Running AI workloads on Google Kubernetes Engine (GKE) presents unique challenges, especially for securing the right hardware. Whether you’re dealing with unpredictable demand and varying job durations or simply looking to control costs, this session will equip you with the knowledge and tools to make informed decisions about your GKE AI infrastructure. We’ll explore recent advancements in Dynamic Workload Scheduler, custom compute classes, and Kueue, demonstrating how these technologies can help you effectively access and manage diverse hardware resources.
talk-data.com
F
Speaker
Fisayo Feyisetan
2
talks
Product Manager
Google Cloud
Filter by Event / Source
Talks & appearances
2 activities · Newest first
with
Fisayo Feyisetan
(Google Cloud)
,
Ed Donahue
,
Theodoros Lappas
,
Thomas Soemo
,
Anthony Mocny
,
Yumao Lu
,
Gina Lee
,
Cenyu Zhang
Join us as we unveil GPT-4 Visual, a new model from OpenAI introducing multimodal input and output capabilities. Explore how GPT-4 Visual is integrated into Azure Cognitive Search and supercharged with vision embeddings, transforming our approach to AI-driven information retrieval. Images and videos can now prompt, or supplement prompts, to large language models (LLMs) like GPT-4. We will also introduce new multimodal models for Azure AI Content Safety, part of our Responsible AI product suite.
To learn more, please check out these resources: * https://aka.ms/Ignite23CollectionsBRK205H * https://info.microsoft.com/ww-landing-contact-me-for-events-m365-in-person-events.html?LCID=en-us&ls=407628-contactme-formfill * https://aka.ms/azure-ignite2023-dataaiblog
𝗦𝗽𝗲𝗮𝗸𝗲𝗿𝘀: * Fisayo Feyisetan * Theodoros Lappas * Thomas Soemo * Anthony Mocny * Cenyu Zhang * Gina Lee * Ed Donahue * Yumao Lu
𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻: This video is one of many sessions delivered for the Microsoft Ignite 2023 event. View sessions on-demand and learn more about Microsoft Ignite at https://ignite.microsoft.com
BRK205 | English (US) | AI & Apps