What Can AI Teach Us About Data Centers? Part 2: Business Considerations
Artificial intelligence can help business leaders understand the characteristics of technologies that promote workers’ ability to innovate. The data center investment use case is of particular interest because data centers have been growing in size and complexity and will continue to do so. Many workers need to leverage the power of data center technology to get their work done and meet their missions responsibly and profitably. ChatGPT can help leaders understand how data centers can contribute to business success and how to evaluate investments in data center technologies.
Part 1 of this series introduced readers to ChatGPT and how it can be used to learn more about a topic (specifically, data centers) that has many different technical, business, environmental, political, and economic considerations. The major focus of Part 1 was describing what ChatGPT is, how it came to be, how it works, and how it can be used to address technical question about data centers. Those issues included descriptions of the types of data centers, the infrastructure required to create these centers, and alternatives to using them, such as edge computing and cloud computing. The utility of data centers for high performance and quantum computing was also described at a high level.
Here in Part 2, we move to business considerations that can help data scientists, managers, and executives weigh the utility of investing in data centers. We focus on staffing, budgeting, and financial issues. By using ChatGPT, we can gain a great deal of knowledge about data centers that would otherwise take much longer and cost much more to obtain. Still, we keep in mind the need for humans to guide the ChatGPT conversation to prevent drifts or ‘hallucinations’ that might otherwise lead the conversation astray.
Business Questions About Data Centers for ChatGPT
The following questions and answers are copied nearly verbatim from my interaction with ChatGPT. I did not correct any logic, or substantive errors in this section, so you can see what to expect from this kind of interaction. I did correct several spelling and grammar mistakes without changing the meaning of any responses, and I underlined several key phrases to make its text easier to track. My questions are in bold italics, and ChatGPT’s answers follow.
How many staff and what kinds of staff are needed to operate a data center?
The number and types of staff needed to operate a data center can vary widely depending on the size and complexity of the data center, but typically a data center will require a combination of technical and non-technical staff.
1. Technical staff: Data centers require technical staff such as systems administrators, network engineers, data center technicians, and engineers, who are responsible for the day-to-day operation and maintenance of the data center’s infrastructure, including servers, storage systems, networking equipment, power and cooling systems, etc.
2. Network and Security staff: Data center operations require specialized network and security staff, who are responsible for maintaining the data center’s network infrastructure and ensuring the data center’s security, such as network administrators, network engineers, security engineers and security analysts.
3. Facility Staff: Data centers require specialized staff to maintain the facility, such as electricians, HVAC technicians, and other facility maintenance staff to ensure the proper functioning of power and cooling systems and to comply with safety regulations.
4. Management Staff: Data centers also require management staff such as a data center manager or director, who is responsible for the overall operation and management of the data center, as well as project managers and others who are responsible for the planning, budgeting and execution of data center projects.
5. Compliance and Auditing Staff: Data centers need staff that can ensure that the organization’s data management practices are compliant with relevant regulations and standards such as HIPAA, SOC2, PCI-
What is the typical training and background for someone who leads a data center?
A typical background for someone who leads a data center is in a technical field such as computer science, electrical engineering, or information technology. Many data center managers have a bachelor’s or master’s degree in one of these fields, or a related field such as mechanical engineering or industrial engineering.
In terms of training and certifications, there are several specific skills and qualifications that are useful for a data center manager to have. Some of these include:
1. Data center infrastructure and operations: Knowledge of data center infrastructure, including servers, storage systems, networking equipment, and cooling systems, as well as best practices for data center operations and management.
2. Project management: Data center managers often oversee large, complex projects, so experience with project management methodologies and tools can be valuable.
3. ITIL or IT service management: ITIL (IT Infrastructure Library) or IT service management (ITSM) frameworks provide best practices for IT service delivery, which can be applied to data centers.
4. Energy efficiency: Data centers consume a large amount of energy, so data center managers should understand energy efficiency best practices and technologies to minimize energy consumption, comply with regulations and reduce costs.
5. Certifications: There are several data center-related certifications available, such as the Certified Data Centre Professional (CDCMP), the Certified Data Center Management Professional (CDCMP), and the Data Center Facility Operations Manager (DCO)
6. Networking: Data Center managers should have a good understanding of networking technologies, protocols and infrastructure to manage the networking in the data center.
7. Cloud computing: With the increased use of cloud computing, Data center managers should have a good understanding of the cloud computing concepts and the various providers and their offerings.
Having a solid understanding of the technologies and industry practices relevant to data centers and staying up to date on industry developments through ongoing education is essential for someone leading a data center.
