HEAVY.AI

I've been diving deeper into the Foursquare Places dataset that was released two weeks ago, and am finding it great not only for accessing individual POIs, but also for performing large-scale analysis of geographic trends. Here I'm using HEAVY.AI to join the Places dataset to the NASA Gridded Population of the World (GPW) dataset, containing population estimates per square km for the entire world, using Uber H3 hexagonal spatial indexes as join keys. This allows for easy calculation and visualization of the number of POIs by category per capita. See below for two visualizations comparing the number of places of worship with the number of bars per 1,000 people. You can easily see the Bible belt light up for the former category, but why the high prevalence of both places of worship and drinking establishments in the Upper Midwest?

It truly was an awesome surprise this morning to see that Foursquare had open sourced over 100M Points-of-Interest (POIs) from its Places dataset. Having such a dataset under a permissive Apache license is going to be a huge boon to developers and open geo enthusiasts alike, as currently there is little high-quality open POI data available. Curious, I pulled the Parquet files from S3 and loaded them into HEAVY.AI, which took about 10 minutes, after which I spent another 10 minutes or so building a dashboard in Heavy Immerse to start interactively exploring the data. As you can see in the quick screen recording here, the dataset is quite stunning. I did a quick sanity check of the number of global McDonald's (roughly 40K), and then even drilled down to the streets of SF to find the locations of my favorite food truck (Señor Sisig !). Major kudos to the Foursquare Places team for their significant contribution to the open geospatial community, and we look forward at HEAVY.AI to exploring this data more and using it to enrich other spatial datasets.

Check out our new 20 billion record AIS ship location demo, running on a single-GPU NVIDIA Grace Hopper GH200 instance on Vultr Cloud!

If you are attending #SC24 in Atlanta, we'd love to show you the combined power of HEAVY.AI and NVIDIA Grace Hopper running on Vultr. Swing by to see how you can visualize your largest geospatial datasets.

I'm excited to present tomorrow (10/31) at 10AM PT/1PM ET with NVIDIA and Vultr on the huge performance gains we're seeing at HEAVY.AI running on the new Nvidia Grace Hopper Superchip. With an ultra-fast NVLink interconnect between CPU and GPU, we no longer have to move data over the "thin straw" of PCIe, dramatically accelerating large-scale data analytics workloads. Join us at the webinar to learn more! https://lnkd.in/egzdU2pG

The extraordinary performance of HEAVY.AI starts from its core engine, the HeavyDB GPU-accelerated analytics database. In this blog post, our Founder and CTO Todd Mostak walks through recent benchmarks we conducted of the system, both Star Schema Benchmark (SSB) and TPC-H, showing the significant performance advantages of the system over CPU-based data warehouses, in some cases returning results over 100X faster.  HeavyDB is tested on an NVIDIA 8XA100 and the new ARM-based Grace Hopper GH200 system, and compared against Snowflake, Databricks, Google BigQuery, and Amazon Redshift. Special thanks to our close partner Vultr for providing the GPU hardware for testing. Read more in the blog and let us know your thoughts! https://lnkd.in/eYY39z9e

Just published my first Forbes piece, digging in a bit on transparency, explicability and how they relate to genAI and SQL. SQL is far from the only choice in constraining an LLM, but I argue that it's the best one. Not coincidentally, that's what we've implemented at #heavyai. https://lnkd.in/gHSACmfp

🎯 Unlock Next-Level Analytics Performance with Vultr’s NVIDIA Grace Hopper GPU Instances and HEAVY.AI Join us for this webinar where our Founder & CTO, Todd Mostak will co-present with Vultr and NVIDIA to dive deep into how Vultr’s NVIDIA Grace Hopper GPU instances have turbocharged HEAVY.AI’s analytics platform to deliver blazing-fast, cost-effective, and scalable insights for massive datasets. 🔥 Real-time analytics at this speed and scale can be a game-changer for your business. 🗓 Date: Thursday, October 31, 2024 ⏰ Time: 10:00-10:45am PT Don’t miss out on this opportunity to revolutionize your analytics capabilities! 👉 Register here: https://lnkd.in/gw_Tif2C #GPU #DataAnalytics #AI #CloudComputing #TechWebinar #Vultr #NVIDIA #HeavyAI #RealTimeData

The Taylor Geospatial Institute is leading a collaborative approach to applying geospatial reseach and innovation to solve complex global challenges. We are pleased to be part of TGI's Industry Partner Program which fosters the exchange of knowledge, tools and data between their institute and industry. We are pleased to have Dr. Michael Flaxman will be presenting to TGIs community on October 22nd.

We launched HeavyIQ earlier this year to our users to effortlessly generate SQL and visualizations in the HEAVY.AI platform using natural language. But now with the new LLM_TRANSFORM SQL operator, you can use the power of the HeavyIQ Large Language Model to add fuzzy intelligence to your SQL itself. Read this blog post from our founder and CTO Todd Mostak on how HeavyIQ and this new SQL AI operator can be used for a wide range of analytics and data cleaning tasks, such as correcting malformed date strings, categorizing data, and sentiment analysis. https://lnkd.in/e9fZfsWp