Wrote this in notes so may not line up correctly. $NBIS NBIS: - NBIS is more of a Ai infrastructure data center play. Buy the GPUs, set them up in clusters for a data center, and rent it out over the cloud. - IaaS - IaaS advantages: - - Finnish DC one of the most efficient in the world. Cold outside weather. Heats nearby homes with excess heat. - - independent DC player. While many hyperscalers are trying to build their own AI accelerator chips, NBIS is not, NVDA sees this and welcomes them as a long term partner over hyperscalers. Coreweave and independent AI infrastructure player also just got NVDA Blackwell and is the first company to offer them on the cloud. Likely NBIS is next or close to. - - With DeepSeek breakthrough, shows what possible for “smaller” AI/LLM players. Also, DeepSeek is available on NBIS AI studio. - - Just focused more on IaaS vs IaaS comparison with coreweave, NBIS trades at a discount to coreweave. Coreweave expected to IPO this year at a monster valuation of $35bn. Coreweave Did $2bn in revenue in 2024 and projects $8bn this year. Had a raise that valued it at $23bn in November. Gives its a 4.4 p/s multiple based on 25 numbers. However, Coreweave has debt while NBIS has none. If I’m not mistake Coreweave has raised over $9.8bn in debt financing over the last 2 years. Their biggest customer is Microsoft who wasn’t able to handle the demand for Azure/Openai and needed help. While it’s helped Coreweave and they have a multi year contract with MSFT, it will end eventually and MSFT is focused on doing cloud in house (instead of paying a 3rd party.) NBIS customers are more differentiated and also work with a lot of smaller sized businesses. Regardless, COREWEAVE IPO will be incredibly bullish for NBIS. Should attract major attention and NBIS should receive a higher multiple. - NBIS ‘24 revenue: 70-80mn, 170-190 ARR based on Q4 est. And 400-600mn for 2025 with and est of .5-$1bn in ARR based of end of 2025. On the latest earnings calls they mentioned they’re likely to arrive above the midpoint. So ARR likely not around $750mn to - $1000bn. That’s over 3x growth in ARR between Q4’24 and Q4’25. They’ve mentioned that this really depends on their access of capital. - PAAS - PaaS advantages. - PaaS. Now NBIS isn’t just a bare metal GPU cloud provider. They’re unique that they offer AI building tools. So they’re more than just a GPU cloud provider like Coreweave. They offer AI building tools like many of the hyperscalers but are independent. Also work and reach out to a lot of smaller businesses. And they offer AI type tools at different levels. So if you want to build your own LLM model or other AI products, you can. You can also fine tune and customize open source models. They had a video where their customers all needed something different. Some needed more in depth tools while others needed more of a ready to go AI product to use. And they accomplished because they have (hundreds I believe) of AI/ML engineers and developers that are working to provide this. This PaaS really differentiates them from the pack of bare metal or GPU cloud providers. MISTRAL AI uses them along with other players. - DeepSeek really showed that smaller players can deliver amazing LLM and ai products. From my understanding, DeepSeek showed that inference is only going to be more important from this point. NBIS is better positioned for this than bare metal AI cloud providers. - OTHER BUSINESSES. - They have a few other businesses. AVride, which is their autonomous driving. They have cars and robots. Did a deal with Uber. Not really that bullish on autonomous riding but in this market these things can catch premium valuations out of nowhere and it feels like it’s being valued as a 0. - Toloka a data labeling type business that sells this data for AI building. Not sure it’s big now but will be huge in the future. The fact they’re focused on this is great to see and they can sell it to many of their customers. - Tobiko. Learning platform online. Basically sell course to teach skills. - THINGS TO NOTE: - CEO has around 90% of net worth in this company. - Spin off Yandex. Had to sell at a fat discount of more than 50%. But basically cut out from under Yandez and Russia. No employees in Russia. Had a lot of yandex employees move out of Russia and work for him now. Condemned russias invasion of Ukraine. Did it late but said he couldn’t risk it because he was trying to get his employees out. - Were getting their business full spun off and separated etc and nasdaq basically told them they’re getting relisted next week. - Almost no analyst coverage. Maybe 1 or 2 people. As more analysts cover it, should get more attention from Investors.