Lateral thinking is known as using an idea or concept in a totally different way than usual. An example of lateral thinking is the use of technology developed in home hardware for computers, phones, and derivatives for other utilities, one of them being the single-board computer in which the Raspberry Pi stands out for its popularity, well, today we bring you the alternative NVIDIA.
What is NVIDIA Jetson?
Jetson is the trade name for a series of single board computers created by NVIDIA around its SoC Tegra family that are currently intended for use in robotics. A world in which disciplines closely related to artificial intelligence, such as computer vision, are booming. And being an SBC, the NVIDIA Jetson includes everything needed to work on the same board: processors, memory, storage, and I / O interfaces.
In other words, the NVIDIA Jetsons are very similar boards to the Raspberry Pi, but with a much higher capacity than these in terms of applications related to artificial intelligence. Indeed, current Tegra SoCs combine the technology of their PC GPUs where their Tensor cores shine in their own light with special purpose accelerators and processors focused on computer vision and combined with powerful ARM cores. These can be architectures created by NVIDIA itself or under ARM license depending on the SoC model we are talking about.
This was from their Tegra X2 or also known as Parker, among the changes NVIDIA implemented in said SoC was the addition of a PCI Express interface to connect an external graphics card. Function used in Drive on-board computers and Clara computers.
The NVIDIA Jetsons are therefore one more product line among many of those sold by the veteran company that are linked to its Tegra SoCs, these being the most commercial of all for the general public, but because they focus on A market Very specific enthusiasts have not reached wide distribution and in some cases can be extremely expensive, costing even more than the most expensive smartphone on the market.
Can I use them as a PC?
Since these have CPU, GPU, memory, and all the necessary interfaces, you can do that. However, it will not be worth it if you choose one of the newer models, especially the Tegra Xavier, because these NVIDIA Jetsons are not only extremely expensive, but due to the fact that you will not be able to enjoy Units like the PVA or NVDLA being focused on deep learning and computer vision, other than that, we insist, they are extremely expensive.
In principle, the most similar hardware that is available in the market is that of the NVIDIA Shield, which is also based on a Tegra model, more precisely the X1 and is the same as that of the Nintendo Switch with the difference is that its clock speeds are not limited and it is an ideal platform as an alternative to the Raspberry Pi as a retro machine and allows things that the popular SBC cannot: emulate consoles like the Nintendo GameCube with no problem. .
Although in case you want to use it as an ARM computer and install GNU / Linux, the best option is the Jetson Nano which is based on a shortened version of the Tegra X1. Why not use the NVIDIA Shield? The answer to this question is simple, for some strange reason the Ubuntu version for Jetson, L4T or Linux for Tegra SBCs does not work with the standard X1.
NVIDIA Jetson Nano vs. Raspberry Pi 4
Okay, you don’t know whether to choose between Jetson Nano and Raspberry Pi 4 to build a low cost, consumer ARM based PC. So you will be interested to know what advantages you can get by opting for the NVIDIA card. So let’s see what their differences are:
- The Raspberry Pi’s SoC is a Broadcom BCM2711 with quad 1.5GHz A72 cores, while the Jetson Nano’s CPU is a 1.42GHz A57. Well, the A72 is a more optimized version than the A57. .in terms of power, but it doesn’t perform better for clock speed.
- Both use LPDDR4 memory, but the Raspberry Pi 4 has a 4GB version, not the Jetson Nano.
- In wireless connectivity, the Raspberry Pi 4 has Bluetooth and WiFi as standard. The Jetson Nano, on the other hand, only includes Ethernet and you have to purchase a separate M.2 card with WiFi.
- As for their I / O ports, both have ports focused on robotics and the world of hobby electronics.
- Let us not forget either that being specific hardware they use specific GNU / Linux distributions.
The big difference between the two, as you might have guessed, is the integrated GPU and that is that there is a huge difference between the two GPUs. Of course, for that you have to take into account which apps take advantage of the capacity, because paying the $ 40 difference for a feature you’re not going to use is a waste of money.
Of course, Jetson Nano for its price is a much better option than an NVIDIA Shield Pro which can cost you double and in which you will have to fight for the simple act of installing the operating system since the operating system which comes with the Jetson he is stubborn and does not want to settle on the Shield.