Can someone without much technical, computer science, or engineering training be an effective leader of a data center?
It is possible for someone without a strong technical background to be an effective leader of a data center, but it would depend on the individual’s experience, skills, and ability to learn quickly.
A data center leader should have an understanding of the technical aspects of data center infrastructure and operations, so that they can make informed decisions, understand the challenges and opportunities that arise, and effectively communicate with the technical staff. They need to be able to understand and manage the technical risks and challenges of the data center, to have a good understanding of industry standards and best practices and to have knowledge of the latest developments in the field.
Someone without a technical background will need to quickly acquire a solid understanding of the technologies and industry practices relevant to data centers and stay current on industry developments through ongoing education and have a team of technical experts to rely on for the day-to-day operations.
However, technical expertise is not the only thing that is needed to lead a data center effectively. A data center leader should also have strong management and leadership skills, including the ability to manage projects, budgets, and staff, as well as strong communication skills to work with stakeholders across the organization. They should also understand business strategy and how the data center aligns with the organization’s goals and objectives.
In short, someone with a non-technical background could be an effective data center leader, but they would need to have a good understanding of the technical aspects of data centers and have the ability to quickly learn the necessary technical details.
Who does a data center leader typically report to?
A data center leader typically reports to a senior manager within the organization’s IT department, such as a Chief Information Officer (CIO) or a Director of IT Operations. In larger organizations, the data center leader may report to a Vice President or Senior Vice President of IT Operations.
The data center leader is responsible for the day-to-day operations of the data center and ensuring that the data center meets the organization’s needs in terms of performance, availability, and security. As such, they need to have a strong understanding of the technical aspects of data center infrastructure and operations, as well as the ability to manage projects, budgets, and staff.
The Data Center leader also often collaborates with other departments, such as Network, Security, Business Continuity and Compliance, because the data center operations are critical for the overall organization operations.
The CIO or Director of IT Operations, who the data center leader reports to, is responsible for the overall IT strategy and operations of the organization and will work closely with the data center leader to ensure that the data center aligns with the organization’s goals and objectives, and that data center performance and availability meet the needs of the business.
Photo by Christina @ wocintechchat.com on Unsplash.com
How much interaction should there be between data center leaders and other types of business leaders?
Data center leaders should have a high level of interaction with other types of business leaders, as the data center is a critical part of the organization’s IT infrastructure and plays a key role in supporting the organization’s overall business goals and objectives. The interactions between data center leaders and other business leaders can take many forms, including:
1. Collaboration on business strategy and planning: Data center leaders should work closely with other business leaders to understand the organization’s overall business strategy and objectives, and to ensure that the data center infrastructure is aligned with these goals. This might involve collaboration on IT roadmaps, business continuity planning, and disaster recovery planning.
2. Communication on performance and availability: Data center leaders should regularly communicate with business leaders to provide updates on the data center’s performance and availability, and to ensure that the data center is meeting the organization’s needs. This might involve regular status reports, dashboards, and performance metrics.
3. Coordination on security and compliance: Data center leaders should work closely with other business leaders, such as the chief information security officer (CISO), to ensure that the data center is compliant with relevant regulations and standards, and to protect the organization’s data and systems from cyber threats. This might involve collaboration on incident response planning, risk management, and security audits.
4. Technical consultation: Data center leaders should be available to provide technical consultation to other business leaders on IT issues, as well as providing advice on technology decisions that may have an impact on the data center.
5. Budget and resource allocation: Data center leaders should work with other business leaders to ensure that the data center is properly budgeted for, and to make sure that resources are allocated in a way that
If a company has limited data resources, does it need a data center?
A company with limited data resources may not need a full-scale data center, but it still needs a secure and reliable infrastructure to store and manage its data. Depending on the size and complexity of the data and the company’s budget, there are several alternatives to a data center that can be considered.
1. Cloud Services: A company with limited data resources may find that cloud services are a cost-effective solution. Cloud services such as Amazon Web Services, Microsoft Azure, or Google Cloud Platform provide scalable and flexible data storage and management services that can be easily scaled up or down as needed. This can be a good option for a company that does not want to invest in building and maintaining its own data center.
2. Colocation Data Center: A company with limited data resources can use a co-location data center where they can rent space to house their servers and other data center equipment. This can be a more cost-effective option than building their own data center, and they can still benefit from the facility’s security, power and cooling systems, and other infrastructure.
3. Remote work: If a company’s data resources are limited, and the data can be stored on employee’s laptops or desktops and the company allows employees to work remotely, they may not need a data center.
