SEARCH SESSIONS

Search All
Refine Results:
 
Year(s)

SOCIAL MEDIA

EMAIL SUBSCRIPTION

 
 

GTC ON-DEMAND

Topic(s) Filter: Data Center & Cloud Infrastructure, GPU Virtualization
Presentation
Media
Opening Keynote (Keynote Talk)
Abstract:
Don't miss this keynote from NVIDIA Founder & CEO, Jensen Huang, as he speaks on the future of computing.
 
Topics:
Deep Learning & AI Frameworks, Intelligent Machines, IoT & Robotics, Autonomous Vehicles, Data Center & Cloud Infrastructure
Type:
Keynote
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9688
Streaming:
Share:
 
Abstract:
We'll discuss how Microsoft and NVIDIA have partnered to bring a broad portfolio of GPU products to Azure to support the demands of the most bleeding-edge customers. Our talk will cover how Azuer's industry-leading accelerator technology, delivered ...Read More
Abstract:
We'll discuss how Microsoft and NVIDIA have partnered to bring a broad portfolio of GPU products to Azure to support the demands of the most bleeding-edge customers. Our talk will cover how Azuer's industry-leading accelerator technology, delivered in multiple formats, puts demanding applications in an environment in which needed resources available on demand. From high performance networking and storage, to AI-aware cluster management and job orchestration tools, Azure takes the work out of running high-performance workloads.  Back
 
Topics:
Data Center & Cloud Infrastructure, GPU Virtualization
Type:
Sponsored Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S91017
Streaming:
Share:
 
Abstract:
Data centers today benefit from highly optimized hardware architectures and performance metrics that enable efficient provisioning and tuning of compute resources. But these architectures and metrics, honed over decades, are sternly challenged by the ...Read More
Abstract:
Data centers today benefit from highly optimized hardware architectures and performance metrics that enable efficient provisioning and tuning of compute resources. But these architectures and metrics, honed over decades, are sternly challenged by the rapid increase of AI applications and neural net workloads, where the impact of memory metrics like bandwidth, capacity, and latency on overall performance is not yet well understood. Get the perspectives of AI HW/SW co-design experts from Google, Microsoft, Facebook and Baidu, and technologists from NVIDIA and Samsung, as they evaluate the AI hardware challenges facing data centers and brainstorm current and necessary advances in architectures with particular emphasis on memory's impact on both training and inference.  Back
 
Topics:
Data Center & Cloud Infrastructure, Performance Optimization, Speech & Language Processing, HPC and AI, HPC and Supercomputing
Type:
Panel
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S91018
Streaming:
Download:
Share:
 
Abstract:
Are you doing data science at scale? Do you need a cluster of GPUs to help accelerate machine learning training and inferencing? Does your data pipeline extend beyond the walls of the data center to remote office, retail stores, or IoT sensors? We'l ...Read More
Abstract:
Are you doing data science at scale? Do you need a cluster of GPUs to help accelerate machine learning training and inferencing? Does your data pipeline extend beyond the walls of the data center to remote office, retail stores, or IoT sensors? We'll discuss how Cisco infrastructure can support data pipelines that extend from ends of the earth, to the data center, and even to the cloud with technologies such as NGC on Red Hat OpenShift, NGC on Hortonworks, and Kubeflow. Learn how infrastructure performance, scale, and flexibility can help accelerate, scale, and operationalize your data pipeline.  Back
 
Topics:
Data Center & Cloud Infrastructure, AI & Deep Learning Business Track (High Level)
Type:
Sponsored Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S91019
Streaming:
Download:
Share:
 
Abstract:
Migrating and building solutions in the cloud is challenging, expensive and not nearly as performant. Oracle Cloud Infrastructure (OCI) has been working with NVIDIA on giving you the on-premises performance you need with the cloud benefits and f ...Read More
Abstract:

Migrating and building solutions in the cloud is challenging, expensive and not nearly as performant. Oracle Cloud Infrastructure (OCI) has been working with NVIDIA on giving you the on-premises performance you need with the cloud benefits and flexibility you expect. In this session we'll discuss how you can take big data and analytics workloads, database workloads, or traditional enterprise HPC workloads that require multiple components along with a portfolio of accelerated hardware and not only migrate them to the cloud, but run them successfully. We'll discuss solution architectures, showcase demos, benchmarks and take you through the cloud migration journey. We'll detail the latest instances that OCI provides, along with cloud-scale services.

  Back
 
Topics:
Data Center & Cloud Infrastructure, Performance Optimization
Type:
Sponsored Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S91026
Streaming:
Download:
Share:
 
Abstract:
While every enterprise is on a mission to infuse its business with deep learning, few know how to build the infrastructure to get there. Short-sighted approaches to data center design can lead to long-term consequences that make the ROI of AI elusive ...Read More
Abstract:
While every enterprise is on a mission to infuse its business with deep learning, few know how to build the infrastructure to get there. Short-sighted approaches to data center design can lead to long-term consequences that make the ROI of AI elusive. We'll talk about the insights and best practices we at NVIDIA have gained from deep learning deployments around the globe and provide prescriptive guidance that every organization can leverage to shorten deployment timeframes, improve developer productivity, and streamline operations.  Back
 
Topics:
Data Center & Cloud Infrastructure, Deep Learning & AI Frameworks
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9120
Streaming:
Share:
 
Abstract:
We'll discuss Nebraska Medicine's transition to virtual desktops with vGPU and describe how that's enhanced user experience while reducing overhead on IT staff. We are also leveraging vGPUs for a better Windows 10 experience for users.
 
Topics:
GPU Virtualization, AI in Healthcare
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9162
Streaming:
Download:
Share:
 
Abstract:
Are you planning to deploy virtual desktops or in the middle of a struggling virtual desktop project? We were there three years ago. We lacked buy-in from executives, IT support, or users. If you are planning a virtual desktop infrastructure (VDI) de ...Read More
Abstract:
Are you planning to deploy virtual desktops or in the middle of a struggling virtual desktop project? We were there three years ago. We lacked buy-in from executives, IT support, or users. If you are planning a virtual desktop infrastructure (VDI) deployment or challenged by your current configuration, this session outlines the critical factors for success. We describe how to correct a languishing VDI implementation through engagement with leadership, VMware expertise, Agile scrum techniques, ITIL and project management best practices. Our talk will include technical details about our environment and how we achieved broad success.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9219
Streaming:
Download:
Share:
 
Abstract:
Learn about the importance of graphics support for unified communications such as Skype for business. We'll explain how using NVIDIA in the data center and IGEL at the endpoint can improve performance in this critical area for Windows 10 virtualizat ...Read More
Abstract:
Learn about the importance of graphics support for unified communications such as Skype for business. We'll explain how using NVIDIA in the data center and IGEL at the endpoint can improve performance in this critical area for Windows 10 virtualization. We'll describe how NVIDIA GPUs help organizations achieve the workstation performance levels that users want. Using case studies and examples, we'll show how the combination of IGEL and Citrix VDI with NVIDIA can also support high-latency and integrations with popular UCC applications, leading to improved communication and collaboration among geographically dispersed workforces.  Back
 
