Hello!

Spring is here at TileDB, and we have updates to share! The biggest of which is our recent strategic investment from Verizon Ventures. We also enjoyed getting to meet some of you at Bio-IT World in Boston this month. Stay tuned for another TileDB appearance, coming soon.

Now, on to our updates for TileDB Cloud, TileDB Embedded, and more.

We’re making it easier to organize your work in TileDB Cloud, and are adding flexibility with new options for Python and R.

Groups

Asset groups are now available in TileDB Cloud to keep your projects nice and tidy. Groups allow you to gather together related arrays, notebooks, files, and other TileDB Cloud assets for simplified data management and sharing.

UDF language options: R & Python

R UDFs are now live in TileDB Cloud. Enjoy the same convenient registration, versioning, and parallelization features as Python UDFs. Speaking of Python UDFs, TileDB Cloud now supports multiple Python versions too: 3.8, 3.9, and 3.10.

Task graph management

You can now register entire task graphs to TileDB Cloud, accessible as their own data asset type. This design makes it easier to manage and share even the most complex task graphs encompassing many UDFs. Task activity logs now also capture the graph’s structure.

Performance improvements

The 2.8 release includes on-disk format adjustments, adding a lightweight commit consolidation mode that dramatically boosts performance on object stores. Both 2.7 and 2.8 feature improvements to compression filters.

Groups

Preparing asset groups to launch on TileDB Cloud also drove improvements to the pre-existing groups functionality in TileDB Embedded 2.8. Here, new features include group metadata, as well as versioning and time-traveling functionality.

New File API

Arbitrary file storage is a feature that was initially developed for TileDB Cloud, but we’re happy to announce that the 2.9 release has introduced native file support (modeled as 1D dense arrays) within TileDB Embedded. Along with groups, file support will provide a unified data management experience for users working with both our open-source and commercial offerings.

TileDB was recently covered in GenomeWeb for our work with Rady Children's Institute for Genomic Medicine. Free registration is required to access the article, “Rady Children's to Launch Consortium for Rapid Whole-Genome Sequencing of Newborns“.

We also recently hosted a webinar with our collaborators at Capella Space on analyzing LiDAR and SAR data. The blog contains the full video and recap. This session includes several example notebooks. See the blog for details on how to access them.

On the tutorial front, we published a new notebook on how to efficiently manage dataframes with TileDB. We cover rapid slicing, Apache Arrow integration, and SQL support. Download the public notebook, or sign up and launch it directly in TileDB Cloud's Jupyter environment.

We were thrilled to present with our partners at Spire Maritime (part of Spire Global) on AIS data management & time-series analytics using TileDB Cloud. The blog post has the recap and full video.

TileDB VP of Geospatial Norman Barker presented a lightning talk at the Open Geospatial Consortium’s Cloud-Native Geospatial Outreach Event. It’s a quick 6-minute preview of TileDB’s plans for spatial queries, while offering a single unified API for both vector and point data.

Finally, TileDB Founder & CEO Stavros Papadopoulos recently made the case for multi-dimensional arrays as a universal data model. It’s a must-read for anyone wanting to understand the underlying technical principles of TileDB Embedded and TileDB Cloud.

Visit our new website!

That’s all for now. In the meantime, check out the new tiledb.com, complete with:

• Fresh product detail pages for TileDB Embedded
• A new product page for TileDB Cloud
• New pages for supported data types and industries via the top menu bar

We look forward to your feedback.

Thank you,

— The TileDB Team