> Oracle plans thousands of job cuts as data center costs rise, Bloomberg News reports

With how many Wog hires to replace the fired Americans?

wrote:

Perhaps we don't need all these data centers?

yes very much like the .com era - basically spend as much as you can until you run out of money - hoping you can become the biggest player.

VP_Spiro_T_Cheney wrote:

AI will be deemed essential and too big to fail, necessitating a taxpayer bailout within 4 years

AI will become comoditized.

wrote:

AI will become comoditized.

And a low margin commodity can’t cover the debt service on a trillion dollars in capex spending

wrote:

Which one of the AI services is worth $200 per year in "pro" usage?

The usage cap is a practical limit on the model's capabilities, more than anything. I am seeing companies burning $1000s/month on tokens without gaining much improvement in the results.

wrote:

Perhaps we don't need all these data centers?

I asked Claude and he said they had to have them.

Phillip_McCavity wrote:

I asked Claude and he said they had to have them.

Hegseth disagrees.

About those data center buildouts:

https://x.com/Gaurab/status/2046755907137413440

wrote:

Does anyone still use Oracle for their data warehouse?

Does Oracle have power to run their data warehouse?  We are talking about Oracle, who would turn the lights out on data warehouses to provide power to AI.

wrote:

Does anyone still use Oracle for their data warehouse?

Only 95% of the Fortune 100 and every government agency in the country.

wrote:

Only 95% of the Fortune 100 and every government agency in the country.

The same Fortune 100 who would dump Oracle at the first opportunity?
Nobody in their right mind wants to deal with Oracle.

wrote:

Those contracts are ironclad. Larry knows what he is doing.

Options exist:

Apache Doris, ClickHouse, and Databend are leading open-source data warehouse solutions that provide scalable, cost-effective alternatives to proprietary cloud platforms like Snowflake.  These tools utilize Massively Parallel Processing (MPP) or columnar storage architectures to deliver high-performance real-time analytics and support for complex SQL queries.

Key open-source options include:

Apache Doris: A modern MPP database designed for real-time analytics with easy usability and high performance, supporting both OLAP queries and search capabilities.

ClickHouse: A fast, open-source column-oriented DBMS optimized for real-time analytical data reports using SQL, capable of handling billions of rows per second.

Databend: A cloud-native, open-source data warehouse focused on cost-efficiency and scalability, offering a Snowflake-like experience for data-driven applications.

StarRocks: An open-source distributed OLAP database designed for sub-second analytics, supporting major open table formats like Apache Iceberg, Hudi, and Delta Lake.

DuckDB: An embedded, in-process OLAP database ideal for "tiny" data warehouses or local analytical workloads, often compared to an OLAP version of SQLite.
For smaller scale or lightweight needs, PostgreSQL combined with tools like dbt for transformations and Airflow for orchestration is frequently recommended as a dependable, simple solution.  Other notable tools in the ecosystem include Apache Hive, Presto, Apache Kylin, and Greenplum, which cater to various big data and batch processing requirements.