Topics:
GPU Virtualization, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9231
Streaming:
Download:
Share:
 
Abstract:
We'll describe how our startup, AITRICS, built time- and cost-efficient machine learning pipelines. We'll explain how to configure the clustering between physical servers and the on-demand cloud server with terraform and kubernetes. We'll highligh ...Read More
Abstract:
We'll describe how our startup, AITRICS, built time- and cost-efficient machine learning pipelines. We'll explain how to configure the clustering between physical servers and the on-demand cloud server with terraform and kubernetes. We'll highlight problems we're trying to solve such as sensitive data management, permission control, model evaluation, and versioning/serving. We will also cover our flexible, scalable training farms for researchers, developers, and collaborative research groups. These farms are also available for deploying trained models.  Back
 
Topics:
Data Center & Cloud Infrastructure, AI Application, Deployment & Inference
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9264
Streaming:
Download:
Share:
 
Abstract:
Learn how to build a data-centric GPU cluster for artificial intelligence. Mellanox is a leader in high-performance, scalable, low-latency network interconnects for both InfiniBand and Ethernet. We will briefly present state-of-the-art techniques for ...Read More
Abstract:
Learn how to build a data-centric GPU cluster for artificial intelligence. Mellanox is a leader in high-performance, scalable, low-latency network interconnects for both InfiniBand and Ethernet. We will briefly present state-of-the-art techniques for distributed machine learning and examine what special requirements these techniques impose on the system. We'll also give an overview of interconnect technologies used to scale and accelerate distributed machine learning including RDMA, NVIDIA's GPUDirect technology, and in-network computing that accelerates large-scale deployments in HPC and artificial intelligence.  Back
 
Topics:
Data Center & Cloud Infrastructure, Deep Learning & AI Frameworks, HPC and Supercomputing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9268
Streaming:
Share:
 
Abstract:
Learn how to run GPU workloads securely in isolated unprivileged containers across a multi-node LXD cluster. We'll explain what unprivileged containers are and why they're safe, and then use demos to show how the power of LXD can be used ...Read More
Abstract:

Learn how to run GPU workloads securely in isolated unprivileged containers across a multi-node LXD cluster. We'll explain what unprivileged containers are and why they're safe, and then use demos to show how the power of LXD can be used to create a whole cluster of unprivileged containers that are isolated from each other. We will also show how LXD makes it trivial to pass through physical GPUs to containers and how it exposes a wide range of NVIDIA-specific options by leveraging NVIDIA's libnvidia-container library. This way each container can easily get a dedicated GPU or GPUs to run workloads. This session will help participants understand how running GPU-Intensive workloads can be effortless with the help of a dedicated container manager that is aware of NVIDIA-specific features.

  Back
 
Topics:
Data Center & Cloud Infrastructure, HPC and Supercomputing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9274
Streaming:
Download:
Share:
 
Abstract:
We'll describe a lightweight GPU counter monitoring tool called GPUPerf that our Alibaba team developed with NVIDIA. It monitors GPU context create and destroy, and records GPU internal counter values, such as active/elapsed cycles, IPC, and memory ...Read More
Abstract:
We'll describe a lightweight GPU counter monitoring tool called GPUPerf that our Alibaba team developed with NVIDIA. It monitors GPU context create and destroy, and records GPU internal counter values, such as active/elapsed cycles, IPC, and memory access bandwidth with little overhead. We'll discuss how we deployed this tool in one of our lab clusters to do real-time monitoring. Combined with information collected from NVIDIA-smi, we now understand our GPU server workload much better. We'll also explain how GPUPerf helps improve GPU cluster orchestra and scheduling algorithms.  Back
 
Topics:
Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9285
Streaming:
Download:
Share:
 
Abstract:
Red Hat and NVIDIA collaborated to bring together two of the technology industry's most popular products: Red Hat Enterprise Linux 7 and the NVIDIA DGX system. This talk will cover how the combination of RHELs rock-solid stability with the incredibl ...Read More
Abstract:
Red Hat and NVIDIA collaborated to bring together two of the technology industry's most popular products: Red Hat Enterprise Linux 7 and the NVIDIA DGX system. This talk will cover how the combination of RHELs rock-solid stability with the incredible DGX hardware can deliver tremendous value to enterprise data scientists. We will also show how to leverage NVIDIA GPU Cloud container images with Kubernetes and RHEL to reap maximum benefits from this incredible hardware.  Back
 
Topics:
Data Center & Cloud Infrastructure, Finance - Quantitative Risk & Derivative Calculations
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9292
Streaming:
Download:
Share:
 
Abstract:
Red Hat Virtualization is an open platform that is built on Kernel-based Virtual Machine (KVM), one of several hypervisors supporting NVIDIA vGPU integration. Learn about RHV installation, the NVIDIA vGPU host driver, deployment of guest VMs with sin ...Read More
Abstract:
Red Hat Virtualization is an open platform that is built on Kernel-based Virtual Machine (KVM), one of several hypervisors supporting NVIDIA vGPU integration. Learn about RHV installation, the NVIDIA vGPU host driver, deployment of guest VMs with single and multiple vGPU enablement, as well as NVIDIA GRID license manager.   Back
 
Topics:
GPU Virtualization, AI in Healthcare
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9299
Streaming:
Download:
Share:
 
Abstract:
Do you have a GPU cluster or air-gapped environment that you are responsible for but don't have an HPC background?   NVIDIA DGX POD is a new way of thinking about AI infrastructure, combining DGX servers with networking and storage to a ...Read More
Abstract:

Do you have a GPU cluster or air-gapped environment that you are responsible for but don't have an HPC background?   NVIDIA DGX POD is a new way of thinking about AI infrastructure, combining DGX servers with networking and storage to accelerate AI workflow deployment and time to insight. We'll discuss lessons learned about building, deploying, and managing AI infrastructure at scale from design to deployment to management and monitoring.   We will show how the DGX Pod Management software (DeepOps) along with our storage partner reference-architectures can be used for the deployment and management of multi-node GPU clusters for Deep Learning and HPC environments, in an on-premise, optionally air-gapped datacenter. The modular nature of the software also allows experienced administrators to pick and choose items that may be useful, making the process compatible with their existing software or infrastructure.  

