// Platform Catalog

Edge AI Platforms — Specs, Sizing & Deployment Fit

Last updated: August 2026

Every platform EdgeAIStack sizes and benchmarks — Jetson modules, Hailo accelerators, Coral, and RK3588 — with hedged mid-2026 pricing, planning capacity estimates, and honest decision guidance. Compute and power are manufacturer specifications; capacity figures are EdgeAIStack planning estimates, not guarantees.

All 11 Platforms

Platform	Best fit	Compute	Power	Est. planning capacity	Indicative price
Jetson Orin Nano Super	Flexible entry Jetson	67 TOPS	10–25W	~6 streams	~$399
Jetson Orin Nano 8GB	Entry CUDA development	40 TOPS	7–15W	~4 streams	~$399
Jetson Orin NX 16GB	Multi-camera production	157 TOPS	10–40W	~12 streams	~$999
Jetson AGX Orin 64GB	High-density pipelines	275 TOPS	15–60W	~20 streams	~$2,999
Jetson AGX Thor T5000	Generative AI & very high density	2,070 TOPS	40–130W	~32 streams	~$4,999
Jetson T4000	Thor generation, lower envelope	~1200 FP4 TFLOPs	40–70W	—	~$2,999
Hailo-8	Efficient dedicated inference	26 TOPS	2.5–8W	~4 streams	~$200
Hailo-8L	Lowest-power detection	13 TOPS	1–1.5W	~2 streams	~$70
Hailo-10H	Higher-capacity M.2 accelerator	20 TOPS	1–5W	~8 streams	~$130
Google Coral Edge TPU	Legacy ultra-low-power classification	4 TOPS	2–2.5W	~1 stream	~$60
Rockchip RK3588	Budget Linux SBC	6 TOPS	5–20W	~3 streams	~$149

Prices are indicative list/street pricing as of mid-2026, not quotes. Capacity assumes YOLO-class 1080p detection at ~15 FPS — validate with the Hardware Selector.

// All Platforms

NVIDIA

Jetson Orin Nano Super

67 INT8 TOPS · 10–25 W · ~6 streams (est.) · ~$399 dev kit

→

NVIDIA

Jetson Orin Nano 8GB

40 INT8 TOPS · 7–15 W · ~4 streams (est.) · ~$399 module

→

NVIDIA

Jetson Orin NX 16GB

157 INT8 TOPS · 10–40 W · ~12 streams (est.) · ~$999 module