AI chip · NVIDIA

NVIDIA Jetson AGX Thor

Blackwell-based embedded AI compute module aimed at humanoid robots and autonomous machines; positioned as the successor to Jetson Orin.

Powers 11 tracked robots.

Market position

AGX Thor is the humanoid successor to Orin: 2,070 TFLOPS FP8 in a 130 W module. NVIDIA's GR00T ecosystem is anchored on it. Every 2025-2026 humanoid launch (Apollo 2, Aeon, Astribot S1, Booster T2) targets Thor.

Efficiency and power

How much work you get per watt and per dollar, and how much power a full rack draws.

Perf per watt: 15.92 FP8 TFLOPS/W
2,070 TFLOPS ÷ 130 W = 15.92

What fits in 128 GB

Which open-source LLMs run on one NVIDIA Jetson AGX Thor, by precision. Weights only: add roughly 15% headroom for real serving. If a model does not fit at FP16, try INT8 or INT4 (smaller quality trade-off than most people expect).

Model	Params	FP16	INT8	INT4
Llama 3.1 8B dense	8 B	16 GB ✓	8 GB ✓	4 GB ✓
Llama 3.1 70B dense	70 B	140 GB ✗	70 GB ✓	35 GB ✓
Llama 3.1 405B dense	405 B	810 GB ✗	405 GB ✗	203 GB ✗
Llama 3.3 70B dense	70 B	140 GB ✗	70 GB ✓	35 GB ✓
DeepSeek V3 MoE (671B total, 37B active per token)	671 B	1342 GB ✗	671 GB ✗	336 GB ✗
DeepSeek R1 MoE (671B total, 37B active per token)	671 B	1342 GB ✗	671 GB ✗	336 GB ✗
Qwen 2.5 7B dense	7 B	14 GB ✓	7 GB ✓	4 GB ✓
Qwen 2.5 72B dense	72 B	144 GB ✗	72 GB ✓	36 GB ✓
Mixtral 8x7B MoE (46.7B total, 12.9B active per token)	46.7 B	93 GB ✓	47 GB ✓	23 GB ✓
Mixtral 8x22B MoE (141B total, 39B active per token)	141 B	282 GB ✗	141 GB ✗	71 GB ✓
Gemma 2 27B dense	27 B	54 GB ✓	27 GB ✓	14 GB ✓
Command R+ dense	104 B	208 GB ✗	104 GB ✓	52 GB ✓
Kimi K2 MoE (1T total, 32B active per token)	1000 B	2000 GB ✗	1000 GB ✗	500 GB ✗

Math: FP16 = params × 2 bytes; INT8 = params × 1 byte; INT4 = params × 0.5 bytes. MoE models sum every expert (full weights on disk), not the per-token active subset.

Common questions

How much does NVIDIA Jetson AGX Thor cost?

NVIDIA Jetson AGX Thor doesn't have a public launch list price. Vendors like this one usually price through direct sales rather than a public sheet. See list price per TFLOP for chips that do publish.

How much memory does NVIDIA Jetson AGX Thor have?

128 GB of LPDDR5X, running at 273 GB/s. Straight from the vendor datasheet. See chips with the most memory for context.

How much power does one NVIDIA Jetson AGX Thor draw?

130 W at the chip. A full server draws more once you add CPU, memory, networking and cooling: see the rack power number above. Compare to other chips on perf-per-watt.

What runs on NVIDIA Jetson AGX Thor?

11 robot models we track use this chip, including Nuro R3, Apollo and Figure 03, plus 8 more. Full list below.

Which open-source LLMs fit on one NVIDIA Jetson AGX Thor?

Of 13 open-source LLMs we track, 4 fit at FP16, 8 at INT8, and 9 at INT4 (for example Llama 3.1 8B, Llama 3.1 70B and Llama 3.3 70B). Full table above with each model's memory need.

See every answer we publish →

Key facts

Class: Compute SoC (AI accelerator)
Designer: NVIDIA
Safety-critical?: Yes (embedded / automotive)
Record as of: 2026-09-27

Specifications (9 fields, click to expand)

Straight from the vendor datasheet. Dense throughput shown first; sparse (2:4) numbers in parentheses where the vendor publishes them. Full datasheet linked below.

Process node: TSMC 4NP
TDP: 130 W
Memory: 128 GB LPDDR5X
Memory bandwidth: 273 GB/s
FP8 (dense): 2,070 TFLOPS
INT8 (dense): 2,070 TOPS
Form factor: Module (AGX Thor)
Announced: 2024-03-18
Released: 2025-08-01

Source: vendor datasheet

Generation

Predecessor:NVIDIA Jetson AGX Orin Full NVIDIA roadmap →

Foundry & process

Fabbed at: TSMC (all chips TSMC makes →)
Process node: TSMC 4NP

Robots running NVIDIA Jetson AGX Thor

Every robot model publicly known to ship NVIDIA Jetson AGX Thor inside it. Ranked by how widely each has been deployed.

Model	Maker	Form factor	Role	Source	Deployments
Nuro R3	Nuro	av	primary inference compute	reported	7
Apollo	Apptronik	humanoid	primary inference compute	reported	6
Figure 03	Figure AI	humanoid	primary inference compute	reported	4
Atlas	Boston Dynamics	humanoid	primary inference compute	reported	3
NEO	1X Technologies	humanoid	primary inference compute	reported	3
Apollo 2	Apptronik	humanoid	onboard perception + planning SoC	reported	2
AEON	Hexagon	humanoid	onboard humanoid compute	reported	1
Astra	Apptronik	humanoid	onboard perception + planning SoC	inferred	0
Astribot S1	Stardust Intelligence	humanoid	onboard humanoid compute	inferred	0
Booster T2	Booster Robotics	humanoid	onboard humanoid compute	inferred	0
NEO Gamma	1X Technologies	humanoid	onboard humanoid compute	reported	0

Sources

Jetson ThorNvidia

Adoption rows appear as we confirm each chip-in-robot pairing from a public source. See every chip we track for the full catalog or the NVIDIA page.