  Back
 
Topics:
Data Center & Cloud Infrastructure, AI Application, Deployment & Inference
Type:
Tutorial
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9334
Streaming:
Download:
Share:
 
Abstract:
Learn our solutions for increasing GPU resource utilization on an on-premise DGX-2 node and public clouds. In this talk we present our operational experiences of a set of multi-tenant deep learning workloads selected through an open competition. To h ...Read More
Abstract:
Learn our solutions for increasing GPU resource utilization on an on-premise DGX-2 node and public clouds. In this talk we present our operational experiences of a set of multi-tenant deep learning workloads selected through an open competition. To host them we use and extend the Backend.AI framework as the resource and computation manager. While tailored for both educational and research-oriented workloads, it offers a topology-aware multi-GPU resource scheduler combined with fractional GPU scaling implemented via API-level CUDA virtualization, achieving higher GPU utilization compared to vanilla setups.  Back
 
Topics:
Data Center & Cloud Infrastructure, GPU Virtualization, Deep Learning & AI Frameworks
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9406
Download:
Share:
 
Abstract:
vMotion is a feature in VMware vSphere that is used to guarantee continuous availability and uptime in the face of planned outages and maintenance operations. It's now available for virtual machines (VMs) with an NVIDIA GRID GPU. We'll present an o ...Read More
Abstract:
vMotion is a feature in VMware vSphere that is used to guarantee continuous availability and uptime in the face of planned outages and maintenance operations. It's now available for virtual machines (VMs) with an NVIDIA GRID GPU. We'll present an overview of the architecture of vMotion for vGPU-enabled VMs and discuss how it affects performance. We'll also outline potential vMotion uses for load balancing on vGPU-enabled clouds.  Back
 
Topics:
Data Center & Cloud Infrastructure, AI Application, Deployment & Inference
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9411
Streaming:
Download:
Share:
 
Abstract:
Video- and audio-based applications now comprise about 80 percent of Internet traffic, but their quality depends on the network condition. Cloud providers must accurately quantify and monitor video and audio quality so they can maintain the quality o ...Read More
Abstract:
Video- and audio-based applications now comprise about 80 percent of Internet traffic, but their quality depends on the network condition. Cloud providers must accurately quantify and monitor video and audio quality so they can maintain the quality of these applications while optimizing cloud resource usage. We'll describe our solution, which uses deep neural networks to measure the quality of video and audio, and demonstrate how we measure the quality of streaming video and audio using VMware Horizon virtual desktops. We'll present our research results showing the capabilities of the latest NVIDIA Pascal and Volta GPUs and NVIDIA GRID on VMware vSphere to accelerate the deep learning-based measurement task for our large-scale performance monitoring need. We will also cover the benefits of using NVIDIA GRID to improve the performance of our application without changing the number of GPUs in the system.  Back
 
Topics:
Data Center & Cloud Infrastructure, AI Application, Deployment & Inference, Accelerated Data Science
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9435
Streaming:
Download:
Share:
 
Abstract:
Our talk will describe the business and technical challenges posed by subsurface exploration for oil and gas and how the high-performance RiVA computing platform addresses these challenges. Oil and gas companies have struggled to find experts to ...Read More
Abstract:

Our talk will describe the business and technical challenges posed by subsurface exploration for oil and gas and how the high-performance RiVA computing platform addresses these challenges. Oil and gas companies have struggled to find experts to manage the complex systems needed for subsurface exploration. In addition, the large data sets these engineers require often take more than eight hours to load. We'll discuss RiVA, which was built to address problems with slow data transfers, and describe how it offers performance 30 times faster than other solutions and reduces deployment time from years to months. We'll cover the technologies that make our solution possible, including NVIDIA GPUs, Mechdyne TGX, and the Leostream Connection Broker. In addition, a RiVA customer will share challenges and show how deploying RiVA helped lower costs during deployment and production.

  Back
 
Topics:
GPU Virtualization, Data Center & Cloud Infrastructure, Seismic & Geosciences
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9454
Streaming:
Download:
Share:
 
Abstract:
NVIDIA GPU Cloud is a single source for researchers and developers seeking access to GPU optimized deep learning framework containers for TensorFlow, PyTorch, and MXNet.  Well cover the latest NVIDIA features integrated into these popular f ...Read More
Abstract:

NVIDIA GPU Cloud is a single source for researchers and developers seeking access to GPU optimized deep learning framework containers for TensorFlow, PyTorch, and MXNet.  Well cover the latest NVIDIA features integrated into these popular frameworks, the benefits of using them through NGC monthly container updates, and tips and tricks to maximize performance on NVIDIA GPUs for your deep learning workloads. Well dive into the anatomy of a deep learning container, breaking down the software that makes up the container, and present the optimizations we have implemented to get the most out of NVIDIA GPUs. For both new and experienced users of our deep learning framework containers, this session will provide valuable insight into the benefits of NVIDIA accelerated frameworks available as easy pull and run containers.

  Back
 
Topics:
Data Center & Cloud Infrastructure, AI & Deep Learning Research
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9500
Streaming:
Download:
Share:
 
Abstract:
Whether it's for AI, data science and analytics, or HPC, GPU-Accelerated software can make possible the previously impossible. But it's well known that these cutting edge software tools are often complex to use, hard to manage, and diffi ...Read More
Abstract:

Whether it's for AI, data science and analytics, or HPC, GPU-Accelerated software can make possible the previously impossible. But it's well known that these cutting edge software tools are often complex to use, hard to manage, and difficult to deploy. We'll exlain how NGC solves these problems and gives users a head start on their projects by simplifying the use of GPU-Optimized software. NVIDIA product management and engineering experts will walk through the latest enhancements to NGC and give examples of how software from NGC can improve GPU-accelerated workflows.

  Back
 
Topics:
Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9504
Streaming:
Download:
Share:
 
Abstract:
NVIDIA offers several containerized applications in HPC, visualization, and deep learning. We have also enabled a broad array of contain-related technologies for GPUs with upstreamed improvements to community projects and with tools that are seeing b ...Read More
Abstract:
NVIDIA offers several containerized applications in HPC, visualization, and deep learning. We have also enabled a broad array of contain-related technologies for GPUs with upstreamed improvements to community projects and with tools that are seeing broad interest and adoption. In addition, NVIDIA is a catalyst for the broader community in enumerating key technical challenges for developers, admins and end users, and is helping to identify gaps and drive them to closure. Our talk describes NVIDIA's new developments and upcoming efforts. We'll detail progress in the most important technical areas, including multi-node containers, security, and scheduling frameworks. We'll also offer highlights of the breadth and depth of interactions across the HPC community that are making the latest, highly-quality HPC applications available to platforms that include GPUs.  Back
 
Topics:
Data Center & Cloud Infrastructure, HPC and AI
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9525
Streaming:
Download:
Share:
 
