Personal Finance
14h
2513
Should I marry a lazy girl?
Tech Industry
9h
412
Which AI company is the most prestigious in 2024?
Cars
15h
1519
Electric cars depreciate 10 times faster than gasoline cars
Tech Industry
19h
1505
Horrible Netflix ML interview experience
Tech Industry
21h
455
Please Help :: Should I pursue CS Masters in USA or remain in India to join Startups?
What valuation will you buy databricks stock at assuming it IPO today. How so? For reference: Snowflake at current price is worth 72B TC: 400k
What does Databricks do that can’t already be done?
DB never fails to trigger this putz from Dell.
This guy here is so salty about databricks. I would say do Leet hard next time and try again
$2B growing 60%+ beyotches. They hate us cuz they anus
What's your current serverless revenue?
Ooooo fight fight fight!
Databricks folks, you guys has two product lines: serverless and on-premise, the later is low margin business because you have to spend human resources to support customers with fragmented deployment. Look at confluent in the similar situation where their gross margin is insanely lower than industry avg. DBX won't enjoy the same multiplier compared with Snow. Snow P/S is 30ish, DBX would get 25ish at most, that gives you ceiling at 35B valuation. I am not saying you guys won't grow market cap, but there is long way to go to convert on-premise to cloud deployment.
We do not have an on premise offering. What are you talking about?
Customers managed deployment is called on premise dude.
We have Serverless = DB manages both cloud resource and the software compute, and Non-Serverless = customer manages and pays for cloud and pays DB only for software compute. Serverless means less absolute margin, but is suppose to be a better experience. Snowflake is all Serverless But the dichotomy is not Serverless or On-Prem (facepalm). Everything is on cloud
Deployment happening in customers account is costing you on maintenance, support, fragmented patches and so on. You see confluent don't reveal how much revenue coming from serverless, they just show subscription and cloud revenues which overlaps each other, because they know Wal Street don't like the non-severless business.
Let me repeat something I wrote in another thread. You gravely exaggerate the costs incurred. We have a very mature system to manage the spark clusters and we spin up dozens of thousands per week fully automated. And I think a significant part of our customers use cases will remain on this deployment model due to compliance constraints, isolation, etc.
Chill, databricks and snowflake now have a new common enemy which is called Palantir. According to recent earnings call, their commercial revenue is up 60% yoy
Eh. Commercial revenues were always dwarfed by their federal/pubsec business I thought - so 60% growth on that is less impressive than +60% revenues overall. Edit: just checked. You were looking at USA Commercial revenues which are just at $131m and were up 70%. Total WW commercial revenue was just +32%. Their total YoY growth is much smaller (20%) compared to DBX, SNOW, and others. Their bread and butter of govt revenues was up only 6%. If you’re betting on them you’re betting on an acceleration in Commercial when options are typically either Snowflake, Databricks, BigQuery, or any other tool that has a much wider talent pool.
The threat for both of us is that Palantir pays shit and Wal Street likes them. You smell something?
This is a dumb thread.
yes feed me more hopium
No dog in this fight. How does the AI story fit into DB product? DB acquired some LLM startup recently. Is Snowflake also trying to explore to AI space?
Both of them just want to ride the hype, they both are not AI or AI infra company. Snow is data management company, DBX is at core a larger confluent which wants to explore to ANY hype they find, first data, then LLM.
^^This is so gold 😂, awesome name
This thread is gold 👍🏻
30B Lemme tell you very clearly. Developer hate click op tools irrespective of how easy it is. I would bet on snowflake than DB. Technology should be serverless but deployment, state management and CI/CD should go through gitops rather than UI
Troll?
This is still high, use their serverless revenues of 100m annually {year oldish} then ramp up from a 30-50x multiplier {Snowflake's current} then add in some growth, maybe 200%? I'm at 15B max.