managers need to work with stakeholders representing the requirements, systems engineering, contracting, cost estimating, and testing communities to design processes around short releases. Acquisition executives a contract strategy, shaping systems engineering processes, managing requirements, and developing cost estimates for programs with a dynamic scope. Experience indicates that cultural changes must .................................................................................. 33 11 Cost Estimation ............................................................................................

stories to concisely define the desired system functions and provide the foundation for Agile estimation and planning. o They describe what the users want to accomplish with the resulting system. User continue to refine it as they learn more from the development sprints and releases. Cost Estimation  Cost estimation in an Agile environment is challenging  The challenge within DoD is often a resistance detailed release and sprint-level estimates as requirements become better defined. The fidelity of the cost estimate increases once a development team is established to help estimate the level of work for

easily”  Experiment: Economist Subscription, Errands for Pen and Suit  Agile tie-in: Size and cost estimation – does anyone really know in absolute terms how many SLOC it will take a build a product or capability, or is it more in keeping with our nature to size something relatively to something else The Cost of Social Norms – Why We Are Happy to Do Things, but Not When We Are Paid to Do Them  Example:

devices and models >> 2 HW Platforms ZCU102 ZCU104 Ultra96 PYNQ Face detection Pose estimation Video analytics Lane detection Object detection Segmentation

numeric values that cannot be represented effectively in native hardware representations, but at the cost of relatively slower performance. The following are Julia's primitive numeric types: • Integer types: functions Generated functions can achieve high efficiency at run time, but come with a compile time cost: a new function body must be generated for every combination of concrete argument types. Typically strongly recommended for interoperable code. In this case, there will be no hidden arguments, at the cost of some language features (e.g. only character(len=1) will be permitted to pass strings). Note

numeric values that cannot be represented effectively in native hardware representations, but at the cost of relatively slower performance. The following are Julia's primitive numeric types: • Integer types: functions Generated functions can achieve high efficiency at run time, but come with a compile time cost: a new function body must be generated for every combination of concrete argument types. Typically strongly recommended for interoperable code. In this case, there will be no hidden arguments, at the cost of some language features (e.g. only character(len=1) will be permitted to pass strings). Note

numeric values that cannot be represented effectively in native hardware representations, but at the cost of relatively slower performance. The following are Julia's primitive numeric types: • Integer types: functions Generated functions can achieve high efficiency at run time, but come with a compile time cost: a new function body must be generated for every combination of concrete argument types. Typically strongly recommended for interoperable code. In this case, there will be no hidden arguments, at the cost of some language features (e.g. only character(len=1) will be permitted to pass strings). Note

numeric values that cannot be represented effectively in native hardware representations, but at the cost of relatively slower performance. The following are Julia's primitive numeric types: • Integer types: functions Generated functions can achieve high efficiency at run time, but come with a compile time cost: a new function body must be generated for every combination of concrete argument types. Typically strongly recommended for interoperable code. In this case, there will be no hidden arguments, at the cost of some language features (e.g. only character(len=1) will be permitted to pass strings). Note

Inference Costs Per Token Falling = Performance Converging + Developer Usage Rising • AI Usage + Cost + Loss Growth = Unprecedented • AI Monetization Threats = Rising Competition + Open-Source Momentum Rising + Inference Costs Per Token Falling = Performance Converging + Developer Usage Rising 3 Cost of Key Technologies Relative to Launch Year % of Original Price By Year (Indexed to Year 0) Note: International Federation of Robotics Industrial Robots Installed Details on Page 289 AI Usage + Cost + Loss Growth = Unprecedented 4 Leading USA-Based AI LLM Revenue vs. Compute Expense Note: Figures

In the past: We viewed EA as primarily concerned with standardization, consistency, planning, and cost reduction. It documented as-is and to-be architectures, demonstrated alignment of systems with business something you can easily roll in any direction you choose. That is its latent value; the inverse of the cost of implementing changes in the future. What does such an EA, brimming with latent value, look like that product. The cost of custom development is falling: More and more logic is abstracted away by frameworks and design patterns. Incremental delivery and staged investments reduce cost and risk.  Custom