Improve Wide & Deep Inference Performance with Azure Esv4 VMs featuring 2nd Gen Intel® Xeon® Scalable processors
Using a subset of machine learning—called deep learning—to make relationships between customer data can deliver real-time recommendations that help consumer locate what they seek. Utilizing wide linear models and deep neural networks to infer relationships between data, Wide & Deep workloads deliver real-time recommendations based on that data. Selecting Microsoft Azure Esv4 VMs enabled by 2nd Gen Intel® Xeon® Scalable processors over Esv3 VMs with previous-generation processors can improve Wide & Deep recommendation engine performance. The 2nd Gen Intel Xeon Scalable processor family features Intel Deep Learning Boost, which improves machine learning performance.
To determine which configuration offers better performance, independent third-party Principled Technologies tested Wide & Deep performance across three different VM sizes. Azure Esv4 VMs featuring Intel Xeon Platinum 8272CL processors handled up to 3.48x more samples per second than Esv3 VMs. With Esv4 VMs, organizations can deliver real-time recommendations based on the data they collect even faster, which can improve customer satisfaction and boost overall sales.
Improve Deep Learning Performance on Small Instances
The faster your cloud VMs can infer meaningful relationships between data, the faster you can make recommendations to consumers. As Figure 1 shows, 8-vCPU Esv4 VMs enabled by 2nd Gen Intel Xeon Scalable processors outperformed 8-vCPU Esv3m VMs in a deep learning Wide & Deep benchmark test. The Intel Xeon processor-based VMs handled 3.48 times the samples per second that the previous-gen VMs did, which means they can process data and make recommendations faster.
Improve Deep Learning Performance on Medium Instances
Organizations with mid-sized datasets can also get improved deep learning inference performance by choosing VMs with newer processors. As Figure 2 shows, 16-vCPU Azure Esv4 VMs enabled by 2nd Gen Intel® Xeon® Scalable processors handled 3.23 times the samples per second in Wide & Deep tests compared to Esv3 VMs with previous-generation processors.
Improve Deep Learning Performance on Large Instances
Larger datasets that require larger VMs similarly benefit from choosing newer processor architecture for deep learning workloads. In tests, Esv4 VMs featuring 2nd Gen Intel Xeon Scalable processors handled 2.99 times the samples per second using the Wide & Deep benchmark test (see Figure 3).
For datasets small, medium, and large, selecting Azure Esv4 VMs with 2nd Gen Intel Xeon Scalable processors over Esv3 VMs with previous-generation processors can boost deep learning performance to form meaningful relationships from data and make fast real-time recommendations to consumers.