Developer Playgrounds & Calculators
A centralized suite of interactive systems sandboxes, VRAM estimators, and inference calculators built to size and optimize modern AI infrastructures.
🎨 Interactive Sandboxes
Visualize and test core LLM data-processing mechanisms in real-time.
Context Tokenizer
Estimate token weights, inspect input lengths, and visualize split frequencies before query submission.
RAG Chunking Sandbox
Simulate Recursive Character Splitting parameters. Inspect overlaps, segment boundaries, and token density maps.
Attention Map Simulator
Hover over token sequences to visualize how multi-head self-attention models calculate contextual word relevance.
Parallelism Split Planner
Model division tool. Plan layer distribution across Tensor Parallelism (TP) and Pipeline Parallelism (PP) ranks.
CIDR Subnet Calculator
AWS & GCP subnetting helper. Calculate IP ranges, netmasks, broadcast addresses, and hosts visually.
Attention Softmax Simulator
Mathematical step-through. Play with Query-Key dot products and dimension sizes to watch softmax scale.
Architecture IaC Designer
Design cloud systems visually. Configure target properties and export validated Terraform config scripts.
📐 Systems Calculators
Size hardware requirements, execution speedups, and compute trajectories.
VRAM Hardware Sizing Calculator
Model weight and GQA KV-cache sizing engine. Match parameters to GPU nodes (A100, H100, etc.) with custom buffers.
Agentic Loop Cost Simulator
Project API costs of multi-turn agentic loops. Simulates recursive token scale and context window inflation overheads.
Speculative Decoding Speedup
Calculate net speedup multipliers and latencies when combining high-performance draft and target model pairs.
Cloud GPU Instance Estimator
Estimate cloud hosting costs for LLMs by mapping memory requirements to AWS, GCP, and Azure GPU instances.
Systems Design & ML Quiz
Prepare for AWS & GCP certifications. Answer scenario questions with detailed explanation traces.