AI in Manufacturing: moving AI from Idea to Execution
AI and HPC convergence is here and is here to stay and accelerate innovations across industries. The increased availability of data, hardware advancements leading to increased computational capabilities, and new algorithms and mathematical models have collectively resulted in the accelerated AI expansion in all sorts of applications. This, however, creates high computational needs which naturally have been more and more successfully addressed by HPC (High-Performance Computing). In that sense, AI & HPC complement each other.
HPC infrastructure is often used to train AI’s powerful algorithms by leveraging huge amounts of sample data (training set) and in that way enables AI models (trained algorithms) to recognize shapes, objects (machine vision), find answers hidden in the data (predictive maintenance, data analytics) or accelerate time to results (predict the outcome of complex engineering simulations).
We at byteLAKE have been closely working with Lenovo, Lenovo Infrastructure Solutions Group, Intel Corporation, NVIDIA and many more to ensure that our AI-powered products not only help our clients efficiently automate various operations and reduce time and cost but also are highly optimized and make the most of the hardware and software infrastructure where they are deployed. Besides our efforts in bringing AI solutions to the paper industry and manufacturing in general (which I described in my previous post), our efforts in bringing value thru AI in the chemical industry highly benefit from HPC’s capabilities to dynamically scale and keep up with performance requirements.
Our product, CFDSuite (AI-accelerated CFD) leverages HPC to efficiently analyze historic CFD simulations and makes it possible for our clients to predict their outcomes on various edge devices i.e. laptops, desktop PCs or local edge servers.
And with that in mind, I am very happy to see the byteLAKE team becoming one of the drivers of AI & HPC convergence and leveraging it to bring innovations to various industries.
Links:
- byteLAKE’s Cognitive Services: www.byteLAKE.com/en/CognitiveServices (Cognitive Services (AI for Paper Industry & Manufacturing)). Related blog post series: www.byteLAKE.com/en/CognitiveServices-toc
- byteLAKE’s CFD Suite (CFD Suite (AI-accelerated CFD)): www.byteLAKE.com/en/CFDSuite. Related blog post series: www.byteLAKE.com/en/AI4CFD-toc
- HPC scalability report: https://marcrojek.medium.com/bytelakes-cfd-suite-ai-accelerated-cfd-hpc-scalability-report-25f9786e6123 (full report: https://www.slideshare.net/byteLAKE/bytelakes-cfd-suite-aiaccelerated-cfd-hpc-scalability-report-april21)
- byteLAKE’s CFD Suite (AI-accelerated CFD) — product community: www.bytelake.com/en/AI4CFD-pt2 (LinkedIn and Facebook groups)
#AI #IoT #Manufacturing #Automotive #Paper #PaperIndustry #ChemicalIndustry #CFD #FluidDynamics #OpenFOAM #ArtificialIntelligence #DeepLearning #MachineLearning #ComputerVision #Automation #Industry40