Abstract:
We'll describe a new benchmark suite proposed by the Deep Learning community for machine learning workloads. We'll present a quantitative analysis of an early version (0.5) of benchmark known as MLPerf and explain performance impact of NVIDIA GPU a ...Read More
Abstract:
We'll describe a new benchmark suite proposed by the Deep Learning community for machine learning workloads. We'll present a quantitative analysis of an early version (0.5) of benchmark known as MLPerf and explain performance impact of NVIDIA GPU architecture across a range of DL applications. This work includes evaluating MLPerf performance on Turing, Volta, and Pascal to demonstrate the performance impact of NVIDIA GPU architecture across a range of DL applications. We'll evaluate the impact of system-level technologies Nvlink vs. PCIe topology using server- and workstation-class platforms to show how system architecture impacts DL training workloads. We also plan to discuss our work to characterize MLPerf benchmark performance using profiling tools (GPU, CPU, memory & I/O), our hyperparameter tuning study (batch size, learning rate, SGD optimizer) on MLPerf performance, and map real world application use cases to MLPerf suite and how to quantify results for specific DL practitioner use cases.  Back
 
Topics:
Data Center & Cloud Infrastructure, Performance Optimization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9553
Streaming:
Download:
Share:
 
Abstract:
Learn how to create distributed, composable resource pools for your on-prem or bare metal Kubernetes-based system. This helps developers determine the most effective use of resources like GPUs, networking, and storage, and enables a dynamically balan ...Read More
Abstract:
Learn how to create distributed, composable resource pools for your on-prem or bare metal Kubernetes-based system. This helps developers determine the most effective use of resources like GPUs, networking, and storage, and enables a dynamically balanced system in which compute power is available where it's needed. We'll describe a heterogeneous, distributed architecture based on Kubernetes that takes system resources from different system nodes in a cluster and gathers them in resource pools. The architecture partitions the resources into pools, which are securely and dynamically allocated to groups of users. Devices within the pools are assigned on demand and released back into the pool when done. We'll provide an overview of how containerized applications operate, including the challenges of distributed composability in these systems. We'll also explain how to effectively allocate and share scarce, expensive devices among competing uses.  Back
 
Topics:
Data Center & Cloud Infrastructure, Deep Learning & AI Frameworks
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9572
Streaming:
Download:
Share:
 
Abstract:
We'll discuss how ImageNet can be trained from scratch to standard benchmark accuracy in 19 minutes on 64 NVIDIA Tesla V100 GPUs by relying solely on open-source tools and publicly available cloud infrastructure. Although previous efforts in this ar ...Read More
Abstract:
We'll discuss how ImageNet can be trained from scratch to standard benchmark accuracy in 19 minutes on 64 NVIDIA Tesla V100 GPUs by relying solely on open-source tools and publicly available cloud infrastructure. Although previous efforts in this area used dedicated cluster hardware and specialized Internet adapters, our implementation performs well over Ethernet in a shared cloud environment. We'll walk through algorithmic improvements needed to achieve this result and lessons we learned in the process.  Back
 
Topics:
Data Center & Cloud Infrastructure, Deep Learning & AI Frameworks, Computer Vision
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9573
Streaming:
Download:
Share:
 
Abstract:
Learn about the fundamental expectations, planning, and implementation considerations for a VDI by Day/Compute by Night environment. We'll present a basic architecture featuring GPU virtualization and show how to apply the architecture to multiple d ...Read More
Abstract:
Learn about the fundamental expectations, planning, and implementation considerations for a VDI by Day/Compute by Night environment. We'll present a basic architecture featuring GPU virtualization and show how to apply the architecture to multiple domain types.  Back
 
Topics:
GPU Virtualization, Computer Aided Engineering
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9670
Streaming:
Download:
Share:
 
Abstract:
Learn how to efficiently share GPUs, NVMe drives, and other IO devices in a PCI Express network using Dolphin Interconnect Solutions' SmartIO. We'll explain how we eliminated restrictions of traditional IO devices, which are statically assigned to ...Read More
Abstract:
Learn how to efficiently share GPUs, NVMe drives, and other IO devices in a PCI Express network using Dolphin Interconnect Solutions' SmartIO. We'll explain how we eliminated restrictions of traditional IO devices, which are statically assigned to a single root complex (host machine) and lack flexible support for features such as hot-add, device migration, and remote access without complex software frameworks. We'll highlight how SmartIO provides a framework that allows PCIe devices and systems to use remote systems to flexibly access devices such as GPUs, NVMe drives, and other IO devices. We will demonstrate how we implemented SmartIO using standard PCIe and non-transparent bridging and show how our system got near-native performance when moving data borrowed GPUs and NVMe drives. We'll also explore how we can dynamically add more GPUs to scale performance.  Back
 
Topics:
Data Center & Cloud Infrastructure, HPC and Supercomputing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9709
Streaming:
Download:
Share:
 
Abstract:
Learn how to distribute GPU workloads on the intelligent edge with Verizon. We'll describe how Verizon engineers developed GPU virtualization and management platforms to accelerate graphics, rendering, audio, and computer vision with edge networks. ...Read More
Abstract:
Learn how to distribute GPU workloads on the intelligent edge with Verizon. We'll describe how Verizon engineers developed GPU virtualization and management platforms to accelerate graphics, rendering, audio, and computer vision with edge networks. We'll share recent developments in the journey to 5G and the edge by describing cases studies and demonstrating edge-based APIs for rendering and inferencing.  Back
 
Topics:
GPU Virtualization, Virtual Reality & Augmented Reality, 5G & Edge, Gaming and AI
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9741
Streaming:
Download:
Share:
 
Abstract:
We'll describe how large data scale (over two millennia of speech data per year) and low-latency requirements have enabled and required novel approaches to several speech and language models. Our talk will cover the GPU speech recognition training p ...Read More
Abstract:
We'll describe how large data scale (over two millennia of speech data per year) and low-latency requirements have enabled and required novel approaches to several speech and language models. Our talk will cover the GPU speech recognition training pipeline, continuous feedback-based training, optimizations for training, and inference on TensorRT for ultra- low latency text-to-speech models for call centers. We will discuss accuracy and latency benchmarks for speech recognition on conversational speech, speech synthesis, data-driven dialogue systems, emotion recognition, and speech act classification. We'll also demonstrate our system running on a scaled simulated call center and show live speech recognition, synthesis, and language processing.  Back
 
Topics:
Data Center & Cloud Infrastructure, AI in Healthcare, Medical Imaging & Radiology
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9776
Streaming:
Download:
Share:
 
Abstract:
To meet the needs of today's distributed teams, many AEC firms are turning to virtualization to efficiently and securely run applications such as Autodesk Revit, AutoCAD, and Sketchup. Learn how this panel of AEC experts takes advantage of the lates ...Read More
Abstract:
To meet the needs of today's distributed teams, many AEC firms are turning to virtualization to efficiently and securely run applications such as Autodesk Revit, AutoCAD, and Sketchup. Learn how this panel of AEC experts takes advantage of the latest innovations of NVIDIA vGPU technology to exponentially speed rendering time, enable greater mobility, optimize resource consumption, reduce maintenance and management work, and increase productivity.  Back
 
Topics:
GPU Virtualization
Type:
Panel
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9814
Streaming:
Download:
Share:
 
