menu
techminis

A naukri.com initiative

google-web-stories
Home

>

Google News

>

How Verve ...
source image

Cloudblog

1M

read

105

img
dot

Image Credit: Cloudblog

How Verve achieves 37% performance gains with C4 machines and new GKE features

  • Verve, the advertising solutions specialist, has benefited from a 37% improvement in platform performance by integrating Google Cloud's high-performance C4 virtual machines into its infrastructure.
  • C4 is fuelling Verve's growth by reducing latency and improving the 5-7% rate of ad-fill achieved by the platform on Google Kubernetes Engine (GKE). This means the company needs to pay for less traffic that automatically scales up when a bid-winning ad, generating revenue, is requested.
  • Another GKE-aided win comes from adopting Google's GKE Gateway, which leverages Google's Envoy-based global load balancers, to reduce computing costs and latency by 20% to 25% and incorporate automated scaling to handle traffic surges.
  • Most recently, Verve has started working with Google's Custom Compute Classes (CCCs), a Kubernetes-native, declarative API that prioritises available compute requirements by metrics like price, availability and performance.
  • The company was one of the first to use the service during its Early Access phase and set benchmarking by migrating its marketplace from N2D machines to C4.
  • Through the CCC service, Verve was able to establish C4 as its preferred machine while also avoiding a false dichotomy of having to choose between saving IT costs and lower machine availability.
  • It plans to improve its compute utilisation and scalability by deploying GKE's Node Autoprovisioning and Custom Compute Classes in the next stage of its growth program.
  • In summary, Verve realised a 37% improvement in platform performance using C4 machines and saved costs leveraging GKE Gateway for load balancing, which translated into a 7.5% revenue boost. Custom Compute Classes optimised its compute preference and usage.
  • C4 machines are designed for performance-sensitive workloads, built on the latest Intel Xeon Scalable processors, and offer high bandwidth, low latency networking up to 200 Gbps.
  • Varying vCPU and memory configurations make the C4 VMs suitable for many mission-critical requirements, including financial modelling, inference, databases and gaming applications.

Read Full Article

like

6 Likes

For uninterrupted reading, download the app