Join Microsoft Azure at SC24

Date / Time (EST)

Type

Title

Speakers/Panelists

Sun, Nov 17
8:30 am - 5 pm

Tutorial

Delivering HPC: Procurement, Cost Models, Metrics, and Value

Microsoft Azure, Stanford University, Mitre, Red Oak Consulting

Mon, Nov 18
8:30 am - 5 pm

Tutorial

Parallel I/O In Practice

Microsoft Azure, Argonne National Laboratory (ANL), Google

Tues, Nov 19
10:30 am - 12 pm

Panel

How the Convergence of HPC and AI Is Accelerating Innovation

Microsoft Azure, AMD, Minds AI, Purdue University, d3VIEW

Tues, Nov 19
3:30 pm - 5 pm

Panel

Impact of Advances in Lower Precision Performance on HPC Simulation Codes

Microsoft Azure, NERSC at LBNL, NVIDIA, AMD, Chevron

Tues, Nov 19
5:15 pm - 6:45 pm 

Birds of a Feather

Benchmarking: A Key Driver for AI Innovation

Microsoft Azure, AMD, NVIDIA, Oracle, Snorkel AI

Tues, Nov 19
5:15 pm - 6:45 pm

Birds of a Feather

LUSTRE Community BOF: Lustre in HPC, AI, and the Cloud

Microsoft, Whamcloud, Los Alamos National Laboratory, Lawrence Livermore National Laboratory, Oak Ridge National Laboratory, University of Cambridge

Paper

RecFlex: Enabling Feature Heterogeneity-Aware Optimization for Deep Recommendation Models with Flexible Schedules

Microsoft, Renmin University of China, University of California San Diego

Wed, Nov 20
12:15 pm - 1:15 pm

Birds of a Feather

The Future of Benchmarks in Supercomputing

Microsoft Azure, NVIDIA, HPE 

Wed, Nov 20
2 pm - 2:30 pm

Paper

LoRAStencil: Low-Rank Adaptation of Stencil Computation on Tensor Core

Microsoft Research, UCAS, Tsinghua University, CAS

Wed, Nov 20
5:15 PM - 6:45pm

Birds of a Feather

Converged HPC and Cloud Computing in the Era of Generative AI

Microsoft Azure, Lawrence Livermore National Laboratory, Barcelona Supercomputing Center, USC Information Sciences Institute, IBM TJ Watson Research Center, Google, University of Tennessee, University of Delaware

Wed, Nov 20
5:15 pm - 6:45 pm

Birds of a Feather

The National Artificial Intelligence Research Resource (NAIRR) Pilot User Experience BOF

Microsoft, University of Colorado Boulder, Slippery Rock University of Pennsylvania, SambaNova Systems, Arizona State University

Thurs, Nov 21 
9:30 am - 10am

Paper

Long Exposure: Accelerating Parameter-Efficient Fine-Tuning for LLMs under Shadowy Sparsity

Microsoft Research, Tsinghua University

Fri, Nov 22 
11:30 am - 12 pm

Workshop

HPCAdvisor: A Tool for Assisting Users in Selecting HPC Resources in the Cloud

Microsoft 

Date/Time (EST)

Session Title

Abstract

Presenter

Mon, Nov 18

7:40 pm - 8:00 pm

Accelerating AI Innovation: Microsoft and NVIDIA Bring the Power of GB200 Superchip to Azure

Microsoft and NVIDIA are bringing the NVIDIA Grace Blackwell 200 (GB200) Superchip to Azure, enabling businesses to handle large-scale AI, data processing, and high-performance tasks with faster performance. This collaboration optimizes GPUs for advanced AI models, improving accuracy, efficiency, and scalability. Paired with upcoming NVIDIA's Quantum-X800 networking platform, this powerful infrastructure helps companies boost AI workloads and stay competitive with cutting-edge technology.

NVIDIA, Microsoft

Mon, Nov 18

8:10 pm - 8:30 pm

Revolutionizing Computational Science: AI-Driven Solutions on Rescale and Microsoft 

Learn how innovative AI capabilities, such as Rescale’s compute recommendation engine and AI Physics raise the bar for what's possible in R&D. We'll discuss how AI on Azure is solving challenges in areas such as Molecular Dynamics, Biomechanics, CFD and more...

Rescale, Microsoft

Tue, Nov 19

9:30 am - 11:30 am

Ignite 2024 Keynote

Live stream of Microsoft CEO, Satya Nadella's Ignite 2024 keynote presentation.