It’s important to remember that no matter the size of a company’s data resources, data security and compliance should be considered. The company will need to ensure that data is properly protected and that the data management systems are compliant with any relevant regulations or standards.
It’s recommended to consult with data center professionals, IT consultants, and security experts to evaluate the organization’s data management needs and select the option that best fits the company’s budget, resources, and data management requirements.
What proportion of a company’s operating budget should be used for the data center?
The proportion of a company’s operating budget that should be used for the data center can vary widely depending on the size and complexity of the data center, as well as the organization’s specific needs and business goals. However, as a general rule of thumb, it is common for the data center to consume 2–3% of a company’s operating budget.
However, this proportion can increase or decrease depending on the company’s industry, the company’s growth plans, and the company’s IT strategy. For example, a company that has a significant online presence or relies heavily on data-driven business decisions, the proportion of the budget allocated for the data center may be higher.
Additionally, as data center’s infrastructure and technology are evolving, it is important for the company to consider continuous investment to keep up with the change, as this will help avoid issues like higher operating costs, lack of scalability, and security breaches. As a result, the proportion of the budget allocated for the data center may also change over time.
It’s also important to note that the budget allocated for the data center also should consider not only the costs of building and maintaining the data center infrastructure, but also the costs of staffing and training, compliance and security, power, cooling, and other operational costs.
Ultimately, the proportion of the budget allocated for the data center should be determined through a comprehensive analysis of the organization’s needs and goals.
Photo by Mathieu Stern on Unsplash.com
How much investment should a medium-sized company put into a data center?
The amount of investment a medium-sized company should put into a data center depends on a number of factors, including the company’s specific needs, budget, and growth plans. The investment can vary from tens of thousands to millions of dollars depending on the complexity and scale of the data center and the level of redundancy and availability required.
When considering building or upgrading a data center, it’s important to consider the costs of equipment, construction and facilities, professional services, and staffing and training. All these costs should be considered in the context of the company’s overall IT budget and long-term business plans.
Another important consideration is the return on investment (ROI) of the data center. The company should aim to make an investment that will provide a positive ROI over time by increasing availability, improving security, enabling new business capabilities, or reducing costs. The company also should consider different options such as a co-location data center, hosting in the cloud, or using a hybrid solution, all of these options have different cost implications, which may or may not make them feasible depending on the company’s budget and needs.
It’s also important to mention that the investment in a data center should not be a one-time thing but a continuous effort, as the needs of the data center will evolve over time and the company needs to keep pace with the changes. As a result, the company should also consider ongoing costs, such as maintenance, monitoring, upgrades, and replacements that will be required to ensure that the data center remains up-to-date and meets the company’s changing needs.
In general, it’s recommended to consult with data center professionals and conduct a thorough analysis of the company’s needs before deciding on how much to invest in a data center.
Does a multinational company typically have more than one data center?
Multinational companies typically have more than one data center. They might have a centralized data center that is responsible for most of the company’s data processing and storage needs, while also having one or more regional data centers that are responsible for serving the company’s local operations.
Having multiple data centers is important for multinational companies because it allows them to:
1. Improve responsiveness and reduce latency: By having data centers located closer to the company’s customers and operations, multinational companies can improve responsiveness and reduce latency, which can improve the customer experience and increase productivity.
2. Increase reliability and availability: By having multiple data centers, multinational companies can increase reliability and availability, by providing redundancy and failover capability. This is important for ensuring continuity of operations in the event of a problem with one data center.
3. Comply with local regulations: Having multiple data centers can also help multinational companies comply with local regulations, such as data sovereignty laws, which require data to be stored within the country where it was generated.
4. Improve security: Having multiple data centers can also improve security, by reducing the risk of data breaches and other security threats. It also allows the company to create different security policies for different regions and data centers.
5. Address specific needs: Multinational companies may also have different types of data centers, depending on their specific needs. For example, some data centers may be focused on providing high-performance computing and others on providing cloud services.
It’s worth noting that a multinational company’s data center strategy may also change over time, as they expand into new markets, or as new technologies or regulatory requirements emerge. It’s important for multinational companies to regularly review and update their data center strategy to ensure that it remains aligned with their business goals and needs.
What are the major differences between how small, medium, and large size companies pay for and use data centers?
The use of and payment for data centers can vary significantly depending on the size of the company. Here are some of the differences:
Small Companies:
1. Cost and Capacity: Small companies often don’t have the financial resources or the need for a full-scale, private data center, so they’re more likely to rely on shared or cloud-based data centers. These options are more cost-effective, providing the necessary computational and storage resources without the need for large initial investments or ongoing maintenance costs.
