I posted about this on the Berkshire Hathaway board, ajm101 replies on SNOW - hold forever?, and I'm just reposting here (slightly cleaned up):

Isn't that a big barrier for Oracle customers to switch to SF so it's unlikely SF will catch up with Oracle?

I work around the space, and I don't get a lot of opportunities where BRK investment discussion and technology overlap.

Snowflake has a number of academic papers floating around, a few of which are quite good (PDFs, view at your own risk)

* event.cwi.nl - Dageville sn... "The Snowflake Elastic Data Warehouse" (slides from talk accompany paper here, https://15721.courses.cs.cmu.edu/spring2018/slides...)
* https://www.usenix.org/system/files/nsdi20-paper-v... "Building An Elastic Query Engine on Disaggregated Storage"

Migration is tedious but it's one time in nature. It can be worth it in order to reduce long-term costs, which Snowflake does in ways the papers describe. Those papers are fairly accessible but technical, so to summarize, Snowflake optimizes storage efficiency (via compression in "blob storage", or s3-like systems), adapts established data warehousing techniques to work with that storage architecture, decouples storage from compute resources, and has invested a lot in very efficient utilization of those separate compute resources. To summarize further, they support extremely big data in generic cloud resources very efficiently (ie, inexpensively).

One difficulty in chipping away at Oracle's position had been the depth of the integration of their database with their CRM, ERP, financial, and ERM systems. Now that CRM/ERP/et al. data may reside in Oracle databases, but periodically gets copied into SNOW where analytical workflows can run without risking the dreaded Oracle license noncompliance or audits.

I don't know why BRK started the SNOW position, particularly given their dearth of tech investments. SNOW has a very high valuation and high risk of disruption (many competitors can emerge in the time necessary to grow into 20 P/S multiple). But I'd guess it came from BRK's own experience in the insurance operations, from analyzing the future market for data, SNOW's structural cost advantage, and the "stickiness" once SNOW acquired a customer.