Microsoft

Tue, Nov 19

11:30 am - 12:30 pm

Special Azure HPC announcement

Join us for a special announcement where we’ll unveil a major leap forward in high-performance computing.

Microsoft Azure, AMD

Tue, Nov 19

12:40 pm - 1:00 pm

Together We Advance Cloud HPC

With Azure H-series infrastructure, based on AMD EPYC processors, Azure customers have access to super computing resources in the cloud. Join this session to learn how to reduce cost per job by using the latest H-series VMs.

AMD

Tue, Nov 19

1:10 pm - 1:30 pm

Fast and efficient processing of complex AI workloads with ND MI300X v5

Join us for an in-depth exploration of the Azure ND MI300X v5 VM series, a groundbreaking addition to Azure’s AI infrastructure. Learn how these VMs meet the demands of advanced AI workflows. 

AMD

Tue, Nov 19

1:40 pm - 2:00 pm

Case Studies in Digital Product Innovation with TotalCAE on Azure

TotalCAE makes Computer-Aided Engineering (CAE), Life Sciences, and AI applications easy to use on Azure in just a few clicks with TotalCAE-managed HPC services and the TotalCAE platform on Azure. Learn through several public customer case studies how clients are adopting Azure HPC optimized H series instances to accelerate their product designs and transform their HPC workflows with award-winning deployments in manufacturing, automotive, and life sciences.

TotalCAE, Microsoft

Tue, Nov 19

2:10 pm - 2:30 pm

Ansys & Azure: How to leverage the power of Azure to accelerate Ansys workloads

Learn how Ansys and Microsoft leverage the compute power of Azure and the solving power of Ansys to accelerate the pace of innovation. This session will focus on the migration of Ansys workloads to Azure cloud infrastructure and Ansys Access on Microsoft Azure™. 

Ansys, Microsoft

Tue, Nov 19

2:40 pm - 3:00 pm

Accelerating Hybrid Quantum-Classical Computing with Microsoft & Quantinuum

This session will cover recent announcements and advancements by Quantinuum and Microsoft in quantum computing and applications.

Quantinuum, Microsoft

Tue, Nov 19
3:10 pm - 3:30 pm

Rescale & Microsoft: Simplifying Advanced Computing to Scale Digital Engineering & AI Innovation 

Join our session to learn how Microsoft and Rescale simplify HPC, enhancing collaboration between engineers, data scientists, and IT leaders to power innovation. 

Rescale, Microsoft

Tue, Nov 19

3:40 pm - 4:00 pm

Hurricane Modeling in Public Cloud

In this session, we'll explore how the 2024 Hurricane Analysis and Forecast System (HAFS) Ensemble in Real-time on the Cloud (HERC) experiment showcased NOAA’s ability to run advanced hurricane models seamlessly on the cloud.  

Parallel Works, Microsoft

Tue, Nov 19

4:10 pm - 4:30 pm

Revolutionizing HPC Workflows with Generative AI assistants

Explore how advanced generative AI is revolutionizing high-performance computing (HPC) workflows. Discover an AI-driven tool designed to streamline HPC processes by reducing manual intervention, with a featured use case in genomics and health sciences.

Microsoft Azure

Tue, Nov 19

4:40 pm - 5:00 pm

Accelerating AI for Science on Supercomputers using DeepSpeed and Megatron

Join this session to learn about Microsoft AI for Science and how Microsoft's AI models for science leverage deepspeed frameworks together with megatron for diverse scientific domains.  

Microsoft Research

Tue, Nov 19

5:10 pm - 5:30 pm

Azure Storage for AI Supercomputers

This is an overview session covering the various announcements in Azure storage products which cater to modern AI workloads.

Microsoft Azure

Wed, Nov 20

10:10 am - 10:30 am

Accelerating EDA workloads on Azure – Best Practice and benchmark on Intel EMR CPU

Learn best practices for running EDA on Azure, including recommended architecture, optimal compute VMs, and high-performance NFS storage solutions. And benchmark results for EDA tools on Azure and Azure VMs with 5th Gen Intel® Xeon® Platinum 8537C processor. 

