As part of the plethora of newly announced products at [email protected] 2022, one of the most important updates come under the name of the NVIDIA AI platform.
Due to the professional nature of the platform, not a lot of people actually know what it is and what it does. Basically, it is a platform comprising of key SDKs and tools to help developers create a complex network of nodes for training and machine-learning workloads and there are several ways to tackle an AI-related issue through one of the following solutions. We’ll top them off with extra bits of information regarding what updates they are getting.
- NVIDIA Triton – A versatile open-source hyperscale model inference solution with the latest release adding a new Model Navigator for accelerated deployment of optimized models, Management Service for efficient scaling in Kubernetes, and Forest Inference Library enabling inference on tree-based models with explainability for fast, optimized, and scalable AI in every application.
- NVIDIA Riva 2.0 – AI SDK developed for speech-related models. Update expands speech recognition languages to 7, human-like deep learning-based text-to-speech with both male and female voices, as well as custom tuning via NVIDIA TAO Toolkit. New paid Enterprise version with direct support from NVIDIA added for more ambitious projects.
- NVIDIA NeMo Megatron 0.9 – Created to train Large Language Models (LLMs), it enables researchers and enterprises the ability to train any model to convergence and scale to trillions of parameters for applications such as conversational AI, recommenders, and genomics. Optimizations and new recipes result in shorter development and training time with cloud support added in latest update.
- NVIDIA Merlin 1.0 – 2 new libraries Merlin Models and Merlin Systems allows this AI-based recommending framework to deploy as microservices and is fully scalable
- NVIDIA Maxine – An audio and video quality enhancement SDK that reinvents real-time communications with AI, and introduces acoustic echo cancellation and audio super-resolution. The two new features enhance audio quality leading to a clearer communication experience/
Other than the updates, the NVIDIA AI Enterprise software suite also welcomed its 2.0 iteration where it has been certified by Red Hat OpenShift and have the OpenShift container framework made available on NVIDIA LaunchPad to create a consistent hybrid cloud foundation for scalable and containerized services either in the form of local bare-metal hardware or through virtualized systems via VMware vSphere. In addition to this, more NVIDIA AI software containers are joining the 2.0 suite like the NVIDIA TAO Toolkit providing fine-tuning and optimizations to NVIDIA pre-trained AI models can be accessed.
Availability
NVIDIA’s collection of AI software is available to developers as part of the NVIDIA Developer Program while Enterprises can experience NVIDIA AI software in curated labs for IT teams and AI developers on NVIDIA LaunchPad.
