Ampere^® Altra^® processors are designed from the ground up to deliver exceptional performance for Cloud Native applications such as video encoding. With an innovative architecture that delivers high performance, linear scalability, and amazing energy efficiency, Ampere Altra allows workloads to run in a predictable manner with minimal variance under increasing loads.

This enables industry leading performance/watt and a smaller carbon footprint for real world workloads like video encoding. Google Cloud offers the cost-optimized Tau T2A VMs powered by Ampere Altra processors for scale-out Cloud Native workloads in multiple predetermined VM shapes – up to 48 vCPUS per VM, 4 GB of memory per vCPU, up to 32 Gbps networking bandwidth, and a wide range of network-attached storage options.

These VMs are suitable for scale-out workloads like web servers, containerized microservices, data-logging processing, media transcoding, and Java applications. This workload brief focuses on x265 which implements the H.265/MPEG-H Part2 standard, the second most widely used video codec today after H.264. Compared to x264, more advanced video codecs such as x265 provide greater video compression at the expense of greater computing resources and power usage.

The Google Cloud Tau T2A VMs powered by Ampere Altra processors offer great performance in a variety of video encoding workloads, including x265. We evaluated the performance of the x265 codec using three different input videos with resolutions of 480p, 720p, and 1080p from the UGC Dataset, uploaded to YouTube and distributed under the Creative Commons license, using x265’s medium, slower, and placebo presets.

Ampere Altra-based Google Cloud Tau T2A VMs outperform the x86 VMs on raw performance. We plot relative performance averaged over 3 input videos each run with x265 medium, slower & placebo presets. The T2A VM has 6% better performance than the N2 VM and 9% better compared to the N2D VM.

Comparing price-performance, the T2A VMs outperform the legacy x86 VMs even further. Altra T2A VM has 34% better price-performance than the N2 VM and 19% better compared to the N2D VM.

Ampere Altra processors have been designed from the ground up to address challenges in multitenant cloud environments like noisy neighbors and scalability. To highlight this, we studied the aggregate FPS on the Ampere Altra-based Google Cloud Tau T2A VM as the number of simultaneous x265 jobs running the placebo preset with the 1080 input video was increased. This is a computationally demanding x265 configuration, and the T2A VM demonstrated perfect, linear scaling from 1 to 8 jobs. The N2 and N2D VMs, on the other hand, showed little improvement in FPS beyond 4 vCPUs due to contention from Hyper-Threading.

We used the following input files from the UGC Dataset, uploaded to YouTube distributed under the Creative Commons license:

• https://storage.googleapis.com/ugc-dataset/original_videos/Sports/480P/Sports_480P-0623.mkv

• https://storage.googleapis.com/ugc-dataset/original_videos/Sports/720P/Sports_720P-00a1.mkv

• https://storage.googleapis.com/ugc-dataset/original_videos/Sports/1080P/Sports_1080P-0063.mkv

For each vCPU available in the VM, we ran the following command to evaluate x265 performance.

Copy
x265 --preset $PRESET --frame-threads 1 --no-wpp $INPUT --input-res $INPUT_RES --fps 30 -o $OUTPUT

Video encoding is a key and popular workload in the cloud and advanced codecs that are compute intensive like x265 are used frequently. The H.265/MPEG-H Part2 compression standard is the second most widely used video format today after H.264 and x265 is the leading H.265 implementation.

In our tests, the Google Cloud Tau T2A VMs powered by the Ampere Altra Cloud Native processors delivered perfect scaling by reducing noisy neighbor issues and provided better performance and price-performance compared to legacy x86 VMs - up to 6% higher performance and 34% higher price-performance.

For more information about the Google Tau T2D virtual machines with Ampere Altra processors, please visit the Google Cloud blog for further details.