The AI-driven economy is booming, and businesses are eager to harness the power of artificial intelligence and machine learning. However, they face a significant challenge: connecting their sensitive, proprietary data sets to high-performance AI training facilities without sacrificing security or investing heavily in infrastructure.
Optical transparency in modern networks eliminates costly optical-electrical-optical conversions, significantly reducing latency, power consumption, and hardware complexity while offering significant economic benefits. Optical circuit switches (OCS) build upon these advantages, bringing optical transparency to data center networks and offering a compelling solution for data center architectures.
The rapid evolution of network technologies, particularly the rise of open network architectures, has brought about significant changes in the way networks are designed, tested, deployed, and managed. Open optical networks are networks that utilize disaggregated hardware and software components from multiple vendors, enabling greater flexibility, innovation, and cost-efficiency.
The rapid advancement of machine learning (ML) has led to a surge in demand for high-performance computing infrastructure with large GPU clusters needed to meet the exponentially increasing size of the training data sets.
It’s hard to read any business article today without seeing a mention about the expected impact of machine learning (ML). Potential benefits include improvements in diagnostic accuracy and personalized treatment plans in medicine, better customer service and chatbots in retail, intelligent tutoring systems in education and enhanced risk managements and fraud detection in finance.
The ever-growing complexity of machine learning (ML) models is fueling a data deluge. These models, used in everything from recommendation engines to large language models (LLMs), require massive amounts of data for training.
The ever-growing demands of the digital age have necessitated the rise of hyperscale data centers – massive infrastructure that support the colossal demands of cloud computing, data centers, and internet applications.
Data centers are hungry for power, and their appetite is only growing. The demand for data storage and online services has fueled a steady rise in data center power consumption.
In the technology world, 2023 was dominated by headlines about machine learning (ML) programs such as ChatGPT and Dall-E. Large language models (LLMs) and generative AI fascinated people with the ability to generate text following almost any prompt and an image created by the generative AI program Midjourney even won an art contest.
Automation can address the “most labor-intensive and error-prone step” during re-striping
If you are reading this blog, it is likely that you are already acquainted with the concept of Large Language Models (LLMs) and generative Artificial Intelligence (AI).
A recent NANOG blog post delved into the critical need for the automation of optical fiber transport networks [1]. The article underscored the imperative for automation to drive down network costs, enhance network resilience and reduce the number of outages – all of which significantly improves customer satisfaction. However, the author candidly noted that “Network automation is moving slower than we thought it should, and slower than we need it to move.”
The ‘cloudification’ of information technology has produced benefits in seemingly every aspect of business. The cloud has created a virtualized compute oversubscription model by automating functions at the higher layers of the Open Systems Interconnection (OSI) stack.
Antonio Bernardo, Dean and Professor of Finance, UCLA Anderson School of Management
Whether and When to Invest in Telescent Technology
This question of whether to deploy automation can be broken down into two decision points: whether to invest and when to invest in the Telescent technology.
Decision 1: Whether to invest
Cost savings are extremely compelling; not hypothetical – cost savings have been achieved with other clients.
10x ROI for unmanned data centers
30% upfront network CapEx savings, 41% network energy savings for hyperscale data centers
Cash flow benefits are large relative to CapEx – will allow Data Center Operator to maintain considerable financial flexibility.
Data center operators will be at a competitive disadvantage if they do not invest in disruptive technologies used by competing firms
Decision to invest is clear and incontrovertible.
Decision 2: When to invest
Why do it now?
Achieve cost savings as soon as possible.
Improve energy efficiency of data center as soon as possible
Avoid continuing/compounding mistakes inherent in manual patch-panel deployments (incorrect records, incorrect billing, service interruption)
Past expense of current manual patch panel technology is irrelevant; these are sunk costs which should not deter the data center operator from investing in new value-creating technology.
Why wait?
Switching to Telescent technology can disrupt service to data center operator’s customers.
Telescent solution is drop-in replacement for manual patch-panels; it is relatively simple to switch from manual to automated.
This is not a good reason to wait because these disruption costs can be minimized.
Concern that rapid transition exposes the customer to considerable systemic risk if the technology doesn’t work.
Experience with other data center operators that have deployed technology indicates that this risk is minimal.
Concern that Telescent won’t be around in the future. Customers want to be assured company will be around to service equipment, provide upgrades, etc.
Telescent can reassure data center operators with its strong financial backing, a growing client base, and strong future prospects.
Flextronics Inc., a multinational contract manufacturer, handles reverse logistics such as spares, returns, repairs, etc.
Telescent may not be able to ramp up quickly enough.
Flextronics Inc. can increase production quickly.
Telescent can scale required personnel through existing network equipment installers and it can train the client for self-installation.
System is maintenance-free and highly reliable so only minimal service infrastructure is required.
Telescent interconnect automation systems already exceed 1 billion interconnect hours in operation across the globe so the deployment risk is small. For operators of data centers, we believe it is advisable to begin the rollout right away to achieve cost savings and improve service delivery as rapidly as possible. The modular and scalable nature of the Telescent solution ensures that the rollout can be flexible and incremental for additional risk mitigation, enabling data center operators to accelerate/decelerate the deployment as they gain experience with the transition process and costs.
Antonio Bernardo Biography
During his 25 years at UCLA Anderson, Dean Antonio “Tony” Bernardo estimates he’s taught more than 5,000 students. Bernardo says some of the best moments are when students walk into his office to dig deeper into a business issue — to learn beyond the classroom material. “Everyone on the faculty I know loves meeting with a hard-working, motivated student,” he says.
In 2019, Bernardo was appointed UCLA Anderson’s ninth dean, effective July 1.
Bernardo’s research spans various areas of corporate finance. His recent research on bailouts provided a set of policy recommendations to lawmakers looking to design bailouts for distressed firms. His current research looks at optimal capital structures, or how much debt and equity financing a firm should have. It suggests that capital structure decisions are determined, in part, by the debt choices other firms in the industry are making.
Another area Bernardo is exploring involves optimal compensation of doctors, and designing compensation systems for health providers. “I tend to work on topics that are very diverse,” Bernardo says. “I learn a lot that way.”
In addition to having an impressive list of published papers and working papers, as well as receiving numerous teaching awards, Bernardo serves as associate editor of the Critical Finance Review and Financial Management.
Education
Ph.D. Economics, 1994, Stanford University
B.A. Economics, 1989, University of Western Ontario
Mark Zuckerberg said it loud and clear, “2023 is the year of efficiency.” For Meta and other tech companies, that means doing more with less. Automation is one way to drive efficiency, taking repeatable tasks and implementing a solution that allows people to be more efficient and consistent at what they do.
While the above Too Long : Didn’t Read summary is said somewhat tongue-in-cheek knowing that there is a lot of complexity in the replacement of optical for electrical switches, the recent paper from Google discussed their use of an optical circuit switch (OCS) in their data center networking and the tremendous value the received due to this replacement.
A huge variety of our daily activities are affected by machine learning algorithms. Machine learning is used in everything from Siri's speech recognition to your choice of movie based on a Netflix recommendation to which ad shows up in your timeline when you open Facebook.
Happy International Data Center Day! Today, we celebrate the vital role that data centers play in our increasingly digital world.
High Performance Computing (HPC) and Machine Learning (ML) rely on collective communication – concurrently aggregating and distributing data collected from processes running on clusters of interconnected compute nodes.
This episode features a conversation with executives from MOX and Telescent as they provide an overview of their respective organizations and reveal how MOX deploys the Telescent Network Topology Manager to remotely automate their customer’s fiber connectivity between data center locations.