2. Services: They typically use data centers for essential services like web hosting, email hosting, data storage, and backup.
3. Management: Because of their limited resources, small businesses usually don’t have dedicated IT teams for data center management. Instead, they rely on the data center provider to handle most of the technical aspects.
Medium-Sized Companies:
1. Cost and Capacity: Medium-sized companies might have a mix of on-premise and cloud-based services. They might have the budget to invest in their own hardware to some extent but still use cloud services to supplement their needs.
2. Services: In addition to the services small companies use, medium-sized companies might also use data centers for more complex operations like high-volume data processing, analytics, and more advanced IT applications.
3. Management: Medium-sized businesses often have more dedicated IT personnel who can handle some of the data center management tasks. However, they still largely depend on the data center provider for many aspects.
Large Companies:
1. Cost and Capacity: Large companies often have the resources to invest in their own private data centers. They may own multiple data centers in different geographic locations to ensure data redundancy, business continuity, and improved network performance.
2. Services: Large corporations use data centers for a wide range of advanced services, including high-performance computing, AI and machine learning processing, big data analytics, and complex web applications.
3. Management: Large corporations usually have their own IT departments that take care of data center management. These teams can handle everything from routine maintenance to complex troubleshooting and disaster recovery planning.
Of course, these are general trends and not strict rules. The actual use of and payment for data centers will depend on various factors, including the company’s industry, its specific needs, and its long-term plans. For instance, a small tech start-up might invest more heavily in data center usage than a large but less tech-oriented company.
Photo by Tuan-nguy-n-minh on Unsplash.com
Is a make vs buy decision paradigm the right way to think about whether a data center or an alternative would be desirable?
The make vs buy decision paradigm is often used to evaluate whether a company should build and operate its own data center or instead outsource its data center needs to a third-party provider. This type of decision is often referred to as a “build-versus-buy” decision.
When making a build-versus-buy decision for a data center, a company should consider a number of factors such as:
1. Business Requirements: The company’s specific business needs and requirements should be taken into account. For example, if the company has strict compliance or security requirements, it may be better to build its own data center.
2. Cost: Building and operating a data center can be a significant investment, and it is important to consider the long-term costs, such as ongoing maintenance, power, and cooling costs, as well as the costs of staffing and training.
3. Scalability: A company should consider the scalability of its data center needs, both in terms of the potential for future growth and the potential for changes in technology.
4. Expertise: a company should also consider the expertise needed to build and operate a data center, and whether the company has the necessary skills and resources in-house or whether it would need to hire additional staff.
5. Cloud Services or co-location: An alternative to building and operating a data center is to outsource to a third-party provider such as a cloud service provider or a co-location facility, which can provide cost savings, scalability, and expertise.
It’s important to note that the decision to build or buy a data center, or to use cloud services or co-location, will depend on the specific needs and circumstances of each organization, so it’s important to evaluate the trade-offs carefully before making a decision.
How do companies pay for data centers?
Companies typically pay for data centers through a combination of upfront capital expenses and ongoing operational expenses. Upfront capital expenses include the cost of building or leasing the physical space for the data center, as well as the cost of equipment such as servers, storage systems, and network infrastructure. Ongoing operational expenses include costs such as electricity and cooling to power and cool the equipment, as well as the salaries of personnel who manage and maintain the data center. Some companies also choose to outsource the management of their data center to a third-party provider, in which case they would pay a fee for those services. Additionally, some companies choose to use cloud-based data center services, in which case they pay for the use of the data center on a pay-as-you-go or subscription basis.
What is the most cost-effective type of data center?
The most cost-effective type of data center can vary depending on the specific needs and requirements of the organization. However, there are a few types of data centers that are generally considered to be more cost-effective than others:
1. Colocation data centers: Colocation data centers are facilities that allow organizations to rent space, power, and cooling for their own servers and other equipment. This can be a cost-effective option for organizations that do not want to invest in building and maintaining their own data center.
2. Hybrid data centers: A hybrid data center is a combination of on-premises data centers and cloud services. This allows organizations to take advantage of the cost-effectiveness of cloud services, while still maintaining some level of control over their data and infrastructure.
3. Micro data centers: Micro data centers are small, modular data centers that are designed to be deployed quickly and easily in a variety of locations. They are typically less expensive than traditional data centers, as they are smaller and use less power and cooling.
