- 11
- 13 183 823
Nebius
Netherlands
เข้าร่วมเมื่อ 11 ม.ค. 2023
We are an AI-centric public cloud platform specifically crafted to serve AI models from ML training to inferencing at a scale.
Learn more: nebius.ai/
Learn more: nebius.ai/
Nebius Founder and CEO Arkady Volozh on AI infrastructure at the FT’s Future of AI Summit 2024
At the Financial Times’s Future of AI Summit 2024, Nebius’s CEO and Founder, Arkady Volozh speaks to Elaine Moore, Tech Comment Editor at the Financial Times. In this fireside chat Arkady discusses AI infrastructure, leading an innovative startup and how to stand out as an unusual player in the AI industry.
00:00 Introduction
00:51 History of Nebius
03:24 Yandex 2.0 or a startup
05:47 Impact of restart of trading on NASDAQ
08:40 AI Infrastructure expansion into US
11:40 Data center expansion in Europe
13:13 NVIDIA and Nebius’s relationship
15:56 Nebius’s business units
18:17 R+D
19:34 Future of AI
21:31 Conclusion
Connect with us:
Website: www.nebius.com
LinkedIn: www.linkedin.com/company/nebius-group
X: nebiusgroup
00:00 Introduction
00:51 History of Nebius
03:24 Yandex 2.0 or a startup
05:47 Impact of restart of trading on NASDAQ
08:40 AI Infrastructure expansion into US
11:40 Data center expansion in Europe
13:13 NVIDIA and Nebius’s relationship
15:56 Nebius’s business units
18:17 R+D
19:34 Future of AI
21:31 Conclusion
Connect with us:
Website: www.nebius.com
LinkedIn: www.linkedin.com/company/nebius-group
X: nebiusgroup
มุมมอง: 4 343
วีดีโอ
Beyond ChatGPT: unlocking the power of open-source LLMs
มุมมอง 4692 หลายเดือนก่อน
As GenAI application builders face escalating costs and privacy concerns with proprietary models like ChatGPT, open-source alternatives such as Llama and Mistral are emerging as game-changers. These models offer comparable quality at a fraction of the cost, with enhanced privacy controls. In this webinar record, Nikita Vdovushkin, product lead, will unveil our in-house inference infrastructure ...
Into the wild building an AI cloud for AI builders
มุมมอง 20K2 หลายเดือนก่อน
On stage is Rashid Ivaev, Business Development Director at Nebius. Rashid is explaining how exactly we gathered insights into what ML engineers need from an AI cloud and how we are advancing the platform today. The French AI community has always welcomed us warmly, which is especially important considering the opening of our GPU cluster in Paris this November. #conferences #events #networking #...
What it takes to train a GenAI model from scratch
มุมมอง 1.1K5 หลายเดือนก่อน
This is a talk at the London AI Summit by Narek Tatevosyan, Product Director at Nebius. He shares an insider's perspective on the key steps, essential tools, and challenges involved in bringing a generative AI model from concept to production. In this talk, he will break down the real story behind: - Data preparation - Experiments to prepare foundational model training - Pre-training - Fine tun...
Decoding GPU pricing - tips from a CFO
มุมมอง 57K6 หลายเดือนก่อน
The GPU cloud market is flourishing, bringing with it a variety of pricing models. In this webinar, Danila Pavlov, CFO at Nebius, guided through the intricacies of GPU pricing. Danila provided practical tips on how to interpret official price lists, identify hidden costs and ask the right questions of your service provider. ✅ Find out more on our website: nebius.ai/ Like, Share, and Subscribe f...
Deploying a knowledge-based chatbot with RAG in production
มุมมอง 18K7 หลายเดือนก่อน
On this webinar Boris Popov, CSA at Nebius AI, talk about the deployment of a knowledge-based chatbot using RAG in a production environment. This implementation leverages open source technologies and is powered by NVIDIA® H100 Tensor Core GPUs. We will also discuss the integration of Kubernetes, Cuda, Triton Server, TensorRT, Milvus, PyTorch, and Llama2. During this session, we will cover: - Te...
