Skip to content

Latest commit

 

History

History
58 lines (48 loc) · 3.06 KB

README.md

File metadata and controls

58 lines (48 loc) · 3.06 KB
Raw

AI Core Station ^ AI Accelerators and a naive guide to AI hardware:

Here is a small guide I wrote while building my lab's AI workstation @ DigiLab, LSBG and relevant information I learned while building the system are documented. I will cover 5 topics in this blog and here is an overview diagram of the workstation:


Topics:

  • AI/ ML / DL Workstation Build
  • Robotics, Electronics and Hardware
  • AI Accelerators
  • Benchmarking GPU Performance
  • Benchmarking Standard DL tasks

#1 AI/ ML / DL Workstation Build

  • Essence of Computation & Networking NOTES
  • AI / ML / DL hardware core components NOTES
  • Parallel Programming with CUDA NOTES
  • Final AI Work Station build NOTES

#2 Robotics, Electronics and Hardware

  • Electronics components & workbench NOTES
  • Robotics Components NOTES
  • Robotics and Mechatronics NOTES
  • IoT & AI on Edge devices NOTES
  • 3D printing & PCB Design NOTES
  • DIY & Advanced Robotics hardware NOTES
  • Robot Operating System [ROS] NOTES
  • Autonomous Systems NOTES

#3 AI Accelerators

  • AI accelerators NOTES
  • Future Computing NOTES
  • Mobile AI workstations/ laptops NOTES

#4 Benchmarking System Performance

To know more about linux : you might find useful content in my repository/@linux_d

  • 3D Mark Benchmark [x report] as no test suite for linux but windows
  • Phoronix test suite report ✅✅
  • Paessler Network benchmark link only windows sys (for linux - phoronix/ethr or phoronix/speedtest-cli)
  • LAMMPS benckmark link ( classical molecular dynamics in phoronix )
  • VASP benchmark link ( atomic scale molecular modelling in phoronix)
  • V-Ray 5 benchmark link ( render latency benchmark in phoronix)

#5 Benchmarking Standard DL tasks

To know more about AI : you might find useful content in my repository/@aicore_n

  • AI environment setup with CUDA, cudnn and python3 in linux report
  • Stanford DAWN Deep Learning Benchmark [x report] - retired benchmark and upgraded to mlperf
  • MLPerf training benchmarks report
  • AI-Benchmark Alpha / Burnout report

@lab : lab_video_draft_22"