Developer workshops are restricted to machine learning practitioners from research institutions and enterprises who are interested in learning how to port code onto novel AI platforms and want to get hands-on access to hardware and SDKs. 

Workshops are application only and subject to eligibility and availability. The workshops are free, and lunch, shared networking sessions, and access to the Meet and Greet function and keynote is included in the developer pass. If you're a machine learning engineer / AI application developer, please apply using the form in the registration section of the website or by emailing [email protected]. There are approximately 30 spaces available.

Using Hugging Face Optimum, we will show how easy it is to run and accelerate an end-to-end state-of-the-art Transformer model workflow on IPUs. We’ll provide a demo to fine-tune a BERT-Large Transformer model in IPUs using Hugging Face Optimum, and then serve the model using the inference API. Join us as we look under the hood of our Intelligence Processing Unit to see how our unique architecture, combined with Optimum’s simple plug-and-play experience, enables faster performance on today’s most popular ML models. By the end of the talk, we would like you to have a better understanding of the wide range of off-the-shelf NLP & Computer Vision Transformer models through our integration with the Hugging Face ecosystem. Plus, how to access this walkthrough by using our free IPU runtimes in the cloud.

Developer workshops are restricted to machine learning practitioners from research institutions and enterprises who are interested in learning how to port code onto novel AI platforms and want to get hands-on access to hardware and SDKs.  

Workshops are application only and subject to eligibility and availability. The workshops are free, and lunch, shared networking sessions, and access to the Meet and Greet function and keynote is included in the developer pass. If you're a machine learning engineer / AI application developer, please apply using the form in the registration section of the website or by emailing [email protected]. There are approximately 30 spaces available.

Developer workshops are restricted to machine learning practitioners from research institutions and enterprises who are interested in learning how to port code onto novel AI platforms and want to get hands-on access to hardware and SDKs.  

Workshops are application only and subject to eligibility and availability. The workshops are free, and lunch, shared networking sessions, and access to the Meet and Greet function and keynote is included in the developer pass. If you're a machine learning engineer / AI application developer, please apply using the form in the registration section of the website or by emailing [email protected]. There are approximately 30 spaces available.

Cerebras Systems builds the fastest AI accelerators in the industry. In this talk we will review how the size and scope of massive natural language processing (NLP) presents fundamental challenges to legacy compute and to traditional cloud providers. We will explore the importance of guaranteed node to node latency in large clusters, how that can’t be achieved in the cloud, and how it prevents linear and even deterministic scaling. We will examine the complexity of distributing NLP models over hundreds or thousands of GPUs and show how quickly and easily a cluster of Cerebras CS-2s is set up, and how linear scaling can be achieved over millions of compute cores with Cerebras technology. And finally, we will show how innovative customers are using clusters of Cerebras CS-2s to train large language models in order to solve both basic and applied scientific challenges, including understanding the COVID-19 replication mechanism, epigenetic language modelling for drug discovery, and in the development of clean energy. This enables researchers to test ideas that may otherwise languish for lack of resources and, ultimately, reduces the cost of curiosity.  ​

Deep neural networks (DNNs), a subset of machine learning (ML), provide a foundation for automating conversational artificial intelligence (CAI) applications. FPGAs provide hardware acceleration enabling high-density and low latency CAI. In this presentation, we will provide an overview of CAI, data center use-cases, describe the traditional compute model and its limitations and show how an ML compute engine integrated into the Achronix FPGA can lead to 90% cost reductions for speech transcription.

As AI makes its way into healthcare and medical applications, the role of hardware accelerators in the successful deployment of such large AI models becomes more and more important. Nowadays large language models, such as GPT-3 and T5, offer unprecedented opportunities to solve challenging healthcare business problems like drug discovery, medical term mapping and insight generation from electronic health records. However, efficient and cost effective training, as well as deployment and maintenance of such models in production remains a challenge for healthcare industry. This presentation will review a few open challenges and opportunities in the healthcare industry and the benefits that AI hardware innovation may bring to the ML utilization.

One of the biggest challenges in the US is managing the cost of healthcare.  Although we have high healthcare costs in the US, our life expectancy is still average.  In this talk we will look at some of the core causes of healthcare costs and what modern AI hardware can do to lower these costs.  We will see that faster and bigger GPUs alone will not save us.  We need detailed models to across a wide swath of our communities and perform early interventions.  We need accurate models of our world and the ability to simulate the impact of policy changes to overall healthcare costs.  We need new MIMD hardware with cores and memory architecture that keep cores fed with the right data.