusted operating income was between $200 million and $220 million, and adjusted operating margin was 13.1%, an increase of 160 basis points from last year's first quarter. This was driven by price-cost tailwinds and productivity programs, partially offset by continued growth investments.
Guidance for Q1 2024
- Net sales $1,575M – $1,625M
- Organic net sales growth rate 3.5% to 6.5%
- Adjusted operating income $200 million – $220 million
- Adjusted operating profit margin 2.9% – 13.3%
- Adjusted diluted EPS $0.32 – $0.36
Full year 2024 guidance
- Net sales $7,515M – $7,655M
- Organic net sales growth rate 9.0% to 11.0%
- Adjusted operating income $1,275M – $1,325M
- Adjusted operating profit margin 16.9% – 17.3%
- Adjusted diluted EPS $2.20 – $2.26
- Adjusted free cash flow $800 million – $850 million
Tying in to the strong 2024 guidance is Vertiv's acquisition of CoolTera on December 8, 2023. Vertiv's CEO, Giordano Albertazzi, said on the fourth quarter earnings call:
“We have already begun manufacturing CoolTera CDUs at one of our Vertiv plants and expect to quadruple our production capacity by the end of the first quarter. We are implementing plans to scale our GPUs by more than 40x. We want to ensure we have the capacity to cover the most aggressive GPU growth scenarios. It means revitalizing our global manufacturing and supply chain footprint.”
This “rapid and significant capacity increase” is illustrated on the right side of Figure 4.
chart 4
Vertiv and CoolTera partnered for three years before Vertiv fully acquired the company by the end of 2023. UK-based CoolTera was founded in 2016 and designs and manufactures coolant distribution units (CDUs) and secondary fluid networks (SFNs). , and manifolds for data center liquid cooling solutions. CoolTera's products will be added to Vertiv's liquid cooling portfolio, and the acquisition is expected to increase Vertiv's capacity by 45x compared to baseline within 12 months of acquisition.
Key points for investors
This section compares VRT with NVDA, a leading AI play. These companies are complementary rather than competitors, and VRT's thermal management systems are used in data centers that incorporate his Nvidia processors.
Exhibit 5 shows a comparison of the two companies' ratings. Quant Hold has an equal rating, but indicates a Strong Buy from Wall Street analysts.
chart 5
Chart 6 shows the 1-year stock price change, and RT (orange line) actually outperformed NVDA.
chart 6
Figure 7 compares the forward P/E ratios of the two companies, with VRT slightly outperforming NVDA.
chart 7
The major difference between the two companies is their future EPS, shown in Figure 8. Here, the VRT value is 2.8 while the NVDA value is 29.2.
However, VRT reported EPS of $1.19 for the past quarter, indicating an increase of 135% for the coming year.
Chart 8
As I mentioned at the beginning of this article, I have shown that VRT is a powerful strategy in AI and an alternative for investors/traders who think some AI stocks are overbought. Masu. We also wrote this article because some investors may prefer to buy relatively cheap AI stocks to take advantage of this fast-growing technology. As of today's close (February 28th), VRT's price was $73.40 and NVDA's price was $908.64.
I rate VTR as a hold. I expect Vertiv Holdings' earnings and revenue to grow 29.5% and 9.4% annually, respectively. This compares to 15.5% and 8.2% for the semiconductor industry, respectively.
EPS is expected to grow at an annual rate of 30.1%. However, the company's EPS for the next fiscal year is 2.82, which is relatively high. VRT is also expensive based on its price-to-earnings ratio (61.4x) compared to the average of its peers (26.8x).
VRT is accelerating CoolTera's asset consolidation and aggressive production through 2024. I'm waiting for the company's next earnings from the impact of his CoolTera acquisition in 2024 guidance.