Download as PDF, PPTX
Uploaded byAlluxio, Inc.
Flexible and Fast Storage for Deep Learning with Alluxio
This document discusses how Alluxio provides fast and flexible storage for deep learning workloads. It summarizes Alluxio's capabilities to accelerate data processing and machine learning workflows by enabling data to be stored, cached, and processed directly in memory across distributed environments. Alluxio uses a unified namespace and intelligent caching to provide high-speed data access to remote data sources and overcome storage bottlenecks.
More Related Content
Similar to Flexible and Fast Storage for Deep Learning with Alluxio
![Vibe Coding vs. Spec-Driven Development [Free Meetup]](https://cdn.slidesharecdn.com/ss_thumbnails/vibecodingvsspecdrivendevelopment-251209105622-43f455e7-thumbnail.jpg?width=640&height=640&fit=bounds)
![Coded Agents – with UiPath SDK + LangGraph [Virtual Hands-on Workshop]](https://cdn.slidesharecdn.com/ss_thumbnails/codedagentsdeck-251215155422-5497c599-thumbnail.jpg?width=640&height=640&fit=bounds)
Flexible and Fast Storage for Deep Learning with Alluxio
- 1. Mike Wendt @mike_wendt FLEXIBLE ANDFAST STORAGE FOR DEEP LEARNING WITH ALLUXIO Yupeng Fu
- 2. 2 ACCELERATE THE DEEPLEARNING STACK GPU-Acceleration by NVIDIA and Fast Storage from Alluxio Apache Arrow +
- 3. 3 DATA PROCESSING EVOLUTION FasterData Access Less Data Movement Store Read HDFS Write HDFS Read HDFS Write HDFS Read Query ETL ML Train Hadoop Processing, Reading from disk
- 4. 4 DATA PROCESSING EVOLUTION FasterData Access Less Data Movement HDFS Read HDFS Write HDFS Read HDFS Write HDFS Read Query ETL ML Train HDFS Read Query ETL ML Train Hadoop Processing, Reading from disk 25-100x Improvement Less code Language flexible Primarily In-Memory Storage bottleneck Spark In-Memory Processing
- 5. 5 25-100x Improvement Less code Language flexible PrimarilyIn-Memory DATA PROCESSING EVOLUTION Faster Data Access Less Data Movement HDFS Read HDFS Write HDFS Read HDFS Write HDFS Read Query ETL ML Train HDFS Read Query ETL ML Train HDFS Read GPU Read Query CPU Write GPU Read ETL CPU Write GPU Read ML Train 5-10x Improvement More code Language rigid Substantially on GPU Storage still a bottleneck GPU/Spark In-Memory Processing Hadoop Processing, Reading from disk Spark In-Memory Processing
- 6.
- 7. 7 ACCELERATE THE DEEPLEARNING STACK GPU-Acceleration by NVIDIA and Fast Storage from Alluxio Apache Arrow +
- 8. 8 APP A GPU-ACCELERATED ARCHITECTURETHEN Too much data movement and too many different data formats CPU GPU APP B Read DataH2O.ai Anaconda Gunrock Graphistry BlazingDB MapD Copy & Convert Copy & Convert Copy & Convert Load Data APP A GPU Data APP B GPU Data
- 9. 9 GPU-ACCELERATED ARCHITECTURE NOW Singledata format and shared access to data on GPU CPU GPU GPU MEM Read DataH2O.ai Anaconda Gunrock Graphistry BlazingDB MapD Load Data Apache Arrow Powered by: GPU Data Frame
- 10.
