| Platform | DockerHub Official Image Tests | BareMetal Tests | Ampere Docker Tests |
|---|---|---|---|
| Ampere CIDR Altra | 1 | ||
| Ampere CIDR Altra Max | 1 | ||
| Azure | 1 | ||
| Equinix | 1 | ||
| 1 | |||
| HPE Altra Max | 1 | ||
| OCI Ampere A1 Compute | 1 |
Ampere AI delivers world class AI Inference solutions! The all new Ampere AI Optimized abstraction layer can deliver significant performance enhancements over typical machine learning frameworks with standard open source math libraries. In this case, Ampere is running our latest released images in full regression on any supported architecture including our own systems and public Cloud instances. See below for more on these exciting new AI tools from Ampere!
More information on Ampere CIDR Regression Suites.
Overview
Ampere AI
Ampere AI optimizes the performance of trained neural networks on standard AI frameworks. Customers can achieve up to 5x better performance for AI workloads by deploying Ampere AI in existing AI frameworks. Ampere AI is a best-in-class CPU-based Inference optimization engine that comes integrated into many AI frameworks (e.g. TensorFlow, Pytorch or ONNX) that accelerates inference on Ampere processors without introducing any accuracy loss, conversions, or model retraining. No API changes are required, and Ampere AI Optimization works with many common neural network architectures (including image classification, object detection, language processing and recommendation engines).
Ampere Optimized AI delivers better CPU-based inference performance (latency or throughput) than standard frameworks on open source libraries along with industry leading performance per watt for highly intensive inference workloads. Ampere AI Optimization can now be downloaded via the new Ampere AI Solutions page
DETAILS
Resources and Test Results
Here at Ampere we've built an extensive infrastructure focused on Continuous Integration, Delivery, and Regression (called CIDR).
Read More About CIDR
Our testing runs 24/7/365 in our regression infrastructure.
Basic Functional Test
Results are categorized as either 'Verified' or 'Unverified'.
Know More
Test and build infrastructure can encounter complexities or unexpected speed bumps. Known incidents and their resolutions will be documented where applicable.
DockerHub Introduces Image Pull Limits (2020.11)
Date,Platform and OS details of verified test results are displayed on hover bubble.
To give feedback on this portal please send us a message at
