Skip to main content

Nvidia RTX 30 Series Review

Hello!

Hello and welcome to todays video. Before we get started don’t forget to hit that subscribe button, press that like button and make sure to comment down below. Enough said, lets get into todays video.

NVIDIA

NVIDIA, the biggest graphics card manufacture in the world just released the NVidia GeForce RTX 30 series, sounds amazing, so what is it and what is it capable of.

First, the RTX 30 series is the top of the line brand new graphics card, just released a couple weeks ago and already pulling in amazingly high stats.

They are powered by the Ampere, the 2nd Generation of their RTX architecture, but what is ampere? If you are the average person and just search up “What is ampere” in Google, you will end up with the result of “a unit of electric current equal to a flow of one coulomb per second”…ok…that’s great but that’s not what we are looking for.

According to Nvidia the Ampere is the connection between AI (artificial intelligence and graphics). With this technology they were able to make groundbreaking results, their Tensor Cores allow the AI in the Graphics card to run faster then ever, giving the AI tons of training of how to control the graphics. Just with this tensor change, not even a single code change they were able to gain 20 times the speed as before. By adding the change of NVIDIAS Automatic Mixed Precision they can get 2X the performance just by adding a few more lines of code.

Multi-Instance GPU

Now that we got the AI Tensions out of the way lets focus on the GPU or what they like to call, the “Multi-Instance GPU”. With the help of their MIG they are able to partition their graphics card so that only the NEEDED amount of graphics was available making it so that the rest was on “standby”, this increased the acceleration for all the applications that were tested.

After the GPU we got the 3rd Generation of NVLink. Applications, especially having multiple applications open can put a lot of strain on the computer and graphics card. This means that a TON OF DATA is being moved from GPU to GPU. With the help of 3rd Gen NVLink they were able to double the bandwidth to 600 GIGABITES PER SECOND!!!, that is about 10 times the amount of the PCIe Gen4. By adding what is known as NVSwitch you can now connect to all the GPU’s that are in your server5 allowing for even faster data transfers. Combine this with the AI and now we have a computer that is faster and has better stability.

Because of the advancement of computes and AI we are now able to stave on parameters allowing us to have what is called a “sparse matrix” meaning that there are more better direct contacts between data rather than connecting to all the data at once which would make it very dense and slower.  By making the connections sparse NVIDIA was able to get 2 times the performance that they were getting before allowing us to have a bigger field in which we can improve on faster and better in the future.

Adding more memory is always nice, isn’t it? What if we were able to add more memory but make it faster? Well, that’s exactly what NVIDIA did, they got their memory and was able to make the speeds up to 1.6 terabytes per second bandwidth. This is a HUGE increase over past generations. Getting this and adding it to more memory makes the Ampere technology even faster.

ConnectX

If you are a tech nerd than you have definitely heard of ConnectX, if you aren’t, ConnectX is Nvidias way of well, connecting everything together. They thought to themselves, how do we make these connections faster, well after a lot of hard work they came out with what is called NVIDIA Mellanox ConnectX-6 DX SmartNIC, man, that was a long title. The Mellanox aka latest generation of ConnectX allows us to decrypt frames at 200 gigabits per second and with the help of what is called GPUDirect, the way to connect to the GPU, we are able to instantly send the frames to the AI processing allowing frames to run faster allowing for some serious gaming.

Wow that was a lot about just one part of the RTX. Now let’s talk about the DLSS AI Acceleration. The DLSS is the way that NVIDIA’s AI renders all the frames, with the faster connections this allows for the rendering to be faster, allowing games to look clearer and run faster. Giving faster FPS and with the rendering the new series now you can access 8K of HDR gameplay and video with NVIDIA ShadowPlay allowing for the best of the best video out there.

At Last

At last, the new Nvidia GeForce RTX 30 series is groundbreaking and it will be worth buying as this will only increase your gameplay by a huge amount, if you are a serious gamer then this is for you, if you are a business then this is for you, if you just work at home and do simple things then the lower end RTX 30’s will be PERFECT for you, there is an RTX 30 for everyone and it can seriously improve your computer.

Thanks

Thanks for reading, please be sure to like, follow and watch my video on YouTube. This is a script to one of my past videos so it it NOT UP TO DATE. Anyways, thanks for reading and we will see you next time.

Video