Abstract:
End-to-end machine learning workloads perform well using NVIDIA virtual GPUs in VMware vSphere. We'll discuss how to combine the performance of NVIDIA GPUs with manageability and scalability features and maximize GPU utilization for machine learning ...Read More
Abstract:
End-to-end machine learning workloads perform well using NVIDIA virtual GPUs in VMware vSphere. We'll discuss how to combine the performance of NVIDIA GPUs with manageability and scalability features and maximize GPU utilization for machine learning workloads using VMware and NVIDIA technology. We will outline end-to-end machine learning, including training, deploying for inferencing, and managing a production environment using VMware vSphere and VMware's Pivotal Kubernetes Service. NVIDIA Turing architecture is positioned for mixed-precision training and inferencing workloads. We'll describe ways to deploy GPU-Based workloads developed with machine learning frameworks like TensorFlow and Caffe2 by using VMware DirectPathIO and NVIDIA virtual GPU (vGPU). We'll also provide case studies that leverage vGPU scheduling options such as Equal Share, Fixed Share, and Best Effort, and illustrate their benefits with our performance study.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9815
Streaming:
Download:
Share:
 
Abstract:
The hardest part of cloud computing engineering is operations because of the complexity of managing thousands of machines, but machine learning can add intelligence to public cloud operation and maintenance. We use RAPIDS to accelerate machine l ...Read More
Abstract:

The hardest part of cloud computing engineering is operations because of the complexity of managing thousands of machines, but machine learning can add intelligence to public cloud operation and maintenance. We use RAPIDS to accelerate machine learning and the NVIDIA TensorRT inference server for GPU load balancing and improved GPU utilization. We'll explain how to use traditional machine learning algorithms such as ARIMA, XGBoost, and RandomForest for load prediction, load classification, user portrait, exception prediction, and other scenarios. Learn how to use GPUs for data preprocessing and algorithm acceleration for large-scale data analysis and machine learning of massive public cloud data. In addition, we'll cover how we implemented a large-scale training and prediction service platform based on Dask and NVIDIA's inference server. The platform can support large-scale GPU parallel computing and prediction requests.

  Back
 
Topics:
Data Center & Cloud Infrastructure, Accelerated Data Science
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9845
Streaming:
Download:
Share:
 
Abstract:
VDI users across multiple industries can now harness the power of the world's most advanced virtual workstation to enable increasingly demanding workflows. This session brings together graphics virtualization thought leaders and experts from ...Read More
Abstract:

VDI users across multiple industries can now harness the power of the world's most advanced virtual workstation to enable increasingly demanding workflows. This session brings together graphics virtualization thought leaders and experts from across the globe who have deep knowledge of NVIDIA virtual GPU architecture and years of experience implementing VDI across multiple hypervisors. Panelists will discuss how they transformed organizations, including how they leveraged multi-GPU support to boost GPU horsepower for photorealistic rendering and data-intensive simulation and how they used GPU-Accelerated deep learning or HPC VDI environments with ease using NGC containers.

  Back
 
Topics:
GPU Virtualization
Type:
Panel
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9870
Streaming:
Download:
Share:
 
Abstract:
The industry-standard graphics performance benchmark tool, SPECviewperf 13, provides valuable insight on GRID profile and Tesla graphic card selection for virtualizing high-performance graphics applications. Learn how Cisco technical marketing engine ...Read More
Abstract:
The industry-standard graphics performance benchmark tool, SPECviewperf 13, provides valuable insight on GRID profile and Tesla graphic card selection for virtualizing high-performance graphics applications. Learn how Cisco technical marketing engineers evaluate performance of converged and hyperconverged platforms with the full spectrum of NVIDIA Tesla GPUs on eight key applications on the two leading desktop broker vendors. We will share our methodology for scoring grid profile/Tesla GPU/server hardware platform combinations for each application in the performance tool. We'll also present our sizing recommendations for these widely used products based on the application user type.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9881
Streaming:
Download:
Share:
 
Abstract:
Learn how to access a virtual workstation in five minutes and see demos of NVIDIA RTX platform, the latest in rendering technology, speeds up design and visualization workloads. We'll share customer examples and best practices for deploying and mana ...Read More
Abstract:
Learn how to access a virtual workstation in five minutes and see demos of NVIDIA RTX platform, the latest in rendering technology, speeds up design and visualization workloads. We'll share customer examples and best practices for deploying and managing cloud based virtual workstations. We'll also discuss rising doption of GPU-Accelerated cloud computing and touch on how NVIDIA Quadro Virtual Workstations running on NVIDIA's Pascal and Turing architectures deliver the workstation performance expected by professionals with the flexibility of the cloud.  Back
 
Topics:
GPU Virtualization, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9882
Streaming:
Download:
Share:
 
Abstract:
With the growth in demand of Intelligent Video Analytics (IVA), NVIDIA virtual GPUs provides a secure solution while optimizing GPU utilization for inference-based deep learning applications for loss prevention, facial recognition, pose estimati ...Read More
Abstract:

With the growth in demand of Intelligent Video Analytics (IVA), NVIDIA virtual GPUs provides a secure solution while optimizing GPU utilization for inference-based deep learning applications for loss prevention, facial recognition, pose estimation, and many other use cases.

  Back
 
Topics:
GPU Virtualization, Consumer Engagement & Personalization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9883
Streaming:
Download:
Share:
 
Abstract:
The latest operating systems such as Windows 10 or Server 2016 include graphically rich features that require GPUs. Applications that run on these operating systems contribute to GPU consumption. We'll describe how GPUs are needed for traditional VD ...Read More
Abstract:
The latest operating systems such as Windows 10 or Server 2016 include graphically rich features that require GPUs. Applications that run on these operating systems contribute to GPU consumption. We'll describe how GPUs are needed for traditional VDI or any cloud-enabled desktop deployment to provide a best-in-class user experience. We'll discuss how we're working to answer questions about how to measure user experience, find the perfect protocol, figure out the ideal combination of remote protocol policies, and determine the best codec to fit each use case.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9884
Streaming:
Download:
Share:
 
Abstract:
NVIDIA virtual GPU (vGPU) technology transformed VDI by delivering a GPU-accelerated user experience - indistinguishable from a physical desktop. Innovations in NVIDIA vGPU technology enable IT to virtualize a GPU to be shared amongst multiple ...Read More
Abstract:
NVIDIA virtual GPU (vGPU) technology transformed VDI by delivering a GPU-accelerated user experience - indistinguishable from a physical desktop. Innovations in NVIDIA vGPU technology enable IT to virtualize a GPU to be shared amongst multiple users, or enable a single user to harness the power of multiple GPUs in a VM. It can deliver powerful Quadro virtual workstations capable of running the most demanding applications, and more easily manage the VDI environment with GPU live migration. Virtualization increases data center efficiency  - enabling multiple workloads on a pool of graphics accelerated compute resources with NVIDIA T4. Join us to hear about the latest enhancements in NVIDIA vGPU.  In this session, you will hear directly from the vGPU product leadership team.  The discussion will include the latest vGPU release, RTX Server and the latest news on global cloud offerings.  We’ll also review how NVIDIA’s vGPU technology is driving innovation and new use cases.  Now you can access the power of an NVIDIA GPU from anywhere you are – from the data center, in the cloud, to any endpoint you choose. Learn about the NVIDIA virtual GPU technology that take your VDI to the next level.  
 
