Twitter | Search | |
Rodrigo Aramburu
CEO , a GPU SQL engine that connects data lakes to AI.
308
Tweets
530
Following
536
Followers
Tweets
Rodrigo Aramburu retweeted
BlazingSQL Jul 22
This single query joins an File, a , and a Dataframe () --- and persists a GDF in memory for the rest of the ecosystem to go to work on! Try it free:
Reply Retweet Like
Rodrigo Aramburu retweeted
Mark Harris Jul 21
A ton of hard work went into the latest RAPIDS release. Hope you enjoy it!
Reply Retweet Like
Rodrigo Aramburu Jul 16
This is all part of our mission of unifying multiple disparate data sets into a seamless SQL experience that is fast and fully integrated with the , , and ecosystem.
Reply Retweet Like
Rodrigo Aramburu retweeted
BlazingSQL Jun 28
Our own leading a Masters class in . Explaining the benefits of formats and stressing that accelerated are the way to scale to data needs of the future. Foundational for next-gen data science.
Reply Retweet Like
Rodrigo Aramburu retweeted
BlazingSQL Jun 24
We just started our team retreat at our office! Engineering team huddled discussing future features of engine. Warning includes sneak preview of sexy new logo! Shhh 🤫🤫🤫🤫🤫🤫🤫🤫🤫🤫🤫
Reply Retweet Like
Rodrigo Aramburu retweeted
RAPIDS AI Jun 25
Lots of data and fast processing aren’t enough in today. Senior Data Scientist, discusses feature engineering and understanding your data to build proper features for better training
Reply Retweet Like
Rodrigo Aramburu retweeted
Adam Breindel Jun 25
Want to scale out w and Jupyter in the cloud, without messing with AWS, AMIs, etc? will have you running in seconds. They do the terraform and k8s stuff, so you don’t have to! And they support GPU instances for stuff like deep learning and !
Reply Retweet Like
Rodrigo Aramburu retweeted
BlazingSQL Jun 24
Why we have been running on : 1. Free: computing isn't cheap, let alone free 2. Powerful: Tesla are ideal for many workloads 3. Easy: + up installed + running in two minutes Try Now:
Reply Retweet Like
Rodrigo Aramburu retweeted
RAPIDS AI Jun 12
Try early access implementation on Dataproc today! Follow our instructions to scale out your accelerated workloads to clusters.
Reply Retweet Like
Rodrigo Aramburu May 29
Want to understand the plight of a economy for those w/out bank accounts. Travel through China with no or . I just wanted a bottle of water! Please enable for foreign credit cards!
Reply Retweet Like
Rodrigo Aramburu May 23
Have you tried to build a product or business on open CL? We have and failed. Many of our partners started out on opencl too for the reasons you mentioned, which are totally reasonable, but had to ultimately switch to CUDA as codebase grew in complexity.
Reply Retweet Like
Rodrigo Aramburu May 20
Can you "add to my drive"? I can check the settings to see if I can enable that. Lmk
Reply Retweet Like
Rodrigo Aramburu retweeted
BlazingSQL May 15
Just released a blog post about how you can process 20M rows with fast 100X faster than on . Check out the power of w/ GPUs!
Reply Retweet Like
Rodrigo Aramburu retweeted
Samsung NEXT May 15
Check out @blazingdb's latest blog post on how you can process 20M rows with fast 100X faster than on :
Reply Retweet Like
Rodrigo Aramburu retweeted
BlazingSQL May 15
Best part: It's free! High-performance computing and with + on 🔥🔥🔥 Thanks !
Reply Retweet Like
Rodrigo Aramburu retweeted
Kyle Weiss May 15
Why this is cool: 1. It's free: High performance GPUs are not cheap, let alone free (T4 16gb vRAM) 2. It's easy: Full setup process for is 3 commands in < 2 mins Standard & interface 3. It's insanely fast: ETL 20m rows in 1.21 seconds (100x>)
Reply Retweet Like
Rodrigo Aramburu May 15
The more I play with the more I think that this opens up analytics and dev for so many new users. It's crazy easy to start , just open a file and boom, fully functional dev environment. Now with you can query at scale mad fast.
Reply Retweet Like
Rodrigo Aramburu retweeted
NVIDIA AI May 15
At STAC Summit in London today, members discussed the newest benchmark for the backtesting of trading strategies. Learn how the power of DGX-2 & Python, which use CUDA-X software w/ , is delivering more than 6000x speedup:
Reply Retweet Like
Rodrigo Aramburu retweeted
Joshua Patterson May 13
Second big news of the day, is on ! If you want to try multi-node multi-GPU on or you now can, easier than ever, on nodes. We started with , now on Dataproc for scale. Try it today!
Reply Retweet Like
Rodrigo Aramburu retweeted
Joshua Patterson May 13
First big announcement, community, blows away record with & on ! Great work everyone! . 6250x is pretty ridiculous.
Reply Retweet Like