On the GPU Expertise Convention NVIDIA introduced the Jetson Nano Module and the Jetson Nano Developer Package. In comparison with different Jetson boards which value between $299 and $1099, the Jetson Nano bears a low value of $99. This places it inside the attain of many builders, educators, and researchers who couldn’t spend a whole bunch of dollars to get such a product.
The Jetson Nano Growth Package (left) and the Jetson Nano Module (proper)
Bringing again AI growth from ‘cloud’
In the previous couple of years, now we have seen loads of advances in AI analysis. Historically AI computing was all the time achieved within the cloud, the place there was loads of processing energy obtainable.
Lately, there’s been a development in shifting this computation away from the cloud and do it domestically. That is known as Edge Computing. Now on the embedded degree, merchandise which might do such complicated calculations required for AI and Machine Studying have been sparse, however we’re seeing an important explosion lately on this product section.
Merchandise just like the SparkFun Edge and OpenMV Board are good examples. The Jetson Nano, is NVIDIA’s newest providing on this market. When related to your system, will probably be in a position to provide the processing energy wanted for Machine Studying and AI duties with out having to depend on the cloud.
That is nice for privateness in addition to saving on web bandwidth. Additionally it is safer since your information all the time stays on the system itself.
Jetson Nano focuses on smaller AI tasks
Beforehand launched Jetson Boards just like the TX2 and AGX Xavier have been utilized in merchandise like drones and automobiles, the Jetson Nano is focusing on smaller tasks, tasks the place it’s essential have the processing energy which boards just like the Raspberry Pi can not present.
NVIDIA’s JetPack SDK offers a ‘full desktop Linux setting based mostly on Ubuntu 18.04 LTS’. In different phrases, the Jetson Nano is powered by Ubuntu Linux.
NVIDIA Jetson Nano Specs
For $99, you get 472 GFLOPS of processing energy as a result of 128 NVIDIA Maxwell Structure CUDA Cores, a quad-core ARM A57 processor, 4GB of LP-DDR4 RAM, 16GB of on-board storage, and 4k video encode/decode capabilities. The port choice can be fairly first rate with the Nano having Gigabit Ethernet, MIPI Digicam, Show outputs, and a few USB ports (1×three.zero, three×2.zero). Full vary of specs might be discovered right here.
CPUQuad-core ARM® Cortex®-A57 MPCore processorGPUNVIDIA Maxwell™ structure with 128 NVIDIA CUDA® coresRAM4 GB 64-bit LPDDR4Storage16 GB eMMC 5.1 FlashCamera 12 lanes (three×four or four×2) MIPI CSI-2 DPHY 1.1 (1.5 Gbps)ConnectivityGigabit EthernetDisplay PortsHDMI 2.zero and DP 1.2USB Ports1 USB three.zero and three USB 2.0Other1 x1/2/four PCIE, 1x SDIO / 2x SPI / 6x I2C / 2x I2S / GPIOsSize69.6 mm x 45 mm
Together with good , you get help for almost all of fashionable AI frameworks like TensorFlow, PyTorch, Keras, and so forth. It additionally has help for NVIDIA’s JetPack and DeepStream SDKs, similar because the dearer TX2 and AGX Boards.
“Jetson Nano makes AI extra accessible to everybody — and is supported by the identical underlying structure and software program that powers our nation’s supercomputer. Bringing AI to the maker motion opens up an entire new world of innovation, inspiring individuals to create the following massive factor.” mentioned Deepu Talla, VP and GM of Autonomous Machines at NVIDIA.
What do you consider Jetson Nano?
The supply of Jetson Nano differs from nation to nation.
The Intel Neural Stick, can be one such accelerator which is competitively costs at $79. It’s good to see competitors stirring up at these lower cost factors from the massive producers.
I’m trying ahead to getting my palms on the product if doable.
What do you guys take into consideration a product like this? Tell us within the feedback beneath.