The Habana team is excited to share with you our deep learning technologies and invite you to see our demos featured at the Intel Innovation 2022 conference happening today and tomorrow at the San Jose McEnery Convention Center. Our focus for the conference is on creative AI applications possible with our new Gaudi2 Deep Learning processor and how we’re making it easy for developers to access and build with Gaudi in the cloud, as well as on-premises.
Here’s what’s happened already this morning and what we look forward to sharing with you over the next two days.
Text-to-image Diffusion Demo
In today’s keynote, Intel CEO Pat Gelsinger, Ria Cheruvu and Jon Markee demonstrated the Gaudi2 accelerator executing a text-to-image diffusion model. When Ria input text prompts, Gaudi2 spontaneously generated batches of 16 unique images depicting her prompts—in roughly 3.5 seconds. This blazing-fast deployment was executed on a single Gaudi2 accelerator. With this new and evolving text-to-image technology, we are making creativity more accessible to businesses and individuals, which is opening up an entirely new realm of creative possibilities.
See the demo up close
- Text-to-image diffusion demo: Is featured in Habana’s booth #548, in the conference showcase, over the next two days at Intel Innovation. Please come by our booth to learn more about Gaudi2 and to see how text-to-image models can be implemented.
Accelerating Transformer Models with Hugging Face on Habana Gaudi and Amazon EC2 DL1 Instances
Habana is teamed up with Hugging Face to make it easier and quicker to train high-quality transformer models. Thanks to the integration of Habana’s SynapseAI Software Suite with the Hugging Face Optimum open-source library, data scientists and machine learning engineers can accelerate their Transformer jobs on Habana processors with just a few lines of code. They can also enjoy great cost-savings training on Gaudi-based Amazon EC2 DL1 Instances, which deliver up to 40% better price performance than comparable GPU-based instances on the AWS cloud.1
Come to the session
- Want to learn more about how to use transformer models? Julien Simon, chief evangelist for Hugging Face, will conduct a one-hour, hands-on presentation and demonstrate how to accelerate model training and inference deployment on Gaudi® and Xeon® Scalable Processors, respectively.
- When and where: Wed, Sep 28, 10:45 – 11:45; Room 210G
See and learn more up close
- Today and tomorrow at the Habana demo booth we will provide a detailed how-to demo showing visitors “How to Get Started on Transformers with Habana Gaudi.” Please come by if you want to get a quick personalized tutorial or ask questions.
Want to learn more about Gaudi and Gaudi2 solutions?
- Inspur: We’re pleased to share that this morning Inspur announced its new NF5688M7 Server, featuring Gaudi2 for training and inference, plus dual-socket host CPU, Intel Sapphire Rapids. Stop by Inspur’s booth #412 to learn more about Gaudi2 and read the details here.
- Supermicro: Check out the Supermicro Gaudi2 Server, which enables high-performance, high-efficiency model training and inference, featured in Supermicro booth #306, and read the product details here.
- Ready-to-deploy AI Development & Deployment Platform for Data Scientists: Be sure to check out this novel system solution at booth #551 and the session on Wednesday, 1 to 2pm in room 210H. This ready-to-deploy AI system solution comprises a validated and packaged bundle of hardware, software and support for data scientists and IT, making it fast and easy for end-users to quickly implement AI models. The system features the Supermicro Gaudi Training Server featuring eight deep learning accelerators and dual-socket host CPUs— 3rd Gen Intel(R) Xeon(R) Scalable Processors, cnvrg.io for MLOps, Red Hat OpenShift Data Science and software tools and support.
- Multi-model video search: This live demo, powered by Intel Xeon and Gaudi processors, demonstrates mining knowledge from video at scale. It features a multi-model transformer developed by Intel Labs’ Cognitive AI team. Stop by #3003 in the Innovation Zone.
Looking to tomorrow:
- Be sure to attend or tune in to see Greg Lavender’s keynote live or streamed at 8:30am tomorrow, Wed Sep 28.
- And, of course come to Julien Simon’s Hugging Face presentation at 10:45.
Want to meet?
The Habana team would love to meet you and learn about your deep learning compute requirements. Feel free to come by our booth or drop us your info here and we’ll respond asap!