Highlights:
- Google’s new starting CPU boost capability, released today, will help businesses increase the efficiency of their Cloud Functions and Cloud Run deployments.
- The number of extra CPUs that Google Cloud sets up for a company depends on how its cloud environment is set up.
Google LLC’s cloud business unveiled a new capability today called startup CPU boost that will allow customers’ applications to perform specific computing tasks with less latency.
The capability is currently in beta testing and is available for the Cloud Run and Cloud Functions services on Google Cloud.
Cloud Run is a service that allows businesses to run software container applications on the Google Cloud infrastructure. The service automates many maintenance tasks associated with running applications, thus saving developers time.
Cloud Functions is designed such that it can help save developers’ time. It’s a serverless computing service that, like Cloud Run, automates many historical manual tasks associated with running software on Google Cloud’s infrastructure. Google Cloud released an upgraded service version with several performance improvements last month.
Google’s new starting CPU boost capability, released today, will help businesses enhance the efficiency of their Cloud Functions and Cloud Run deployments. It does this by lowering the time it takes for apps to start processing data. This is an essential metric as it determines how quickly an application can complete its intended activities.
Most enterprise applications aren’t built as a single file but as a group of separate software modules. Each module runs in a separate software container. Often, companies set up their applications so that modules are launched only when needed, and then they are deleted automatically when they are no longer required. This helps reduce the amount of hardware used.
Launching an application module can take a long time, and requests to the application may be processed slowly until the task is done. Because of this, users may experience more delays. Google Cloud’s new CPU boost feature is meant to help companies reduce the time it takes for application modules to start up or “launch.”
The feature can be accessed as an opt-in setting via the Google Cloud management console. When the feature is turned on, it finds out when a new application module is about to launch and gives additional central processing units to the cloud instance that runs the module. With additional CPUs, the module can load faster, reducing user latency.
The number of additional CPUs that Google Cloud sets up for a company depends on how its cloud environment is set up. Google Cloud says that its platform provisions between two and eight CPUs to speed up the time taken by the application to load.
The company analyzed how well the new startup CPU boost worked in a series of internal benchmark tests. During the tests, the capability minimized the start-up time of Java applications by up to 50%. Google says that it can also improve the performance of applications written in other programming languages in a big way.
“Customers testing the feature in private preview with Node.js have observed startup time reductions of up to 30%, a significant improvement, a bit less than Java due to the single-threaded nature of Node.js.”, Steren Giannini, Google Cloud group project manager, explained in a blog post. “Each language, framework, and code base will see different levels of benefit.”
