Quantum-centric HPC starts with user enablement
Quantum-centric HPC adoption is a user-enablement problem first. Hardware access alone does not produce useful workloads. Portable runtimes and schedulable resources do.
Short essays and notes on software, quantum, AI, and the engineering practice behind them.
Rebuilding Lakera’s Gandalf prompt-injection game with a plain LLM to make failure modes concrete and highlight input/output controls.
When microservices help, when they hurt (always), and how to avoid a distributed monolith by defaulting to a modular monolith.
Why production outcomes lag demos, and how evaluation, data boundaries, and cost control determine adoption.
Quantum-centric HPC adoption is a user-enablement problem first. Hardware access alone does not produce useful workloads. Portable runtimes and schedulable resources do.
AI productization fails when demos are mistaken for systems. Adoption slows when evaluation, data boundaries, and cost controls are missing. Build operating discipline to make prototypes reliable.
Microservices technical debt accumulates quickly. Weak domain boundaries create a distributed monolith with heavy coordination costs. Default to fewer interfaces until scale proves service splits.
software, architecture, microservices
Read the postShipping early in deep tech reduces execution risk. Waiting for full clarity creates demo-ware and long feedback loops. Deliver thin end-to-end slices to validate constraints early.
Prompt injection defenses must be operational. Treating LLMs as untrusted components exposes controls needed for policy, cost, and audit. Build gateway guardrails before scaling AI features.
Prompt injection becomes obvious in Lakera’s Gandalf game. System prompts alone fail once user text is treated as trusted context. Test reinjection paths to design safer LLM interactions.
LLM security starts with prompt trust boundaries. Prompt injection and multimodal inputs bypass instruction-only defenses. Add controls on inputs, tools, and outputs to reduce exploitability.
Travel tech ecosystem complexity breaks naive AI plans. Search, pricing, inventory, and support pipelines conflict in practice. Set tight system boundaries so AI improves outcomes.
Abstraction and performance are not enemies. Clean interfaces fail when data layout and optimization boundaries are ignored. Place optimizations at the right layer to keep code fast and maintainable.
mathematics, quantum, quantum-computing
Read the postProject architecture starts with iterative requirements. Treating greenfield design as one-shot creates brittle code and stakeholder drift. Use small PoCs to align scope, ownership, and evolvability.
Quantum software architecture debates at IEEE QCE are maturing. Performance, usability, and quality attributes still get mis-scoped. Use clearer boundaries to make trade-offs explicit.
Quantum-HPC integration at ISC-HPC shows real momentum. European programs still miss near-term algorithm work and rush standards too early. Focus on deployable workflows to improve impact.