Des

I am researching software supply chain optimization tools (think CI/CD pipelines, SBOM generation, dependency scanning) and want your take on the technologies behind them. I am comparing Discrete Event Simulation (DES) and Multi-Agent Systems (MAS) used by vendors like JFrog, Snyk, or Aqua Security. I have analyzed their costs and adoption trends, but I am curious about your experiences or predictions. Here is what I found. Overview: - Discrete Event Simulation (DES): Models processes as sequential events (like code commits or pipeline stages). It is like a flowchart for optimizing CI/CD or compliance tasks (like SBOMs). - Multi-Agent Systems (MAS): Models autonomous agents (like AI-driven scanners or developers) that interact dynamically. Suited for complex tasks like real-time vulnerability mitigation. Economic Breakdown For a supply chain with 1000 tasks (like commits or scans) and 5 processes (like build, test, deploy, security, SBOM): -DES: - Development Cost: Tools like SimPy (free) or AnyLogic (about $10K-$20K licenses) are affordable for vendors like JFrog Artifactory. - Computational Cost: Scales linearly (about 28K operations). Runs on one NVIDIA H100 GPU (about $30K in 2025) or cloud (about $3-$5/hour on AWS). - Maintenance: Low, as DES is stable for pipeline optimization. Question: Are vendors like Snyk using DES effectively for compliance or pipeline tasks? -MAS: - Development Cost: Complex frameworks like NetLogo or AI integration cost about $50K-$100K, seen in tools like Chainguard Enforce. - Computational Cost: Heavy (about 10M operations), needing multiple GPUs or cloud (about $20-$50/hour on AWS). - Maintenance: High due to evolving AI agents. Question: Is MAS’s complexity worth it for dynamic security or AI-driven supply chains? Cost Trends I'm considering (2025): - GPUs: NVIDIA H100 about $30K, dropping about 10% yearly to about $15K by 2035. - AI: Training models for MAS agents about $1M-$5M, falling about 15% yearly to about $0.5M by 2035. - Compute: About $10^-8 per Floating Point Operation (FLOP), down about 10% yearly to about $10^-9 by 2035. Forecast (I'm doing this for work): When Does MAS Overtake DES? Using a logistic model with AI, GPU, and compute costs: - Trend: MAS usage in vendor tools grows from 20% (2025) to 90% (2035) as costs drop. - Intercept: MAS overtakes DES (50% usage) around 2030.2, driven by cheaper AI and compute. - Fit: R² = 0.987, but partly synthetic data—real vendor adoption stats would help! Question: Does 2030 seem plausible for MAS to dominate software supply chain tools, or are there hurdles (like regulatory complexity or vendor lock-in)? What I Am Curious About - Which vendors (like JFrog, Snyk, Chainguard) are you using for software supply chain optimization, and do they lean on DES or MAS? - Are MAS tools (like AI-driven security) delivering value, or is DES still king for compliance and efficiency? - Any data on vendor adoption trends or cost declines to refine this forecast? I would love your insights, especially from DevOps or security folks!