Known stays known.
Measurements and specifications you provide are preserved as inputs rather than quietly replaced with assumptions.
PhoneFuturist Labs / Interactive Leaks
Build it. Test it.
Open workspace ↓Explore community builds →FoldablesTabletsLaptops
Simulator
Start with the physical envelope.
Define the primary panel and its geometry.
These describe the current content/workload, not the panel's maximum specification. APL is the average picture level: 0% is black, 100% is full-white-equivalent content.
Capacity, runtime and optional physical battery analysis.
Enter your expected active-use hours. Interactive Leaks converts those hours into workload shares and applies calibrated reference power internally. You never need to enter watts.
Optional. Cell dimensions and weight are used only for packaging and energy-density calculations, not to calculate battery life.
Optional aging scenario. Enter only rates you have evidence for or explicitly want to test. PhoneFuturist does not assign a default degradation rate from battery chemistry.
Estimate wired charging from real released-phone curves.
Charging is capacity-aware and uses measured released-phone curves. Wired and wireless evidence stay separate, while battery chemistry, cell architecture and charging standard refine comparable-device selection when known.
Model one explicit radio/network workload.
Network type and signal labels do not secretly change power. Enter measured or explicit radio power when you want an energy estimate.
Model sustained SoC power, heat and throttling.
Enter up to four CPU clusters. Relative IPC is optional and normalized to your chosen reference, where 1.0 means equal IPC. Node size is metadata only and never creates an automatic performance uplift.
Optional GPU structure for raster, gaming and ray-tracing context. Shader cores and execution units are not assumed to be equivalent across vendors. Relative architecture efficiency is only used when you explicitly supply it. Nothing here is automatically converted into a 3DMark score.
Optional demand-side inputs. These calculate frame budget, pixel throughput and RT ray demand. They describe what the workload asks from the GPU, not what FPS the phone will achieve.
The lumped resistance/mass fields drive the legacy thermal trace. The optional two-node fields separately model the SoC/package and chassis/surface. Supply all four two-node physical parameters or leave them blank. Do not guess them.
Peak benchmark evidence and sustained-performance evidence are intentionally separate. Peak scores use protocol-compatible measured benchmarks. Any future-generation uplift is an explicit assumption. Sustained performance remains evidence-calibrated thermal behavior.
Add only what is actually known.
Optional. Ask the simulator which supplied camera can reach a target equivalent focal length while retaining a requested native output resolution. Computational fusion and AI reconstruction are not assumed.
Model the known internal volume budget.
Battery and camera volume are calculated automatically when enough geometry is supplied.
Unknown inputs stay unknown. The engine does not silently invent missing data.
Methodology
Measurements and specifications you provide are preserved as inputs rather than quietly replaced with assumptions.
Missing information remains missing when the engine cannot calculate it honestly from the available data.
The engine version is attached to the result so the same supported inputs can be evaluated against the same model.