  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9885
Streaming:
Download:
Share:
 
Abstract:
Learn how HPE and NVIDIA are simplifying infrastructure and delivering extreme graphics and performance on the HPE SimpliVity HCI platform. We'll talk about EUC offerings and use cases for HPE SimpliVity with NVIDIA GPUs and highlight perfor ...Read More
Abstract:

Learn how HPE and NVIDIA are simplifying infrastructure and delivering extreme graphics and performance on the HPE SimpliVity HCI platform. We'll talk about EUC offerings and use cases for HPE SimpliVity with NVIDIA GPUs and highlight performance metrics achieved through industry-standard benchmarks.

  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9886
Streaming:
Download:
Share:
 
Abstract:
Modern workers expect a satisfying experience over any network from any device. IT admins want the flexibility to deliver workloads from their hypervisor or cloud of choice. And modern applications like Windows 10 are more graphic-intensive than ever ...Read More
Abstract:
Modern workers expect a satisfying experience over any network from any device. IT admins want the flexibility to deliver workloads from their hypervisor or cloud of choice. And modern applications like Windows 10 are more graphic-intensive than ever. Learn how Citrix and NVIDIA provide a superior a customer experience for graphic-accelerated virtual apps and desktops with true hypervisor and cloud flexibility. We'll discuss our latest innovations in graphics virtualization and describe other Citrix HDX innovations that enhance graphics remoting and optimize user experience.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9887
Streaming:
Download:
Share:
 
Abstract:
Universities have increasing demand for Deep Learning/AI classrooms or labs but are constrained by cost and availability of physical classroom labs. Students require access to a lab 24x7 to work on projects and assignments and find that they have to ...Read More
Abstract:
Universities have increasing demand for Deep Learning/AI classrooms or labs but are constrained by cost and availability of physical classroom labs. Students require access to a lab 24x7 to work on projects and assignments and find that they have to wait for HPC clusters to be free when submitting their jobs for training. In the past, students and researchers are tethered and require expensive data scientist workstations. Virtual GPUs provide a highly secure, flexible, accessible solution to power AI and deep learning coursework and research. Learn how Nanjing University is using virtual vGPUs with NGC for teaching AI and Deep learning courses, empowering researchers with the GPU power they need, and providing students with mobility to do coursework anywhere. Similarly, discover how other universities are maximizing their data center resources by running VDI, HPC and AI workloads on common infrastructure and even how companies like Esri are using virtualized deep learning classes to educate their user base. Discover the benefits of vGPUs for AI and how you can setup your environment to achieve optimum performance, as well as the tools you can use to manage and monitor your environment as you scale.  Back
 
Topics:
GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9888
Streaming:
Download:
Share:
 
Abstract:
We'll talk about how we achieved lower costs, better density, and guaranteed performance while migrating from legacy large-scale GPU passthrough VDI architecture to the latest NVIDIA vGPU solution. We'll also discuss our work delivering ...Read More
Abstract:

We'll talk about how we achieved lower costs, better density, and guaranteed performance while migrating from legacy large-scale GPU passthrough VDI architecture to the latest NVIDIA vGPU solution. We'll also discuss our work delivering remote workstations to subcontractors through our VDI solution for 6,000 concurrent users since 2014.

  Back
 
Topics:
GPU Virtualization, Product & Building Design
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9889
Streaming:
Download:
Share:
 
Abstract:
NVIDIA's virtual GPU software with Nutanix AHV is focused on end-user environments to meet the growing demand for GPU-Accelerated graphics. But this demand could lead to an oversubscribed private cloud because it blurs the lines between graphics and ...Read More
Abstract:
NVIDIA's virtual GPU software with Nutanix AHV is focused on end-user environments to meet the growing demand for GPU-Accelerated graphics. But this demand could lead to an oversubscribed private cloud because it blurs the lines between graphics and compute and breaks down infrastructure silos. We'll discuss how the ability to run VDI by day, compute by night helps achieve high utilization and ROI for large capital investments that would be underused when users go home at the end of the workday. Coupled with vGPU support for Tesla, this allows ML/AI workloads to leverage capital investment to drive better ROI and bring operational benefits of virtualization workloads to applications that have traditionally run bare-metal. We'll cover work done to make GPU configuration extremely simple and build general purpose infrastructure pods that offer high-performance storage on demand and improve business agility.  Back
 
Topics:
GPU Virtualization, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9890
Streaming:
Download:
Share:
 
Abstract:
Learn how to deploy deep learning applications for multi-tenant environments based on KVM. These virtual machines (VM) can be created with simple commands and are tuned for optimal DL performance leveraging underlying NVSwitches, NVLINKs, and NVIDIA ...Read More
Abstract:
Learn how to deploy deep learning applications for multi-tenant environments based on KVM. These virtual machines (VM) can be created with simple commands and are tuned for optimal DL performance leveraging underlying NVSwitches, NVLINKs, and NVIDIA GPUs. We'll show examples for creating, launching, and managing multiple GPU VMs.  Back
 
Topics:
Data Center & Cloud Infrastructure, GPU Virtualization
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9893
Streaming:
Download:
Share:
 
Abstract:
Users of HPC Systems have diverse needs and requirements for their applications and ML/DL environments. Containers help streamline and simplify environment creation, but security concerns generally prohibit popular container environments such as Dock ...Read More
Abstract:
Users of HPC Systems have diverse needs and requirements for their applications and ML/DL environments. Containers help streamline and simplify environment creation, but security concerns generally prohibit popular container environments such as Docker from running in shared computing environments. Alternate container systems for HPC address security concerns but have less documentation and resources available for users. We'll describe how our pipeline and resources at MITRE enable users to quickly build custom environments and run their code on the HPC system while minimizing startup time. Our process implements LXD containers, Docker, and Singularity on a combination of development and production HPC systems using a traditional scheduler.  Back
 
Topics:
Data Center & Cloud Infrastructure, HPC and AI, HPC and Supercomputing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9958
Streaming:
Download:
Share:
 
Abstract:
This customer panel brings together AI implementers who have deployed deep learning at scale. The discussion will focus on specific technical challenges they faced, solution design considerations, and best practices learned from implementing the ...Read More
Abstract:

This customer panel brings together AI implementers who have deployed deep learning at scale. The discussion will focus on specific technical challenges they faced, solution design considerations, and best practices learned from implementing their respective solutions.

