SEARCH FINANCIAL SERVICES INFRASTRUCTURE SECURITY SCIENCE INTERVIEWS

 

     

AWS, GCP and Azure Performance Measured

November 9, 2018

ThousandEyes issued the results of the 2018 Public Cloud Performance Benchmark Report, the industry's first report to compare global network performance of the three major public cloud providers—Amazon Web Services (AWS), Google Cloud Platform (GCP) and Microsoft Azure. The report was created to provide an agnostic resource for enterprise IT business leaders to help guide their multi-cloud decision-making process.

According to Gartner Research, Forecast: Public Cloud Services, Worldwide, 2016-2022, 2Q18 Update, "The worldwide public cloud services market is projected to grow 17.3 percent in 2019 to total $206.2 billion, up from $175.8 billion in 2018. The fastest-growing segment of the market is cloud system infrastructure services (infrastructure as a service or IaaS), which is forecast to grow 27.6 percent in 2019 to reach $39.5 billion, up from $31 billion in 2018."

Given the massive growth and investments in public cloud providers, ThousandEyes published the Public Cloud Performance Benchmark Report to give IT teams a look at detailed point-in-time performance metrics of the big three providers.

"Multi-national organizations that are embracing digital transformation and venturing into the cloud need to be aware of the geographical performance differences between the major public clouds when making global multi-cloud decisions," said Archana Kesevan, report author and senior product marketing manager at ThousandEyes. "To help global businesses with this assessment, ThousandEyes is providing an unbiased, third-party perspective on public cloud performance as it relates to end-user experience—and at the same time, breaking the mold of survey-based research and vendor-led reporting."

Key findings of the 2018 ThousandEyes Public Cloud Performance Benchmark Report:

Architectural differences between providers impacts service delivery: AWS sends traffic over the Internet for the majority of the service delivery path, whereas GCP and Azure do not, instead using their own backbone networks. Increased exposure to the Internet means there is greater operational risk and impact on performance predictability.

Performance variations by region: geographical performance variations exist across the three cloud providers, most noticeably in the LATAM and Asia regions. Decision-makers should consult the detailed findings to choose the best cloud provider on a per-region basis to ensure optimal performance globally.

Multi-cloud network performance is strong: despite being competitors, the three providers peer directly with one another, eliminating the dependence on third-party ISPs. Plus, traffic almost never leaves the provider backbone networks, meaning there is very little loss and jitter in end-to-end communication. Decision-makers need not worry about performance in multi-cloud architectures.

Top Regional Findings on Public Cloud Performance:

Importantly, all findings in the Public Cloud Performance Benchmark Report should be examined through the lens of the individual business planning or evaluating their cloud architectural choices as regional performance differences can make a significant impact in terms of performance gains or losses.

Most notably:

When connecting Europe to India, GCP exhibited three times the network latency compared to AWS and Azure.

In Asia, GCP and Azure exhibited more network performance stability than AWS, which demonstrated 35% less network performance stability than GCP and 56% less than Azure.

When connecting Europe to Singapore, AWS and GCP were 1.5 times slower than Azure.

"As enterprises increasingly move to cloud-first IT strategies, IT increasingly depends upon infrastructure assets they don't control," said Jason Bloomberg, President at Intellyx. "The 2018 ThousandEyes Public Cloud Performance Benchmark Report lifts the veil obscuring region-to-region and cloud-to-cloud performance within and between public clouds, providing IT infrastructure execs critical intelligence on the performance they can expect from single cloud, multi-cloud, and hybrid IT environments."

The 2018 ThousandEyes Public Cloud Performance Benchmark Report is based on metric data from network performance measurements, such as packet loss, latency and jitter. Performance data was collected over a period of 30 days and results derived from over 160 million data points, using both end-to-end performance and in-depth path measurement techniques.

Terms of Use | Copyright © 2002 - 2018 CONSTITUENTWORKS SM  CORPORATION. All rights reserved. | Privacy Statement