Jetson AI Labs – E02 – February 25, 2021

Join the NVIDIA Jetson team for the latest episode of our AMA-style live stream, Jetson AI Labs.

Performance Required for Deep Learning

There is this question that I wanted to find out about deep learning. What are essential System, Network, Protocol that will speed up the Training and/or Inferencing. There may not be necessary to employ the same level of requirements from Training to Inferencing and Vice Versa. I have received this information during a Nvidia Presentation

Training:

  1. Scalability requires ultra-fast networking
  2. Same hardware needs as HPC
  3. Extreme network bandwidth
  4. RDMA
  5. SHARP (Mellanox Scalable Hierarchical Aggregation and Reduction Protocol)
  6. GPUDirect (https://developer.nvidia.com/gpudirect)
  7. Fast Access Storage

Influencing

  1. Highly Transactional
  2. Ultra-low Latency
  3. Instant Network Response
  4. RDMA
  5. PeerDirect, GPUDirect

 

 

What is the difference between a DPU, a CPU, and a GPU?

An interesting blog to explain what is the difference a DPU, CPU, and GPU?

 

So What Makes a DPU Different?

A DPU is a new class of programmable processor that combines three key elements. A DPU is a system on a chip, or SOC, that combines:
An industry standard, high-performance, software programmable, multi-core CPU, typically based on the widely-used Arm architecture, tightly coupled to the other SOC components

A high-performance network interface capable of parsing, processing, and efficiently transferring data at line rate, or the speed of the rest of the network, to GPUs and CPUs

A rich set of flexible and programmable acceleration engines that offload and improve applications performance for AI and Machine Learning, security, telecommunications, and storage, among others.

For more information, do take a look at What’s a DPU? …And what’s the difference between a DPU, a CPU, and a GPU?

NVIDIA to Acquire Arm for $40 Billion, Creating World’s Premier Computing Company for the Age of AI

NVIDIA and SoftBank Group Corp. (SBG) today announced a definitive agreement under which NVIDIA will acquire Arm Limited from SBG and the SoftBank Vision Fund (together, “SoftBank”) in a transaction valued at $40 billion. The transaction is expected to be immediately accretive to NVIDIA’s non-GAAP gross margin and non-GAAP earnings per share.

The combination brings together NVIDIA’s leading AI computing platform with Arm’s vast ecosystem to create the premier computing company for the age of artificial intelligence, accelerating innovation while expanding into large, high-growth markets. SoftBank will remain committed to Arm’s long-term success through its ownership stake in NVIDIA, expected to be under 10 percent.

For more information, see NVIDIA to Acquire Arm for $40 Billion, Creating World’s Premier Computing Company for the Age of AI

Addressing The Challenges In Higher ED and Research

Date: Wednesday, June 17, 2020
Time: 11:00am – 12:00am SGT
Duration: 1 hour

Universities are undergoing an unprecedented challenge to provide staff to work from home, remote teaching and learning , and still provide high value learning to students and cutting edge tools and services to faculty and researchers. While remote learning is not a new phenomenon, providing quality service at scale is now a requirement, along with a new set of challenges that span user experience, mobility, effective management of a distributed deployment.

Solutions that enable remote learning and research, such as NVIDIA virtual GPU (vGPU) technology, enable you to meet these new requirements across various workloads with cost-effective solutions for existing on-premise infrastructure assets and in the cloud.

By attending this webinar, you’ll learn:
How NVIDIA vGPU technology solutions enable remote work and learning
How vGPU solutions are helping universities, across both education and research
How to get started with vGPU and vComputeServer to accelerate VDI and computational workloads in your institution