  Back
 
Topics:
AI & Deep Learning Business Track (High Level), Data Center & Cloud Infrastructure, Deep Learning & AI Frameworks
Type:
Panel
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9121
Streaming:
Share:
 
Abstract:
We will present NVIDIA's solution for interactive, real-time streaming of VR content (such as games and professional applications) from the cloud to a low-powered client driving a VR/AR headset. We will outline few of the challenge ...Read More
Abstract:

We will present NVIDIA's solution for interactive, real-time streaming of VR content (such as games and professional applications) from the cloud to a low-powered client driving a VR/AR headset. We will outline few of the challenges, describe our  design, and share some performance and quality metrics. 

  Back
 
Topics:
Virtual Reality & Augmented Reality, Data Center & Cloud Infrastructure, Video & Image Processing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9156
Streaming:
Download:
Share:
 
Abstract:
Learn how CannonDesign has incorporated NVIDIA's RTX technology into their visualization workflows. During this presentation, we will discuss how CannonDesign is leveraging the power of the new Quadro RTX video cards to optimize rendering ti ...Read More
Abstract:

Learn how CannonDesign has incorporated NVIDIA's RTX technology into their visualization workflows. During this presentation, we will discuss how CannonDesign is leveraging the power of the new Quadro RTX video cards to optimize rendering times using VRay Next and Unreal Engine. We will share our evolutionary path to better rendering solutions, initial challenges with RTX and current workflow through case studies. This session will be of interest to attendees with basic understanding of visualization workflows.

  Back
 
Topics:
Product & Building Design, GPU Virtualization, Rendering & Ray Tracing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9184
Streaming:
Download:
Share:
 
Abstract:
As the use of AI has increased, so has the need for a production-quality AI inference solution. We'll discuss the latest additions to NVIDIA's TensorRT Inference Server and describe deployment examples to help plan your data center production infer ...Read More
Abstract:
As the use of AI has increased, so has the need for a production-quality AI inference solution. We'll discuss the latest additions to NVIDIA's TensorRT Inference Server and describe deployment examples to help plan your data center production inference architecture. NVIDIA TensorRT Inference Server makes it possible to efficiently leverage inference in applications and to do so without reinventing the wheel. We'll talk about how TensorRT supports the top AI frameworks and custom backends, and maximizes utilization by hosting multiple models per GPU and across GPUs with dynamic request batching. Our talk will also cover how the inference server seamlessly supports Kubernetes with health and latency metrics and integrates with Kubeflow for simplified deployment.   Back
 
Topics:
AI Application, Deployment & Inference, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9438
Streaming:
Download:
Share:
 
Abstract:
Building machine learning pipelines is challenging. Doing that in a portable way that supports multi-cloud deployments is even harder. We'll discuss the open source project, Kubeflow, which is designed to allow data scientists and machine learning e ...Read More
Abstract:
Building machine learning pipelines is challenging. Doing that in a portable way that supports multi-cloud deployments is even harder. We'll discuss the open source project, Kubeflow, which is designed to allow data scientists and machine learning engineers to focus on building great ML solutions instead of setting up and managing the infrastructure. We'll detail the latest version of Kubeflow and its integration with TensorRT, the inference server from NVIDIA.  Back
 
Topics:
AI Application, Deployment & Inference, Data Center & Cloud Infrastructure, Tools & Libraries
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9456
Streaming:
Download:
Share:
 
Abstract:
MATLAB's deep learning, visualization, and C++/CUDA code generation technology make it a uniquely complete solution for your entire AI workflow. In MATLAB, you can easily manage data, perform complex image and signal processing, prototype and train ...Read More
Abstract:
MATLAB's deep learning, visualization, and C++/CUDA code generation technology make it a uniquely complete solution for your entire AI workflow. In MATLAB, you can easily manage data, perform complex image and signal processing, prototype and train deep networks, and deploy to your desktop, embedded or cloud environments. Using GPU Coder technology MATLAB generates CUDA kernels that optimize loops and memory access, and C++ that leverages cuDNN and TensorRT, providing the fastest deep network inference of any framework. With MATLAB's NVIDIA docker container available through the NVIDIA GPU Cloud, you can now easily access all this AI power, deploy it in your cloud or DGX environment, and get up and running in seconds. In this presentation we will demonstrate a complete end-to-end workflow that starts from 'docker run', prototypes and trains a network on a multi-GPU machine in the cloud, and ends with a highly optimized inference engine to deploy to data centers, clouds, and embedded devices.  Back
 
Topics:
AI & Deep Learning Research, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9469
Streaming:
Download:
Share:
 
Abstract:
Learn from NVIDIA customers who will share their best practices for extending AI compute power to their teams without the need to build and manage a data center. These organizations will describe innovative approaches that let them turn an NVIDIA DGX ...Read More
Abstract:
Learn from NVIDIA customers who will share their best practices for extending AI compute power to their teams without the need to build and manage a data center. These organizations will describe innovative approaches that let them turn an NVIDIA DGX Station into a powerful solution serving entire teams of developers from the convenience of an office environment. Learn how teams building powerful AI applications may not need to own servers or depend on data center access and find out how to take advantage of containers, orchestration, monitoring, and scheduling tools. The organizations will also show demos of how to set up an AI work group with ease and cover best practices for AI developer productivity.  Back
 
Topics:
Deep Learning & AI Frameworks, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9483
Streaming:
Download:
Share:
 
Abstract:
We'll discuss our team's work to determine how GPU virtualization, 5G, and edge computing, together with cutting-edge hardware and software solutions, can make cloud AR/VR a reality. Our talk will cover how 5G will shift computing and data storage ...Read More
Abstract:
We'll discuss our team's work to determine how GPU virtualization, 5G, and edge computing, together with cutting-edge hardware and software solutions, can make cloud AR/VR a reality. Our talk will cover how 5G will shift computing and data storage to the cloud and enable new business models and commercial opportunities. We'll also talk about how 5G will enable high-resolution (4K or 8K) AR and VR, how these can revolutionize content consumption, and our role in making this possible.  Back
 
Topics:
Virtual Reality & Augmented Reality, GPU Virtualization, 5G & Edge
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9620
Streaming:
Download:
Share:
 
Abstract:
We'll discuss Project MagLev, NVIDIA's internal end-to-end AI platform for developing its self-driving car software, DRIVE. We'll explore the platform that supports continuous data ingest from multiple cars producing TB of data per h ...Read More
Abstract:

We'll discuss Project MagLev, NVIDIA's internal end-to-end AI platform for developing its self-driving car software, DRIVE. We'll explore the platform that supports continuous data ingest from multiple cars producing TB of data per hour. We'll also cover how the platform enables autonomous AI designers to iterate training of new neural network designs across thousands of GPU systems and validate the behavior of these designs over multi PB-scale data sets. We will talk about our overall architecture for everything from data center deployment to AI pipeline automation, as well as large-scale AI dataset management, AI training, and testing.

