{"id":10452,"date":"2022-10-16T00:09:08","date_gmt":"2022-10-15T21:09:08","guid":{"rendered":"https:\/\/globalit.com.tr\/record-from-google-compute-engine\/"},"modified":"2023-08-30T14:20:21","modified_gmt":"2023-08-30T11:20:21","slug":"record-from-google-compute-engine","status":"publish","type":"post","link":"https:\/\/globalit.com.tr\/en\/record-from-google-compute-engine\/","title":{"rendered":"Record from Google Compute Engine"},"content":{"rendered":" \n
Google’s Compute Engine, a solution with high-performance virtual machines, has calculated the 100 trillionth digit of Pi, surpassing its own and a university’s records. This activity, which took more than 157 days, was made possible by Google Compute Engine’s infrastructure that can scale to heavy workloads. <\/p> \n \n
In general terms, the number Pi, the first two digits of which we know, stands out as a mathematical constant that gives the ratio of the circumference of a circle to its diameter and is used as about 3.14. The number Pi, which is an irrational number, in other words cannot be expressed exactly as the ratio of two integers, remains one of the most mysterious elements of the discipline of mathematics. The never-ending decimal representation of pi, the digits that extend to infinity and do not turn into a repeating pattern, also allow for new experiments. <\/p> \n \n
While many civilizations since the Egyptians have needed highly accurate Pi number approximations for practical calculations, for thousands of years mathematicians have been trying to calculate this mysterious number. Calculations that could have taken centuries with human power can now set new records in less than a year with human-developed computing power. <\/p> \n \n
With the recent record <\/a>broken by Google Compute Engine (GCE), humanity knows more today about the number Pi, for which special occasions are held, with William L. Schaaf, author of The Nature and History of Pi, saying, “Probably no symbol in mathematics has aroused as much mystery, romance, misunderstanding, and human interest as the number Pi.” <\/p>
\n
\n As Global IT, Google’s first cloud solution partner in Turkey, today Google offers a virtual machine (virtual machine | VM) solution, we focus on the second record that Google Compute Engine has set in digits of the number Pi. <\/p>
\n
\n Google set a world record as the calendars showed 2019: It calculated the 31.4 trillionth digit of pi using the power of cloud platforms and virtual machines. This record was broken in 2021 by scientists at Grisons University of Applied Sciences, bringing the total number of digits calculated to 62.8 trillion. Not long after its own record, in just 3 years, Google set a new record last June, increasing the number of calculated digits to 100 trillion, as if to show how far the limits and capacity of Google Compute Engine have come. <\/p>
\n
\n This success, which is a testament to how much Google Cloud infrastructure has evolved year after year, was made possible by the power of Compute Engine, the secure customizable computing service of Google Cloud, Google’s family of cloud solutions. The process, which started on the morning of Thursday, October 14 with a system with features such as 128 virtual CPUs and 864 GB of RAM, was completed at similar times on March 21. 515 terabytes of 663 terabytes of data space were used for the process, which took 157 days, 23 hours, 31 minutes. For the calculation, 43.5 petabytes of data were read and 38.5 petabytes of data were written. <\/p>
\n
\n This experiment, which corresponds to a compute, storage, and network-intensive workload, was performed on the n2-highmem-128 virtual machine running Debian Linux 11. Offering a high-performance CPU, large memory, and 100 Gpbs of throughput bandwidth, this machine was a member of the most popular family of general-purpose virtual machines offered by Google Compute Engine. <\/p>
\n
\n In the project where a network-based shared storage architecture was adopted, the N2 machine was preferred, which provides balanced price\/performance. Terraform, an open-source Google Cloud tool, was chosen for creating and managing the working set, while automating mission-critical tasks. Terraform scripts were leveraged to help ensure that the required software packages were installed automatically. <\/p>
\n
\n “We knew that the calculation would take several months, and that even a small performance difference could change the runtime by days or weeks,” Google’s blog post on the subject said, noting that the final program and product design achieved through technical improvements allowed the calculation to be completed in 157 days instead of 300 days. <\/p>
\n
\n What does it mean that Google Compute Engine calculates the 100 trillionth digit of Pi?<\/strong><\/p>
\n
\n By the end of the project, where Chudnovky was preferred, which stood out as a widely used algorithm for calculating the number pi, the result obtained was confirmed by the Bailey-Borwein-Plouffe formula. The auditory and visual equivalents of the 100 trillion digits obtained can be accessed from this link<\/a> . With a note given to each digit of the number pi, it can be experienced in an entertaining way how 100 trillion digits turn into a composition and how a visual work emerges by associating different steps with different colors. <\/p>
\n
\n This study, pioneered by Google, is seen as an experiment and is important in terms of revealing the processing capacity of the virtual machines offered within the scope of the GCE. It demonstrates how Google Compute Engine’s flexible infrastructure enables teams around the world to push the boundaries of scientific experimentation. <\/p>
\n
\n GCE, which can run for more than five months without errors and accurately handles every bit of 82 petabytes of data, is the result of the improvements and improvements Google has made over the past three years. The calculation of this 100 trillion-digit number, which would take 3.1 million years <\/a>if each of them were tried to be said out loud, in a period of 157 days, shows the last point that cloud computing has reached. On the other hand, it proves that a high-performance computer from Google for general use can run the heavy workloads of any business, regardless of scale, without interruption. <\/p>
\n
\n Google Compute Engine to break the boundaries of business innovation<\/strong><\/p>
\n
\n Compute Engine, which runs on Google’s cloud platform and turns high-performance computers located in the company’s different data centers into virtual machines for a business anywhere in the world, offers an infrastructure that can sustain workloads in every industry without interruption at highly affordable costs. <\/p>
\n
\n Google Compute Engine, which is based on mathematics and can fit an experiment that could take years if done with muscle power into 157 days, can be customized for many different sectors and business processes, from streaming platforms to enterprise databases, from game servers to artificial intelligence applications, from scientific systems and development environments to retail companies’ store and e-commerce platforms, with predefined virtual machines optimized in terms of performance, memory or cost. <\/p>
\n
\nCalculating 100 trillion digits of pi using Google Compute Engine<\/h2>
\n
\n
How did the GCE calculate the 100 trillion digits of the number Pi?<\/h2>
\n
\n