AAEON MAXER-5000 AI inference server powered by NVIDIA Jetson Thor T5000 — 2,070 FP4 TFLOPS, 128GB LPDDR5X, QSFP28 4×25GbE, and modular 5G/Wi-Fi 7/NVMe expansion for edge AI server deployments.
NVIDIA Jetson Thor T5000
2,070 FP4 / 1,035 FP8 TFLOPS
14-core Arm Neoverse-V3AE 64-bit
NVIDIA Blackwell, 2560 CUDA cores + 5th-gen Tensor Cores
128GB 256-bit LPDDR5X
QSFP28 ×1 (4× 25GbE) + RJ-45 ×1 (5GbE)
M.2 2280 M-Key ×1 (NVMe)
M.2 3052 B-Key (4G/LTE) + M.2 2230 E-Key (Wi-Fi 7) + NANO SIM
USB 3.2 Gen 2 Type-A ×2 + USB 2.0 ×2 + Micro USB ×1
HDMI ×1 (HDMI 2.0)
NVIDIA JetPack 7.0 and above
Available
Quote
The AAEON MAXER-5000 is an AI inference server built on NVIDIA's Jetson Thor T5000 module, bringing 2,070 FP4 TFLOPS of Blackwell-generation AI compute to the edge. With a 14-core Arm Neoverse-V3AE CPU, 2,560 CUDA cores, fifth-generation Tensor Cores, and 128GB of LPDDR5X memory, it runs large vision-language models, LLMs, and multi-modal inference pipelines entirely on-device — ideal for smart-city video analytics, edge AI server clusters, and industrial automation.
High-bandwidth networking is a core strength: a QSFP28 port provides four 25GbE links while a dedicated 5GbE RJ-45 handles management and uplink. Modular expansion includes an M.2 B-Key slot for 4G/LTE, an M.2 E-Key slot for Wi-Fi 7, a NANO SIM slot, and an M.2 2280 NVMe slot for high-speed local storage. Shipping with NVIDIA JetPack 7.0 pre-installed, the MAXER-5000 gives developers immediate access to NVIDIA Isaac, Metropolis, and Holoscan for robotics and AI agent workloads at the edge.
Need AAEON MAXER-5000?
Jetson Thor T5000 AI inference server for edge deployment. Contact QS Compute for pricing, availability, and configuration.
Request Quote