Anton Alekseev
Avito
If you have a ticket, log in to watch the video
LoginTalk
GPU Inference in K8s: Acceleration, Sharing and Scaling Without Pain
In RussianComplexity -
How can I speed up GPU inference in Kubernetes and not go crazy? It's all about scaling, sharing, speeding up the start and choosing shaders. With examples, hacks, and conclusions from real production.
Speakers
Invited experts
Timur Gilmullin
Automator, independent expert
Other talks on «K8s»
- Watch recording3 hallIn RussianComplexity -
- Watch recording
Scheduling GPU Workloads in Kubernetes: From Standard Mechanisms to Custom Solutions
Makarii Balashov
Yandex Cloud
2 hallIn RussianComplexity - - Watch recording
L2 Announcements in Cilium: Access to Load Balancer in Bare-Metal Kubernetes
Mikhail Petrov
Yandex Cloud
2 hallIn RussianComplexity - - Watch recording1 hallIn RussianComplexity -
- Watch recording
Deploying the Open-Source Site of the Chelyabinsk Zoo in Kubernetes
Aleksandr Shinkarev
Tourmaline Core
Maxim Rychkov
Tourmaline Core
3 hallIn RussianComplexity -
- Watch recording1 hallIn RussianComplexity -
- Watch recording1 hallIn RussianComplexity -
Other talks on «Infrastructure»
- Watch recording
Scheduling GPU Workloads in Kubernetes: From Standard Mechanisms to Custom Solutions
Makarii Balashov
Yandex Cloud
2 hallIn RussianComplexity - - Watch recording2 hallIn RussianComplexity -
- Watch recording2 hallIn RussianComplexity -
- Watch recording1 hallIn RussianComplexity -
- Watch recording
Evolution of the Logging System at ecom.tech: From Elastic Stack to VictoriaLogs
Valerii Evdokimov
ecom.tech
1 hallIn RussianComplexity - - Watch recording2 hallIn RussianComplexity -
- Watch recording
Building Centralized Configuration Management of Infrastructure with K8s and SaltStack
Ivan Gulakov
MWS Cloud Platform
2 hallIn RussianComplexity - - Watch recording
Fast or Slow: The Story of the Struggle to Speed up Image Builds
Mikhail Kozhukhovskii
Analytical Program Solutions
Avenir Voronov
KORUS Consulting
Konstantin Dipež
DeusOps
In RussianComplexity - - Watch recording1 hallIn RussianComplexity -
- Watch recording
The Server Went Down, but the Cache Didn’t Give Up — It Went to S3 and the Production Went Up!
Nikolay Gubin
Avito
1 hallIn RussianComplexity -
Other talks on «ML/AI»
- Watch recording
The Battle of the Code Assistants. Efficiency, Safety and Cost
Anton Chernousov
Yandex Cloud
Aleksandr Kirillov
Evrone
1 hallIn RussianComplexity - - Watch recording
From Rag for Operators to a Rag Platform for a Large Bank
Mark Kuznetsov
Alfa-Bank
Alexey Fateev
Alfa-Bank
1 hallIn RussianComplexity - - Watch recording1 hallIn RussianComplexity -
- Watch recording
How To Get the Most out of GPU and Ray: Our Production ML Infrastructure Pipeline
Mikhail Untura
Orion soft
1 hallIn RussianComplexity - - Watch recording1 hallIn RussianComplexity -
- Watch recording
Break Me Completely: How AI (Does Not) Help with Pentesting
Avenir Voronov
KORUS Consulting
Konstantin Dipež
DeusOps
In RussianComplexity - - Watch recording
AI in SDLC
Avenir Voronov
KORUS Consulting
Ilia Atarshchikov
KORUS Consulting
