Edge AI Computer RV2-X2FM-NVIDIA RTX4070/RTX4060/RTX4050/MX550
RugCon-RV2-X2FM , edge AI inference platform, Equipped with NVIDIA RTX4070/RTX4060/RTX4050/MX550 and support Intel® 14th/13th/12th Core™ i7/i5/i3 low power(15W/28W) high performance processor, 5x M12 LAN (2.5GbE)/4x PoE+, 2x CAN/6x COM/6x USB/8x IDIO, 5G/4G/LTE/WIFI/BT, IO screw-lock, Ignition/Remote Control,9~48V DC INPUT
- Option : RTX4070-8G/65W, RTX4060-8G/60W, RTX4050-6G/65W,MX550-4G/35W /RTX4060-8G/60W
- Intel® 14th/13th/12th Core™ i9/i7/i5 (P/H Series) 28W/45W)
- 5x 2.5GbE, 4x POE+(option) , M12 X-code Connector
- 2x Isolated CAN B.0
- 4x RS232/422/485, 2x RS232
- 6x USB ( Support 3x USB3)
- 1x Mic In/Line Out
- 4 Channel isolated DI, 4 Channel isolated DO
- 2x Displays (1x DVI-D, 1x DP)
- 2x M.2 2280 M key (SATA III SSD) , support Raid 0/1
- 1x M.2 2242/3052 B key with SIM slot, support 4G/5G
- 1x Full Size Mini-PCIe with SIM slot, support 4G/5G
- 3x M.2 Key E Type 2230 for WIFI/BT/GPS
- Optional support 1x M.2 2280 Key M ( PCIe x4 ), Support AI acceleration card/image capture card/NVME SSD, etc.
- 6~48V Wide DC In Voltage Input
- Support 1x DC Out 12V/1A Power output
- Ignition/Remote Power Control

ENBIK RV2-X2FM is an edge AI inference platform specially designed for edge application, providing developers with excellent edge AI computing power. After installing the ENBIK AI computing power box, it is only 21×17.6×11 cm in size. With such a small size, it can output a single-precision computing power of up to 15.71 TFLOPS, and the computing power of converting to integer operations is also up to 321TOPS. In addition, it has 4,608 CUDA numbers and 144 tensor cores. In such an excellent computing power output, only 65W of power consumption can achieve excellent performance, seamless artificial intelligence multitasking and amazing visual experience, all of which are carried out in an amazingly compact size.
ENBIK RV2-X2FM can be equipped with Intel’s 12/13/14th generation Core™ i7/i5/i3 low-power high-performance 15W/28W processor, supports 5xM12 LAN (optional 4xPoE power supply)/6 USB/6 serial ports/2 isolation CAN BUS/8 isolation DIO, and supports 2 M.2 2280 SATA III (Raid 0/1 protection), supports 5G/4G/LTE/WIFI/BU, etc. in the communication part. In addition, it supports vehicle ignition switch and remote switch control, and 9~48V wide voltage input protection in the input power supply part.
The ENBIK RV2-X2FM artificial intelligence edge computing platform can be easily installed and integrated into a narrow space, and effectively solves the edge artificial intelligence deployment problem of limited power supply. It can run the API transfer learning function through the pre-training model, and then directly implement the fine-tuning model and low-latency reasoning at the edge.
What is exciting is that the MF2A computing power box of ENBIK RV2-X2FM adopts AI chips such as RTX4070/RTX4060/RTX4050/MX550 of NVIDIA architecture, which undoubtedly makes software developers more excited. The rich NVIDIA software ecosystem makes edge artificial intelligence deployment faster and easier to implement, effectively solves the installation and low-latency reasoning problems of various artificial intelligence models at the edge, and quickly makes various decisions and judgments at the edge. Therefore, after integrating the MF2A artificial intelligence computing box, the ENBIK RV2-X2FM is definitely the best choice for demanding artificial intelligence tasks, limited space and limited power edge deployment in various edge applications.

