Andi Zimmerer

I hope you've had a great start to the year! I'm excited to announce our blog. We're kicking things off with a look back at everything that happened in 2025.

utndatasystems.github.io/blog/2025/re...

Can your database system predict underprovisining before it even happens?

Meet ◒ xBound, the very first framework for join size lower bounds. xBound tells you how many tuples your SQL query will produce *at least*.

Brought to you by @microsoft.com Gray Systems Lab & @utndatasystems.bsky.social.

This is hilarious. I wonder if the chocolate windmill still sits somewhere on a shelf or if it had been devoured in the process of trying to prove it

"we shall be happy to award a chocolate windmill to the first person..."

I love older papers.

Lenstra and Kan, 1979 "Computational Complexity of Discrete Optimization Problems" Annals of Discrete Mathematics

#orms

Andi Zimmerer | Pruning in Snowflake: Working Smarter, Not Harder Modern cloud-based data analytics systems must efficiently process petabytes of data residing on cloud storage. A key optimization technique in state-of-the-art systems like Snowflake is partition pru...

"The fastest way of processing data is to not process it."

Our SIGMOD 2025 paper shows how Snowflake skips 99.4% of data with new pruning techniques for LIMIT, top-k, and JOIN queries.

Blog: snowflakepruning.github.io
Paper: arxiv.org/abs/2504.11540

@sigmod2025.bsky.social

Paper clickbait

a woman is sitting at a table with a dart board in the background and says `` i 'm done '' . ALT: a woman is sitting at a table with a dart board in the background and says `` i 'm done '' .

Camera-ready version of the paper submitted => 115 tabs in Chrome closed.

GitHub - utndatasystems/redbench: Redbench is a set of 30 analytical SQL workloads that can be used to benchmark workload-driven optimizations. Redbench is a set of 30 analytical SQL workloads that can be used to benchmark workload-driven optimizations. - utndatasystems/redbench

We just released Redbench, a new benchmark that contains 30 analytical SQL workloads that can be used to benchmark workload-driven optimizations. Go check it out!

GitHub: github.com/utndatasyste...

This.

The first day of the BTW Conference in Bamberg is coming to an end.

Some personal favorites:
- Ismail's talk on Pruning in Snowflake
- @stefan-grafberger.com's talk on what-if analysis in ML pipelines and automatically patching ML pipelines in the background
- Observe Inc's presentation

Please help spread the word by reposting!

We've just created the official DEEM Workshop account: @deem-workshop.bsky.social

The review comment I'm most proud of: "The paper is [...] a pleasure to read". Thanks anonymous reviewer 🙏😊

My very first paper got accepted to @sigmod2025.bsky.social! Yay! Means I'll be playing a home game in Berlin

Strong agree. It's just that the "normal" model starts to cause pain on high velocity teams only

I agree that something has to improve, following the points you outlined. But I also see that c++ has a large path dependency, so change has to be done carefully. I still prefer Rust, but I like the overall discussion around the future of C++

Thanks for sharing, that was an interesting read!

Currently reading through your post.
Nit: typo in "falls in the “contacts” and “profiles” camps of Steve’s list above."

Valid points! I'm still curious what the next decade of C++ will look like

Reading cacm.acm.org/blogcacm/21s... makes me think that Rust was just a giant research project and valuable findings are now being streamed back into C++, making them usable to a broader audience.

Agreed. Buddy C++ goes undefined behavior if it's not a strict weak ordering and I've seen it crashing a couple of times as well.

47 *is* really good! It very much depends on age, gender and body weight - I am just fortunate enough to be in a lucky group

Garmin showing a VO2 Max of 58

My professor jokingly threatened me that I would get fired if my VO2 Max is too low. After a run with him it's at 58 now. I guess I can continue my PhD 😋

Happy for everyone in that room who got the chance to listen to Ismail. He's smart, pragmatic, and inspiring

And it took only a tiny amount of month-long convincing given that the other platform is completely broken.

The Nuremberg Data Systems Lab is now on Bluesky 🙌 @utndatasystems.bsky.social

In academia, everyone always has a Colleague Working On Exactly This Problem. I still have to find one. Applications open.

Exciting News! 🎉
#Tampere will host EDBT/ICDT 2026! ✨
Even before the 2025 edition, the important dates are already out:
📅 Round 1 starts on:
February 5 for EDBT Papers
March 13 for ICDT Papers
edbticdt2026.github.io

We can’t wait to see your great submissions and welcome you to Tampere! 🙌

The @sigmod2025.bsky.social Programming Contest goes into another round. We (Bo Tang, Tilmann Rabl, and myself) just published the timeline and task overview:
sigmod-contest-2025.github.io/index.html

Thanks to Carlo Curino and @microsoft.com for the continued support.

If I’m ever a professor again, I want to give a graduate seminar, topics to include:

- how not to say stupid shit about fields outside your expertise
- what is your expertise, anyway?
- how not to be an insufferable bore
- your PhD doesn’t make you a better person: coping with that

Other ideas?

I love how dedicated some students are. They are supposed to create a 5min video about a topic and one of them sends me their slides for review. 15(!) detailed(!) slides(!)