We benchmark multi-channel photoplethysmography (PPG), dual-frequency L1/L5 GNSS multipath mitigation, and autonomic recovery algorithms for competitive endurance athletes.
A laboratory-style comparison of eight 2026 endurance wearables: Garmin, Apple, Suunto, Coros, Polar, Whoop/Oura, Google Pixel and Samsung. We compare PPG sampling rates, MARA, running power models, dual-band GNSS endurance and PSG-validated sleep staging.
A laboratory-style deep dive into the Apple Watch Ultra 2: continuous 250 Hz optical PPG, Neural Engine motion-artifact deconvolution, metabolic-equivalent running power vs. lab metabolic carts, and watchOS sleep staging concordance with polysomnography.
A laboratory-style deep dive into the Garmin Fenix 8: Elevate Gen 5 optical cluster (6 green emitters, 4 photodiodes), Firstbeat EWMA acute:chronic workload modeling, wrist running dynamics error margins and on-device routable vector cartography.
A biomechanics-first review of the Google Pixel Watch 3: convex multi-path optical cluster, Google Health AI motion-artifact rejection, wrist running form (GCT ±9.5 ms, vertical ratio, stride length), the Cardio Load / Target Load engine and honest battery limits.
A biomechanics-focused review of the Samsung Galaxy Watch Ultra: its BioActive optical array, verified accuracy metrics, dual-frequency L1/L5 GNSS on mountain trails, Wear OS 5 running power and 60-hour battery limits.
A laboratory-style benchmark of the Suunto Race and Suunto Vertical: L1+L5 dual-frequency GNSS battery life (40–90 h), free offline vector maps, and the native TrainingPeaks TSS/CTL/ATL/TSB load architecture.
A laboratory-style comparison of Whoop 4.0 (wrist/bicep capillary PPG) and Oura Ring Gen 4 (palmar digital arterial PPG): nocturnal rMSSD derivation, polysomnography sleep staging kappa, logarithmic Strain vs. Readiness modeling and two-year cost.