| GPU Performance | ||
| RTX4070-65W | 8GB GDDR6 128bit/15.71 TFLOPS/321 TOPS/4608 CUDA/144 Tensor/36 RT Core | |
| RTX4060-60W | 8GB GDDR6 128bit/12.22 TFLOPS/232 TOPS/3072 CUDA/96 Tensor/24 RT Core | |
| RTX4050-65W | 6GB GDDR6 96bit/11.835 TFLOPS/194 TOPS/2560 CUDA/80 Tensor/20 RT Core | |
| MX550-35W | 4GB GDDR6 64 bit/4.5 TFLOPS/1024 CUDA/32 Tensor/16 RT Core | |
| IO Interface | ||
| USB |
6 |
3x USB 3.2 Gen2x1(10Gbps)/USB 2.0,
3x USB 2.0 |
| Display |
2 |
1x DVI-D (1920 x 1200) , 1x DP(4096x 2304) |
| Audio |
1 |
1x Mic-In , 1x Line Out |
| LAN |
5 |
1x 2.5 GbE with TSN support,
4x 2.5 GbE with PoE+ support 802.3 af/at 25W /48V (PoE+ is option) |
| COM |
6 |
1x Isolated RS232/422/485, 3x RS232/422/285, 2x RS232 |
| CAN |
2 |
2x CAN (Isolation) |
| DIO |
8 |
4 Channel isolated DI, 4 Channel isolated DO |
| Storage | ||
| M.2 |
2 |
2x M.2 2280 SATA, support Raid 0/1 |
| 2.5” |
0 |
N/A |
| On Board Expansion | ||
| Mini-PCIe
M.2 Expansion |
5 |
1x Full Size Mini-PCIe with SIM supporting 4G/5G/LTE
1x M.2 2242/3052 B key with SIM supporting 5G/LTE 3x M.2 Key E Type 2230 for BT/GPS |
| M.2 2280 M key |
1 |
Optional support 1x M.2 2280 Key M ( PCIe x4 ), Support AI acceleration card/image capture card/NVME SSD, etc. |
| Mechanical | ||
| Mechanical | Dimension : 211(W) x 176(D) x 113(H) mm
Weight : 2.65 kg Mounting : Wall Mount |
|
| Mechanical | ||
| AI BOX | Support installing MF2A, this is ENBIK AI BOX which equipped with MXM GPU AI Card | |
| Environmental | ||
| Operating Temp. | -25°C ~ 60°C
-40°C ~ 85°C ( Extended) |
|
| Storage Temp. | -40°C ~ 85°C | |
| Humidity | 0%~90% , non-condensing | |
| Vibration | EN 50155:2017/ IEC 61373, Category I, Class B – Body mounted
MIL-STD-810H, Method 514.8, Category 4 |
|
| Shock | EN 50155:2017/ IEC 61373, Category I, Class B – Body mounted
MIL-STD-810H, Method 516.8, Procedure I |
|
| EMC | E-Mark, EN 50121 (EN 50155 EMC)
CE/FCC Class A, according to EN 55032 & EN 55035 |
|
| Power |
| 1x Power Button
1x 3-pin pluggable terminal block for 6~48V DC Input , 1x 4-pin pluggable terminal block forIgnition control and Remote Control 1x 2-pin pluggable terminal block for12V/1A DC output |
| Warranty |
| 3 Year Warranty |
|
Intel® 13th Core™ (45W) |
Intel® 12th Core™ (45W) |
|||||||||
|
CPU |
I9-13900H |
I7-13700H |
I5-13600H |
I7-13800HE |
I5-13600HE |
I9-12900H |
I7-12700H |
I5-12600H |
I7-12800HE |
I5-12600HE |
|
DDR |
DDR5 : up to 4800MT/s ( Max. 64 GB) |
|||||||||
|
Cores / Threads |
14 C / 20 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
12 C / 16 T |
|
P Turbo |
up to 5.4GHz |
up to 5.0GHz |
up to 4.8GHz |
up to 5.0GHz |
up to 4.8GHz |
up to 5.0GHz |
up to 4.7GHz |
up to 4.5GHz |
up to 4.6GHz |
up to 4.5GHz |



ENBIK RV2-X2FM is an edge AI inference platform specially designed for edge application, providing developers with excellent edge AI computing power. After installing the ENBIK AI computing power box, it is only 21×17.6×11 cm in size. With such a small size, it can output a single-precision computing power of up to 15.71 TFLOPS, and the computing power of converting to integer operations is also up to 321TOPS. In addition, it has 4,608 CUDA numbers and 144 tensor cores. In such an excellent computing power output, only 65W of power consumption can achieve excellent performance, seamless artificial intelligence multitasking and amazing visual experience, all of which are carried out in an amazingly compact size.
ENBIK RV2-X2FM can be equipped with Intel’s 12/13/14th generation Core™ i7/i5/i3 low-power high-performance 15W/28W processor, supports 5xM12 LAN (optional 4xPoE power supply)/6 USB/6 serial ports/2 isolation CAN BUS/8 isolation DIO, and supports 2 M.2 2280 SATA III (Raid 0/1 protection), supports 5G/4G/LTE/WIFI/BU, etc. in the communication part. In addition, it supports vehicle ignition switch and remote switch control, and 9~48V wide voltage input protection in the input power supply part.
The ENBIK RV2-X2FM artificial intelligence edge computing platform can be easily installed and integrated into a narrow space, and effectively solves the edge artificial intelligence deployment problem of limited power supply. It can run the API transfer learning function through the pre-training model, and then directly implement the fine-tuning model and low-latency reasoning at the edge.