Microsoft Azure 

Wed, Nov 20

10:40 am - 11:00 am

Azure CycleCloud and the WEKA Data Platform Integration

The integration of Azure CycleCloud with the WEKA Data Platform empowers businesses to achieve high performance, scalability, and efficiency in their high-performance computing (HPC) environments.
CycleCloud orchestrates and manages HPC clusters in Azure, providing autoscaling and configuration management for complex deployments. With WEKA, users gain a high-performance file system optimized for low-latency and high-throughput workloads, ideal for AI, analytics, and machine learning applications.
This session will describe the solution and demonstrate the advantages with a live demo. 

Weka, Microsoft

Wed, Nov 20

11:10 am - 11:30 am

A Full Stack Approach to Accelerated Computing

This session will discuss how NVIDIA and Azure are partnering to provide accelerated computing for HPC and AI workloads. Ashraf will highlight NVIDIA's latest innovations in GPUs and software, showcasing how they are used in supercomputing platforms to address the opportunities and challenges faced by scientists, institutions, and governments worldwide. Listeners will gain insights into how Microsoft and NVIDIA are collaborating to deliver a comprehensive, full-stack solution to these challenges.

NVIDIA, Microsoft

Wed, Nov 20

11:40 am - 12:00 pm

Deploy HPC & AI Infrastructure at Scale on Azure in Minutes! – Part 1

Join us for an engaging demonstration on how to build your HPC and AI solutions with ease using Microsoft Azure CycleCloud Workspace for Slurm (CCWS). Learn key features that can bring up HPC workloads in minutes and significantly boost productivity and efficiency.

Microsoft Azure

Wed, Nov 20

12:10 pm - 12:30 pm

Accelerating Enterprise Computing with Altair Workload Managers and Azure CycleCloud

Altair's workload managers and cloud enablement solutions, in synergy with Azure CycleCloud, empower Azure HPC/AI users with unmatched performance and agility. This presentation will showcase the seamless integration of our cutting-edge scheduling solutions—Altair PBS Professional and Altair Grid Engine —within Azure's purpose-built HPC/AI infrastructure. Join us to witness firsthand how this collaboration transforms efficiency and accelerates innovation in high-performance computing and artificial intelligence.

Altair, Microsoft

Wed, Nov 20

12:40 pm - 1:00 pm

Running Small Language model inference on Azure NC series VMs

Join us for a live demonstration of deploying small language models using Azure’s NC H100 v5 VMs. Discover how to leverage open-source tools like Ollama to deploy the Phi3 Mini model, showcasing the seamless integration of NVIDIA H100 GPUs with Azure’s powerful infrastructure. Learn the streamlined deployment process, performance optimization techniques, and best practices for managing NLP projects on Azure. Whether you’re an AI developer or data scientist, this session will provide the essential skills to efficiently deploy and run language models on Azure’s NC series. Don’t miss this opportunity to enhance your AI initiatives with Azure’s advanced GPU offerings!

NVIDIA, Microsoft

Wed, Nov 20

1:40 pm - 2:00 pm

Increase Cryo-EM performance, discover more at impossibly small scale

See how global pharma companies are working with Hammerspace and running on AMD powered instances in Azure to offer users the ability to scale their infrastructure as needed to meet the changing requirements of Cryo-EM and their research.

Hammerspace, Microsoft

Wed, Nov 20

2:10 pm - 2:30 pm

High Performance Computing with AMD EPYC Processors on Azure H-series VMs

Come get an update on how to get the most out of your HPC applications using Azure’s HB-series VMs with AMD EPYC Processors.

Microsoft Azure

Wed, Nov 20

2:40 pm - 3:00 pm

The European Environment for Scientific Software Installations (EESSI)

Have you ever wished that all the scientific software you use was
available on all the resources you have access to without having to go
through the pain of installing it the way you want/need? In this talk,
we will introduce the European Environment for Scientific Software
Installations (EESSI — pronounced "easy") as a common stack of optimized
scientific software installations for HPC systems and beyond all around
the globe.

EESSI, Microsoft

Wed, Nov 20

3:10 pm - 3:30 pm

Esri using Reality Capture for Climate Science in the City of Stuttgart.

Learn how Esri's ArcGIS Pro, powered by Azure and NVIDIA, is used for Climate Science to predict and mitigate flooding in the City of Stuttgart. We'll demonstrate how high-resolution 3D models of the city are created and analyzed. 

NVIDIA, Microsoft Azure

Wed, Nov 20

3:40 pm - 4:00 pm

Deploy HPC & AI Infrastructure at Scale on Azure in Minutes – Part 2!