- 11. 11 GPU ACCELERATION ACROSSTHE ECOSYSTEM Apache Arrow
- 12. 12 25-100x Improvement Less code Language flexible PrimarilyIn-Memory DATA PROCESSING EVOLUTION Faster Data Access Less Data Movement HDFS Read HDFS Write HDFS Read HDFS Write HDFS Read Query ETL ML Train HDFS Read Query ETL ML Train HDFS Read GPU Read Query CPU Write GPU Read ETL CPU Write GPU Read ML Train Alluxio Read Query ETL ML Train 5-10x Improvement More code Language rigid Substantially on GPU 25-100x Improvement Same code Language flexible Primarily on GPU Alluxio Fast and Flexible Storage End to End GPU Processing (GoAi) GPU/Spark In-Memory Processing Hadoop Processing, Reading from disk Spark In-Memory Processing
- 13. 13 ACCELERATE THE ENTIREANALYTICS STACK GPU-Acceleration by NVIDIA and Fast Storage from Alluxio Apache Arrow +
- 14. 14 DATA & AIECOSYSTEM EXPLODES … • Many Compute Frameworks • Many Storage Systems • Most not co-located …
- 15. 15 Data & AIEcosystem Issues • Each app manages multiple data sources • Data source changes require global updates • Storage optimizations requires app change • Poor performance due to lack of locality … …
- 16. 16 Data & AIEcosystem with Alluxio • Apps only talk to Alluxio • Simple Add/Remove • No App Changes • Highest performance in Memory Java File API HDFS Interface Amazon S3 Interface REST Web Service HDFS Interface Amazon S3 Interface Swift Interface NFS Interface … …
- 17. 17 Storage Challenges forDL& ML 2 Data Freshness • Cross-network movement is slow • Copies create lag • Data quality suffers with copies 4 Security & Governance • Data security & governance is increasingly complex 1 Speed & Complexity • Integration and interoperability issues (on prem, hybrid, cloud) • Many departments & groups 3 Cost • Cloud storage is cheap and reliable, but slow • Data duplication 17 Heavy integrations create painful organizational drag
- 18. 18 Alluxio Design Principles 2Data Sharing • Don’t own the data • Multiple apps sharing common data • Data stored in multiple, hybrid systems 4 Enterprise Class • Distributed architecture • Commodity hardware • Service-oriented • High availability • Security 1 Big Data & Machine Learning • Interoperability with leading projects • Large scale data sets • High IO 3 High Speed Data Access • Remote data • Hot/warm/cold data • Temporary data • Read/write support 18
- 19.
- 20. 20 Filesystem in Userspace(FUSE) Running file system code in user space
- 21. 21 Alluxio Innovation: Unified Namespace Enableseffective data management across different Under Stores Uses Mounting with Transparent Naming
- 22. 22 Alluxio Innovation: Unified Namespace Createa catalog of available data sources for Data Scientists /finance/customer-transactions/ /finance/vendor-transactions/ /operations/device-logs/ /operations/phone-call-recordings/ /operations/check-images/ /research/us-economic-data/ /research/intl-economic-data/ /marketing/advertising-dataset/ /marketing/marketing-funnel-dataset/ alluxio://
- 23. 23 Alluxio Innovation: Intelligent Cache Localperformance from remote data using native multi-tier storage RAM SSD HDD Hot Warm Cold
- 24. 24 Deep Learning InputPipeline Deep Learning training involves three stages of utilizing different resources: • Data reads (I/O): e.g. choose and read image files from source. • Data Preprocessing (CPU): e.g. decode image records into images, preprocess, and organize into mini-batches. • Modeling training (GPU): Calculate and update the parameters in the multiple convolutional layers
- 25. 25 Alluxio overcomes I/Obottleneck
- 26. 26 Alluxio Architecture Alluxio Master Zookeeper Standby Master Alluxio Worker Alluxio Worker Under Store UnderStore Alluxio Client Application RAM / SSD / HDD RAM / SSD / HDD Control Path Data Path
- 27. 28 Read data inAlluxio, on same node as client Alluxio Worker RAM / SSD / HDD Memory Speed Read of Data Application Alluxio Client Alluxio Master
- 28. 29 Read data notin Alluxio + Caching 29 RAM / SSD / HDD Network / Disk Speed Read of Data Application Alluxio Client Alluxio Master Alluxio WorkerUnder Store
- 29. 30 Read data inAlluxio, not on same node as client + Caching RAM / SSD / HDD Network Speed Read of Data Application Alluxio Client Alluxio Master Alluxio Worker RAM / SSD / HDD Alluxio Worker
- 30. 31 Write data onlyto Alluxio on same node as client Alluxio Worker RAM / SSD / HDD Memory SpeedWrite of Data Application Alluxio Client Alluxio Master
- 31. 32 Write data toAlluxio and Under Store synchronously RAM / SSD / HDD Network / Disk SpeedWrite of Data Application Alluxio Client Alluxio Master Alluxio Worker Under Store
- 32. 33 100+ known productiondeployments AND MORE!
- 33. Twitter.com/alluxio Linkedin.com/alluxio Website www.alluxio.com E-mail info@alluxio.com @ Social Media á ™ 34 Thank you! YupengFu yupeng@alluxio.com Github: yupeng9 Twitter.com/alluxio Linkedin.com/alluxio Website www.alluxio.org E-mail info@alluxio.com @ Social Media á ™