Not logged in » Login
Mar 18 2020

Optimizing data centers resources using quantum-inspired computing

Given the world’s increasing dependency on IT, the complexity of data centers is growing significantly. Infrastructure landscape optimization, which businesses are heavily dependent on to serve customers, will be a great ally in operational and maintenance cost reductions. According to best practices, capacity management is an essential IT process with the objective of ensuring sufficient resource capacity for any given IT service as part of its design stage. Aiming at optimal performance, Fujitsu’s Digital Annealer is here to help you overcome your data center challenges, optimizing capacity while reducing costs, resource wastage, etc.

Image

The work of data center managers in relation to IT infrastructure capacity management is often heavily under-estimated. They have to make sure their organization’s data center strategy stays ahead of organization expansion needs as well as watching out for unexpected peak requirements that may overwhelm current systems. This can be achieved through data center capacity planning. There are various tools, either in-house or professional tools used for this purpose. For the SAP environment there are SAP infrastructure inspection services which are used to optimize the resource capacity distribution.

However, the tools can be as effective as the underlying hardware. Identifying the most optimal distribution pattern could be extremely cumbersome with evaluation required across 1010 possibilities or more, depending upon how large the datacenter is. If existing CPU and GPU-based hardware were used to evaluate across all the possibilities, it would take more than a lifetime. So, to challenge one of our latest cutting edge innovations – Digital Annealer (DA), a quantum-inspired computing technology, we did a multi-case datacenter capacity management study on DA.

Before we get in to the details of the test case, let’s take a step back to understand what Digital Annealer, quantum-inspired computing is. It is a computational architecture based on a digital circuit design, inspired by quantum phenomena and unique to Fujitsu, that overcomes the current gap in quantum technology and paves the way for a much faster and more efficient approach of solving today’s business problems. Digital Annealer is not only able to solve optimization problems, but to solve them in seconds.

While quantum computers promise to calculate problems that are not solvable with conventional CPU- and GPU-based technology, its usage in real business scenarios is not economically viable at this time. In order to overcome this limitation, Fujitsu came up with a solution which aims at solving the real world optimization problems today.

Now, coming back to utilizing Digital Annealer for data center capacity management: The need for resource capacities can vary across the year, depending on your type of market (whether you operate in manufacturing, automotive, retail etc.). In those scenarios where your company needs to increase its production and your IT capacity should be reallocated accordingly. Digital Annealer can be beneficial to successfully manage the varying capacity needs, ensuring the customer requirements are met in a much faster, near real-time manner.

 

Multi-case data center capacity management study

The optimization problem to work on was the data center capacity optimization in terms of its resources. We conducted a study on a multi-case data center capacity management, monitoring 246 data centers hosting enterprise systems. As infrastructure optimization challenges of this kind are addressed by the Fujitsu SAP System Inspection Service, the Digital Annealer was applied across all servers, in order to find optimal SAP landscape designs and, therefore, to identify savings potential. With the goal of solving server consolidation problems that are modeled from real-world cases, the solution space comprised 1*10¹⁰ solution candidates, making it extremely difficult for CPU- and GPU-based systems to evaluate all possibilities and determine which is the most optimal.

In this scenario, what Digital Annealer is able to do is to evaluate all the possibilities at the same time in a matter of seconds, identifying the optimal solution in no time. When comparing the results to other state-of-the-art optimization algorithms, the Digital Annealer finds optima more often than any other algorithm and, in 13.4% of the cases, a better solution than any other algorithm. In the majority of the tested cases, the Digital Annealer outperforms state-of-the-art algorithms in terms of solution quality and runtime. For complete technical information on how this solution was conducted and how Fujitsu mapped the algorithm, please take a look at the available whitepaper.

There are several other business applications where the Digital Annealer can help you, and while working closely with your organization, Fujitsu will map the specific business scenario, applying the mathematical model and algorithms in order to ensure the optimal solution is delivered.

Image

 

Sushant Dhamnekar

 

About the Author:

Sushant Dhamnekar
close share field

Sushant Dhamnekar studied Embedded System Design and worked as software developer and tester. Started working as solution architect for analytics with Fujits...

SHARE

Comments on this article

No comments yet.

Please Login to leave a comment.

X

Please login

Please log in with your Fujitsu Partner Account.

Login


» Forgot password

Register now

If you do not have a Fujitsu Partner Account, please register for a new account.

» Register now