What is exciting is that the MF2A computing power box of ENBIK RV2-X2FM adopts AI chips such as RTX4070/RTX4060/RTX4050/MX550 of NVIDIA architecture, which undoubtedly makes software developers more excited. The rich NVIDIA software ecosystem makes edge artificial intelligence deployment faster and easier to implement, effectively solves the installation and low-latency reasoning problems of various artificial intelligence models at the edge, and quickly makes various decisions and judgments at the edge. Therefore, after integrating the MF2A artificial intelligence computing box, the ENBIK RV2-X2FM is definitely the best choice for demanding artificial intelligence tasks, limited space and limited power edge deployment in various edge applications.

| GPU Performance | ||
| RTX4070-65W | 8GB GDDR6 128bit/15.71 TFLOPS/321 TOPS/4608 CUDA/144 Tensor/36 RT Core | |
| RTX4060-60W | 8GB GDDR6 128bit/12.22 TFLOPS/232 TOPS/3072 CUDA/96 Tensor/24 RT Core | |
| RTX4050-65W | 6GB GDDR6 96bit/11.835 TFLOPS/194 TOPS/2560 CUDA/80 Tensor/20 RT Core | |
| MX550-35W | 4GB GDDR6 64 bit/4.5 TFLOPS/1024 CUDA/32 Tensor/16 RT Core | |
| IO Interface | ||
| USB |
6 |
3x USB 3.2 Gen2x1(10Gbps)/USB 2.0,
3x USB 2.0 |
| Display |
2 |
1x DVI-D (1920 x 1200) , 1x DP(4096x 2304) |
| Audio |
1 |
1x Mic-In , 1x Line Out |
| LAN |
5 |
1x 2.5 GbE with TSN support,
4x 2.5 GbE with PoE+ support 802.3 af/at 25W /48V (PoE+ is option) |
| COM |
6 |
1x Isolated RS232/422/485, 3x RS232/422/285, 2x RS232 |
| CAN |
2 |
2x CAN (Isolation) |
| DIO |
8 |
4 Channel isolated DI, 4 Channel isolated DO |
| Storage | ||
| M.2 |
2 |
2x M.2 2280 SATA, support Raid 0/1 |
| 2.5” |
0 |
N/A |
| On Board Expansion | ||
| Mini-PCIe
M.2 Expansion |
5 |
1x Full Size Mini-PCIe with SIM supporting 4G/5G/LTE
1x M.2 2242/3052 B key with SIM supporting 5G/LTE 3x M.2 Key E Type 2230 for BT/GPS |
| M.2 2280 M key |
1 |
Optional support 1x M.2 2280 Key M ( PCIe x4 ), Support AI acceleration card/image capture card/NVME SSD, etc. |
| Mechanical | ||
| Mechanical | Dimension : 211(W) x 176(D) x 113(H) mm
Weight : 2.65 kg Mounting : Wall Mount |
|
| Mechanical | ||
| AI BOX | Support installing MF2A, this is ENBIK AI BOX which equipped with MXM GPU AI Card | |
| Environmental | ||
| Operating Temp. | -25°C ~ 60°C
-40°C ~ 85°C ( Extended) |
|
| Storage Temp. | -40°C ~ 85°C | |
| Humidity | 0%~90% , non-condensing | |
| Vibration | EN 50155:2017/ IEC 61373, Category I, Class B – Body mounted
MIL-STD-810H, Method 514.8, Category 4 |
|
| Shock | EN 50155:2017/ IEC 61373, Category I, Class B – Body mounted
MIL-STD-810H, Method 516.8, Procedure I |
|
| EMC | E-Mark, EN 50121 (EN 50155 EMC)
CE/FCC Class A, according to EN 55032 & EN 55035 |
|
| Power |
| 1x Power Button
1x 3-pin pluggable terminal block for 6~48V DC Input , 1x 4-pin pluggable terminal block forIgnition control and Remote Control 1x 2-pin pluggable terminal block for12V/1A DC output |
| Warranty |
| 3 Year Warranty |
|
Intel® 13th Core™ (45W) |
Intel® 12th Core™ (45W) |
|||||||||
|
CPU |
I9-13900H |
I7-13700H |
I5-13600H |
I7-13800HE |
I5-13600HE |
I9-12900H |
I7-12700H |
I5-12600H |
I7-12800HE |
I5-12600HE |
|
DDR |
DDR5 : up to 4800MT/s ( Max. 64 GB) |
|||||||||
|
Cores / Threads |
14 C / 20 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
14 C / 20 T |
12 C / 16 T |
14 C / 20 T |
12 C / 16 T |
|
P Turbo |
up to 5.4GHz |
up to 5.0GHz |
up to 4.8GHz |
up to 5.0GHz |
up to 4.8GHz |
up to 5.0GHz |
up to 4.7GHz |
up to 4.5GHz |
up to 4.6GHz |
up to 4.5GHz |