The ultimate cloud for AI explorers
มุมมอง 13M8 หลายเดือนก่อน
Nebius AI is an AI cloud platform that is one of the largest providers of GPU capacity in Europe. Launched in November 2023, the Nebius AI platform provides high-end, training-optimized infrastructure for AI practitioners. We’ve filmed this video 60 kilometers from Helsinki, the home of the Nebius data center. This is where we built ISEG, a top-16 supercomputer in the world - and we recently al...
Fail fast & recover faster: infrastructure resilience of multi-node LLM training
มุมมอง 3898 หลายเดือนก่อน
Training an LLM model in a multi-node setup is a complex and expensive process. Training failures can't be eliminated, but downtime can be reduced. In this talk, Filipp Fisin, Senior ML Engineer at NebiusAI, provide an overview of techniques for more resilient training that we've found useful in our JAX-based multi-node training setup, namely: - multi-node training orchestration in Kubernetes v...
Choosing the right platform: Slurm vs Kubernetes
มุมมอง 9K9 หลายเดือนก่อน
In this video, Panu Koskela, our cloud solutions architect, will explore Slurm and Kubernetes, and cover their architecture, design, original purposes, and adaptations for ML. He’ll also discuss key considerations for choosing between them. nebius.ai/ 00:00 - 1:30 Intro 1:30 - 3:15 What is Nebius AI? 3:15 - 10:45 Challenges of training large models 10:45 - 19:40 Slurm specifics 19:40 - 27:48 Ku...
Recraft 20B foundational model: what it takes to train it
มุมมอง 64010 หลายเดือนก่อน
Nebius and Recraft shared the experience of training the 20B parameters foundational model. Recraft started using Nebius at the technical preview stage. As one of the early adopters, they contributed to creating the new AI cloud. You will learn about all the challenges and how both teams worked together to overcome them. www.recraft.ai/ nebius.ai/ 00:00 - 2:11 - Into 2:12 - 5:21 - Meet Nebius A...
Cloud optimised for AI model training
มุมมอง 110K10 หลายเดือนก่อน
Nebius is an AI-centric cloud platform that is one of the leading providers of GPU capacity in Europe. Launched in late 2023, it’s specifically crafted to serve AI models from ML training to inferencing at a scale. Find more information about Nebius AI and our offering on our website. ➡️ Learn more: nebius.ai/ #ai #nvidia #machinelearning #datascience #cloudcomputing
bullish $150 target.
Kubernetes supports the "finite resources, infinite work" paradigm via extensions like Kueue and Yunikorn. The downside is that this requires managing additional machinery on top of the base complexity of Kubernetes.
Искренне желаю удачи, Аркадий!
Why dont we simply ask to aig & marsh mclennon....or enron for that matter
We all know ur onions Amsterdam stop bragging, but can u cicada ur way through in2my phone w/ aracnia, locate my CV, run it on a virtual box n write me 7zip css htm check.iso'?or I might highjack ur basal ganglia infiltrate amygdala n do something non commutative casual entropic cool idunno...data broke ur nmda(mg+)may b( look look imma gonna trigger u...drth sidious beats Tolkien Tolkien beats Potter et al..u cant even assembly bruv..tu stuit est,QEDemonstrantum)
Good luck, Arkady!
This is the guy that started the war
Yes, cheering up for Nebius & its entire team!!!
Davai Arkady
LONG - BULLSIH
You bag holding now. 😂
@@derekaugust872no, my avg cost basis is $20.30. Clown “😂”
@@derekaugust872 avg. cost basis 20.30, who’s a bag holder?
@ cool story bro!
@@derekaugust872 🤣🤣 sad your little dig didn’t go according to plan? Get a life man
fringe
why all the memes lol
learn to stop scamming
Explain for me
Hi
Great explanation, really liked the graph with AI tools for training! Thanks!
Это что, бывший заграничный Яндекс?
да Nebius они называются
35:25 Why is slurm better for mpi workloads? Compared to eg K8s with MPI Operator.
Way to go Nebius AI, it seems as if compute is getting less expensive already in 2024!