엔비디아 젯슨 모듈 사양 비교(Compare NVIDIA Jetson Module Specifications)

 

	NANO	TX2 NX	TX2 4GB	TX2	TX2i	Xavier NX	AGX Xavier 8GB	AGX Xavier
AI Performance	472 GFLOPS	1.33 TFLOPS			1.26 TFLOPS	21 TOPS	5.5 TFLOPS (FP16) 11.1 TOPS (INT8)	11 TFLOPS (FP16) 22 TOPS (INT8)
GPU	128-Core NVIDIA Maxwell™ GPU	256-Core NVIDIA Pascal™ GPU				384-Core Volta™ GPU with 48 Tensor cores	384-Core Volta™ GPU with 48 Tensor cores	512-Core Volta™ GPU with 64 Tensor Cores
CPU	4-core ARM Cortex®-A57 MPCore Processor	Dual-core Denver 2 64-bit CPU and 4-core ARM Cortex®-A57 MPCore Processor Complex				6-core NVIDIA Carmel ARM®v8.2 64-bit CPU 6MB L2 + 4MB L3	6-core ARM v8.2 64-bit CPU, 8MB L2 + 4MB L3	8-Core ARM v8.2 64-Bit CPU, 8MB L2 + 4MB L3
Memory	4GB 64-bit LPDDR4 25.6GB/s	4 GB 128-bit LPDDR4 51.2GB/s		8 GB 128-bit LPDDR4 59.7GB/s	8 GB 128-bit LPDDR4 (ECC Support) 51.2GB/s	8GB 128-bit LPDDR4x 51.2GB/s	8GB 256-bit LPDDR4x 1333MHz - 85GB/s	16GB 256-bit LPDDR4x 2133MHz - 137GB/s
Storage	16 GB eMMC 5.1	16 GB eMMC 5.1		32 GB eMMC 5.1	32 GB eMMC 5.1	16 GB eMMC 5.1	32GB eMMC 5.1
Power	5W, 10W	7.5W, 15W			10W, 20W	10W, 15W	10W, 15W, 30W
PCIe	1 x4 (PCIe Gen2)	1 x1 + 1 x2 (PCIe Gen2)	1 x1 + 1 x4 OR 1 x1 + 1 x1 + 1 x2 (PCIe Gen2)			1 x1 (PCIe Gen3) + 1 x4 (PCIe Gen4)* 총 144GT/s*	1 x8 + 1 x4 + 1 x2 + 2 x1 (PCIe Gen4, Root Port & Endpoint)
CSI Camera	Up to 4 cameras 12 lanes MIPI CSI-2 D-PHY 1.1 (up to 18 Gbps)	Up to 5 cameras (12 via virtual channels) 12 lanes MIPI CSI-2 D-PHY 1.2 (up to 30 Gbps)	Up to 6 cameras (12 via virtual channels) 12 lanes MIPI CSI-2 D-PHY 1.2 (up to 30 Gbps)			Up to 6 cameras (24 via virtual channels) 14 lanes MIPI CSI-2 D-PHY 1.2 (up to 30 Gbps)	카메라 최대 6개 (가상 채널을 통해 36개)1 6 lanes MIPI CSI-2, 8 lanes SLVS-EC D-PHY (40 Gbps) C-PHY (64 Gbps)	Up to 6 cameras (36 via virtual channels) 16 lanes MIPI CSI-2 | 8 lanes SLVS-EC D-PHY 1.2 (up to 40 Gbps) C-PHY 1.1 (up to 91 Gbps))
Video Encode	1x 4K30(H.265) 2x 1080p60(H.265)	1x 4K60(H.265) 3x 4K30(H.265) 4x 1080p60(H.265)				2x 4Kp30 | 6x 1080p60 | 14x 1080p30(H.265 & H.264)	2x 4K @ 30 (HEVC)	8x 4K @ 30 (HEVC)
Video decode	1x 4K60(H.265) 4x 1080p60(H.265)	2x 4K60(H.265) 7x 1080p60(H.265) 14x 1080p30(H.265)				2x 4Kp60 | 4x 4Kp30 | 12x 1080p60 | 32x 1080p30(H.265) 2x 4Kp30 | 6x 1080p60 | 16x 1080p30(H.264)	4x 4K @ 30 (HEVC)	12x 4K @ 30 (HEVC)
Networking	10/100/1000 BASE-T	10/100/1000 BASE-T		10/100/1000 BASE-T, Wi-Fi	10/100/1000 BASE-T	10/100/1000 BASE-T	10/100/1000 BASE-T
Display	2 multi-mode  DP 1.2/eDP 1.4/HDMI 2.0 1 x2 DSI(1.5Gbps/레인)	2 multi-mode  DP 1.2/eDP 1.4/HDMI 2.0 1 x2 DSI(1.5Gbps/레인)	HDMI 2.0 / eDP 1.4 / 2x DSI / 2x DP 1.2			2 multi-mode  DP 1.4/eDP 1.4/HDMI 2.0	Three multi-mode DP 1.2/eDP 1.4/HDMI 2.0
DL Accelerator	-	-				(2x) NVDLA Engines	(2x) NVDLA Engines 4.1 TFLOPS (FP16) 8.2 TOPS (INT8)	(2x) NVDLA Engines 5 TFLOPS (FP16) 10 TOPS (INT8)
Vision Accelerator	-	-				7-Way VLIW Vision Processor	7-Way VLIW Vision Processor
UPHY			Gen 2 | 1x4 + 1x1 또는 2x1 + 1x2, USB 3.0 + USB 2.0				3x USB 3.1, 4x USB 2.0 1 x8 or 1 x4 or 1 x2 or 2 x1 PCIe (Gen3)	3x USB 3.1, 4x USB 2.0 1 x8 or 1 x4 or 1 x2 or 2 x1 PCIe (Gen4)
기타							UART, SPI, CAN, I2C, I2S, DMIC, GPIOs
크기	69.6 mm x 45 mm	69.6mm x 45mm	87 mm x 50 mm			69.6mm x 45mm	100 mm x 87 mm
Mechanical	260 핀 SO-DIMM 커넥터	260핀 SO-DIMM 커넥터	TTP(Thermal Transfer Plate)가 있는 400핀 커넥터			260핀 SO-DIMM 커넥터	699핀 Molex Mirror Mex 커넥터 통합 열 전달 플레이트

Ref: www.nvidia.com/en-us/autonomous-machines/embedded-systems/