When Intel previewed an array of data-centric innovations in August 2018, one that captured media attention was Intel® Deep Learning Boost, an embedded AI accelerator in the CPU designed to speed deep learning inference workloads.

Intel DL Boost will make its initial appearance in the upcoming generation of Intel® Xeon® Scalable processors code-named Cascade Lake. In this Chip Chat podcast, Intel Data-centric Platform Marketing Director Jason Kennedy shares details about the optimization behind some impressive test results.

The key to Intel DL Boost – and its performance kick – is augmentation of the existing Intel® Advanced Vector Extensions 512 (Intel® AVX-512) instruction set. This innovation significantly accelerates inference performance for deep learning workloads optimized to use vector neural network instructions (VNNI). Image classification, language translation, object detection, and speech recognition are just a few examples of workloads that can benefit.

Early tests have shown image recognition 11 times faster using a similar configuration than with current-generation Intel Xeon Scalable processors when launched in July 2017. Current projections estimate 17 times faster inference throughput benefit with Intel® Optimized Caffe ResNet-50 and Intel Deep Learning Boost that can be achieved with a new class of advance performance CPUs debuting in the upcoming generation.

For more information about AI activities across Intel visit ai.intel.com.

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more information go to www.intel.com/benchmarks. 

Performance results are based on testing or projections as of 6/2017 to 11/7/2018 and may not reflect all publicly available security updates. See configuration disclosures in https://intel.ly/2Mw7KKE and https://intel.ly/2VUvY2I for details. No product can be absolutely secure. 

[1] Future Intel Xeon Scalable processors codenamed Cascade Lake provide up to 11x inference performance with Intel® Optimized Caffe, ResNet-50, and Intel® Deep Learning Boost (VNNI) in comparison to current generation Intel Xeon Scalable Platinum 8180 Processor with FP32 instructions (at launch, measured July, 2017), for details see https://intel.ly/2Mw7KKE, slide 33.

[2] Future Intel Xeon Scalable processors codenamed Cascade Lake advanced performance provide up to 17x images per second (deep learning inference)  in comparison to current generation Intel Xeon Scalable Platinum 8180 Processor (at launch, measured July, 2017), for details see https://intel.ly/2VUvY2I.

Accelerating AI Inference with Intel® Deep Learning Boost – Intel® Chip Chat episode 632

Lisa Spelman, Vice President and General Manager Intel® Xeon® Products, Data Center Marketing at Intel Corporation, shares the origin of Intel’s data centric computing message as a customer-focused framework to tackle the today’s ultimate challenge and opportunity - data. Intel works with broad partner ecosystems to deliver hardware and software innovation in compute, storage and networking that supports the movement, storage and processing of every bit of data for better business outcomes.  

Cloudification of services and infrastructure is pervasive, which accelerates the utilization and creation of this data. Intel has invested over ten years in building the silicon foundation and software ecosystem for the cloud and cloud native architecture. That revolution has moved into the communication service provider market to support delivery of new services in a more cost-effective manner. Lisa talks about the convergence of 5G, cloud, edge and Internet of Things (IoT) that drives requirements for artificial intelligence, visual workloads and smart computing on the edge. 

Follow Lisa on Twitter @LisaSpelman.

Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration.  No product or component can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com. 

Intel, Xeon, and the Intel logo are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries.  

*Other names and brands may be claimed as the property of others. 

© Intel Corporation

Data-Centric Computing at Intersection of AI, Edge, IoT, 5G and Cloud – Intel® Chip Chat episode 639

Intel Chip Chat is a recurring podcast series of informal interviews with some of the brightest minds in the industry, striving to bring listeners closer to the innovations and inspirations of the people shaping the future of computing, and in the process share a little bit about the technologists themselves.

Accelerating AI Inference with Intel® Deep Learning Boost – Intel® Chip Chat episode 632

Intel Chip Chat 2.0

Related tracks