4. Cloud data centers: Cloud data centers are data centers that are owned and operated by a third-party provider and are available to users on a pay-as-you-go or subscription basis. This can be a cost-effective option for organizations that do not want to invest in building and maintaining their own data center.
It is worth mentioning that the most cost-effective option will depend on the specific use case, organization size and other factors, and it’s recommended to consult with professionals and conduct a cost-benefit analysis before making a decision.
Discussion and Limitations
As evidenced here, there are several business issues to be considered before investing in data centers. The type and location of the data center will influence the ability to ingest, store, and use data for analytics, reporting, operations, and innovation. These in turn will influence the ability to compete with other firms and adjust to ever-changing market conditions.
While this post addressed many interrelated business issues to consider, it is not possible to conceptualize and describe all of them in detail in one article. ChatGPT can help here too. My forthcoming article in CXOTech Magazine reports my interaction with ChatGPT as I asked it to list the major factors that technology leaders should consider about data centers. That is still just a conversation starter though. Data scientists and other leaders are encouraged to investigate other interpersonal and technological avenues to gain the insights they need to find out more.
As mentioned in Part 1of this series, two key limitations to using ChatGPT relate to its timeliness and the possibility that it may report erroneous information. ChatGPT uses information on the internet that was obtained in or before 2021 — more recent information can only be found via other searches of the professional and on-line literature, and by mining the minds of knowledgeable colleagues. These should be oft-used supplements to ChatGPT or any other large language model.
Thus, buyer beware. Take the information reported above with a grain of salt. For example, I tried to use other sources to verify the statistics ChatGPT reported about the proportion of a company’s budget that it should reflect. That was surprisingly difficult. The publicly available annual reports for the few large companies I investigated had no details about data center costs per se, though they did report the costs of property and software that one might use to generate a very rough guess. The other tactic I tried was more helpful and will likely help you as well — just use your other favorite search engine to find current information on data center prices, then follow up with vendors of interest.
A third limitation of ChatGPT and other large language models is that they may hallucinate (i.e., drift into irrelevant conversational paths, some of which can be entertaining while others can be offensive or hurtful). Users should take back control of the conversation and steer ChatGPT back to relevance when this happens.
Conclusion
Readers are advised to use multiple platforms to find information about any topic of interest to them. AI’s utility for data centers is no exception. Looking beyond ChatGPT for information about its pros and cons and for information about your subject matter of interest will generate a more complete knowledge base, leading to better decision making. Despite its limitations ChatGPT still is likely to be a helpful tool because of its ability to provide several quick and reasonably detailed insights, as it did regarding data centers. Part 3 of this series will address issues related to the economics and politics of data centers, along with environmental considerations that data scientists and other leaders should consider to round out their information search. I will also provide a summary of current thinking about how to guide ChatGPT so its results are as useful as possible. Stay tuned.
References
1. Roose K. A Conversation With Bing’s Chatbot Left Me Deeply Unsettled: A very strange conversation with the chatbot built into Microsoft’s search engine led to it declaring its love for me, on https://www.nytimes.com/2023/02/16/technology/bing-chatbot-microsoft-chatgpt.html February 16, 2023.
2. Neumister L. Lawyers blame ChatGPT for tricking them into citing bogus case law, on https://apnews.com/article/artificial-intelligence-chatgpt-courts-e15023d7e6fdf4f099aa122437dbb59b, June 8, 2023.
3. Alkaiss H, MacFarlane SI. Artificial Hallucinations in ChatGPT: Implications in Scientific Writing on https://assets.cureus.com/uploads/editorial/pdf/138667/20230219-28928-6kcyip.pdf, February 19, 2023, DOI DOI: 10.7759/cureus.35179.
Article by Ron Ozminkowski, PhD and ChatGPT
Ron is an internationally recognized consultant, analytics leader, and chief scientist whose published research has been viewed by people in over 90 countries. Educated in the SUNY system and The University of Michigan. Specializing in healthcare analytics, business strategy, economics, policy, and program evaluation. Founder & President of Analytics Strategy & Consulting LLC. Known for exceptional service to clients in The White House, federal and state agencies, life sciences companies, large employers and insurers, industry groups, and health systems. Over 100 endorsements each for leadership, analytic skills, and industry knowledge, on LinkedIn (see https://www.linkedin.com/in/ronozminkowski/). More info also can be found on https://www.analyticstrategiesconsulting.com and ronozminkowski.wordpress.com.
Originally posted on OpenDataScience.com
Read more data science articles on OpenDataScience.com, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Subscribe to our fast-growing Medium Publication too, the ODSC Journal, and inquire about becoming a writer.