This is a continuation of Part 1 session on deploying a full HPC/AI infrastructure using Azure CycleCloud Workspace for Slurm (CCWS). Get a full walkthrough demo on how to deploy and scale your OSS toolchain including familiar tooling, like Open OnDemandand and Linux VDI from with ThinLinc from Cendio.

Microsoft Azure

Wed, Nov 20

4:10 pm - 4:30 pm

Deploying small-to-medium AI workloads on Azure’s PCIe GPUs

Discover how Azure's NC and NV series VMs are revolutionizing the deployment of small-to-medium AI workloads by optimizing performance, scalability, and insights. We’ll introduce you to Azure’s comprehensive PCIe GPU portfolio and the ideal GPU solutions.

Microsoft Azure

Wed, Nov 20

4:40 pm - 5:00 pm

Empowering Workforce Development through AI-Powered Power BI Training

Presenting examines how Azure AI and Power BI are integrated into Emory University’s workforce development curriculum, equipping professionals with skills in AI-driven data analytics. Using tools like sentiment analysis, keyword extraction, and vision recognition, participants generate actionable insights and visualize data in Power BI.

Emory University, Microsoft Azure

Wed, Nov 20

5:10 pm - 5:30 pm

Azure Storage solutions for your most demanding HPC and AI training workloads

This is an overview session covering the various announcements in Azure storage products which cater to modern AI workloads.

Microsoft Azure

Thurs, Nov 20

10:40 am - 11:00am

Introducing SimOps and how to build and operate and optimal and cost-efficient HPC/AI infrastructure on Azure

By collaborating with Azure and manufacturing customers like Rimac Automobili, FLSmidth, and others, Wolfgang from Simr will share insights gained from Azure clients and demonstrate how these lessons have been transformed into SimOps Best Practices. They will outline the SimOps Framework, explaining how it can be used to build, operate, and manage HPC/AI infrastructures in a way that is both efficient and cost-effective

Simr, Microsoft Azure

Thurs, Nov 20

11:10 am - 11:30am

Azure NetApp Files, boosting HPC productive with Innovation

In this session, we will explore the pivotal role of file systems in optimizing High-Performance Computing (HPC) at an Azure scale, utilizing tens of thousands of compute cores working in unison to tackle the most advanced HPC challenges. The key solution is Azure NetApp Files (ANF), which Microsoft Azure customers are leveraging with remarkable efficacy today. Thanks to ANF’s low latency, high throughput, and simplified management, users are experiencing significant productivity gains by eliminating the hardware constraints traditionally associated with running HPC workloads.

NetApp, Microsoft Azure

Thurs, Nov 20

12:10pm- 12:30pm

Qumulo Global Data Platform: Removing data boundaries to deliver infinite possibilities

Imagine if you could share your data - globally, securely and efficiently - to accelerate time-to-results and insights. Imagine if you could elastically scale cloud performance and capacity on-demand to meet your application demands. Learn how Qumulo Global Data Platform unlocks the value of your data to enable faster time-to-results and enables transformative opportunities in accelerated computing. In this presentation you will discover how Qumulo Run Anywhere with Global Namespace enables you to the choice to host your data on-premise or in the cloud, while ensuring your data can be accessed anywhere, whether in the cloud or in your data center, while maintaining control of your most valuable asset; your data.

Qumulo, Microsoft

Thurs, Nov 20

2:10pm- 2:30pm

Quantum & AI to Accelerate Scientific Discovery

AI and quantum computing have the potential to help solve some of the most pressing challenges of our time -- from the climate crisis to world hunger to finding cures for disease. As we continue to improve the fidelity and scale of quantum machines, we will first be able to unlock scientific quantum advantage, solving a growing class of complex and intractable problems. As we scale to quantum supercomputers, we will be able to achieve commercial quantum advantage and solve the world’s most pressing challenges through quantum-enabled advances in chemistry, biochemistry, and materials science. This session will explain Microsoft’s latest advancements in quantum computing and how Microsoft is building a new science platform using AI, HPC, and quantum to accelerate scientific discovery in chemistry, biology, and physics.

Azure Quantum

Date

Start

Demo

Mon Nov 18

7:00 am EST – 9:00 AM

NVIDIA demo

Tues Nov 19

FULL DAY

AMD demo

Wed Nov 20

10:00 AM – 2:00 PM

NVIDIA demo

Wed Nov 20

2:00 PM – 4:00 PM

Rescale demo 

Wed Nov 20

4:00 PM – 6:00 PM

Ansys demo