  Back
 
Topics:
AI Application, Deployment & Inference, Autonomous Vehicles, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9649
Streaming:
Download:
Share:
 
Abstract:
Learn how high-resolution imaging is revolutionizing science and dramatically changing how we process, analyze, and visualize at this new scale. We will show the journey a researcher can take to produce images capable of winning a Nobel prize. We'll ...Read More
Abstract:
Learn how high-resolution imaging is revolutionizing science and dramatically changing how we process, analyze, and visualize at this new scale. We will show the journey a researcher can take to produce images capable of winning a Nobel prize. We'll review the last two years of development in single-particle cryo-electron microscopy processing, with a focus on accelerated software, and discuss benchmarks and best practices for common software packages in this domain. Our talk will include videos and images of atomic resolution molecules and viruses that demonstrate our success in high-resolution imaging.  Back
 
Topics:
Computational Biology & Chemistry, In-Situ & Scientific Visualization, Data Center & Cloud Infrastructure, HPC and AI, Medical Imaging & Radiology
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9664
Streaming:
Download:
Share:
 
Abstract:
There are numerous problems which have been exposed by the creation of AI models due to the total capability of the current generation of GPUs to create and run a large volume of models, and we are going to show people how to fix them. The exponentia ...Read More
Abstract:
There are numerous problems which have been exposed by the creation of AI models due to the total capability of the current generation of GPUs to create and run a large volume of models, and we are going to show people how to fix them. The exponential compute growth which has occurred in this area has opened the doors to creating and testing hundreds or thousands more models than the, one-by-one approach which was performed in the past. These models use and generate data from both batch and real-time sources. As data becomes enriched, and parameters tuned and explored, there is a need for versioning everything, including the data. Issues found here are similar to other software engineering problems, but new approaches must be taken to create solutions given the complexity of the problems with the inclusion of vast amounts of data. We will discuss the very specific problems and approaches to fix them.  Back
 
Topics:
AI Application, Deployment & Inference, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9683
Streaming:
Download:
Share:
 
Abstract:
Despite enormous excitement about the potential of deep learning, building deep learning-powered practical applications remains an enormous challenge. The necessary expertise is scarce, hardware requirements can be prohibitive, and current software t ...Read More
Abstract:
Despite enormous excitement about the potential of deep learning, building deep learning-powered practical applications remains an enormous challenge. The necessary expertise is scarce, hardware requirements can be prohibitive, and current software tools are immature and limited in scope. We'll describe how deep learning workflows are supported by existing software tooling. Learn about promising opportunities to dramatically improve these workflows via novel algorithmic and software solutions, including resource-aware neural architecture search and fully automated GPU training-cluster orchestration. This talk draws on academic work at CMU, UC Berkeley, and UCLA, as well as our experiences at Determined AI, a startup that builds software to make deep learning engineers more productive.  Back
 
Topics:
Deep Learning & AI Frameworks, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9704
Streaming:
Download:
Share:
 
Abstract:
Many applications like telemetry, intrusion detection and anomaly detection can be accelerated by doing Packet Processing on a GPU. Using a GPU will also enable applying ML/DL for smarter Packet Processing. However, one of the bottlenecks to do Packe ...Read More
Abstract:
Many applications like telemetry, intrusion detection and anomaly detection can be accelerated by doing Packet Processing on a GPU. Using a GPU will also enable applying ML/DL for smarter Packet Processing. However, one of the bottlenecks to do Packet Processing on a GPU is to be able to ingest data at high bandwidth and low latency. The recent developments in this field will be reviewed in this session. Following this, the latest development from Nvidia along with preliminary benchmarking results will be presented. Nvidia has extended DPDK library to leverage direct RDMA to GPU memory, enabling close to line rate ingestion of network data into a GPU on 100GigE networks.  Back
 
Topics:
5G & Edge, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9730
Streaming:
Download:
Share:
 
Abstract:
We'll talk about applying ML/AI to the crucial task of identifying and isolating faults in computer and telecommunications networks. A problem with part or all of a single device can quickly propagate through the network, making it essential to iden ...Read More
Abstract:
We'll talk about applying ML/AI to the crucial task of identifying and isolating faults in computer and telecommunications networks. A problem with part or all of a single device can quickly propagate through the network, making it essential to identify a fault before it causes a hardware component to fail. We'll discuss cost-effective expert systems for network monitoring that are designed to minimize the number of service-affecting incidents, while keeping development, personnel, and maintenance costs at an acceptable level. We'll also explain how streaming telemetry enables access to real-time, model-driven, and analytics-ready data that can help with network automation, traffic optimization, and preventive troubleshooting.  Back
 
Topics:
5G & Edge, Data Center & Cloud Infrastructure
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9758
Streaming:
Download:
Share:
 
Abstract:
We'll provide a deep dive into implementing irradiance fields with moment visibility for local and cloud rendering using RTX. This is a real-time, fully dynamic diffuse global illumination solution that prevents light leaks and screen-space noise. T ...Read More
Abstract:
We'll provide a deep dive into implementing irradiance fields with moment visibility for local and cloud rendering using RTX. This is a real-time, fully dynamic diffuse global illumination solution that prevents light leaks and screen-space noise. The technology can operate on local GI on a Turing GPU for PC games or a streaming cloud GI from a Turing-powered server for VR and mobile gaming.  Back
 
Topics:
Gaming and AI, Data Center & Cloud Infrastructure, Rendering & Ray Tracing
Type:
Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9900
Streaming:
Download:
Share:
 
Abstract:
Our talk covers architecture considerations for federating ML and DL data pipelines to exploit GPU acceleration for a seamless tier of data science deployments across edge, core, and cloud. We'll take a look at different requirements, architecture o ...Read More
Abstract:
Our talk covers architecture considerations for federating ML and DL data pipelines to exploit GPU acceleration for a seamless tier of data science deployments across edge, core, and cloud. We'll take a look at different requirements, architecture options to meet them, and the resulting benefits to deliver distributed deployments of the data pipeline stages across data ingestion, data prep, training, inference validation, data science, and model serving. We'll also explore a few ways in which customers are deploying these. We will be joined by implementers of stages of the AI and data pipeline today to hear about their deployments and experiences.  Back
 
Topics:
Accelerated Data Science, AI Application, Deployment & Inference, Data Center & Cloud Infrastructure
Type:
Sponsored Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S9997
Streaming:
Share:
 
Abstract:
Global enterprises need to compress analysis time frames to update the business in real-time, a process called active analytics. We will discuss and demo how to bring together the key elements of an active analytics architecture, including histo ...Read More
Abstract:

Global enterprises need to compress analysis time frames to update the business in real-time, a process called active analytics. We will discuss and demo how to bring together the key elements of an active analytics architecture, including historical, streaming, and graph analytics, location intelligence, and machine learning for predictive analytics.

  Back
 
Topics:
Finance - Accelerated Analytics, GPU Virtualization
Type:
Sponsored Talk
Event:
GTC Silicon Valley
Year:
2019
Session ID:
S91001
Streaming:
Share: