New 40 Series Graphics Cards

Jesse Richardson, VyStar Student Intern

With Quarter 4 of the year having started a few weeks ago, many companies like AMD and NVIDIA have been coming out with new advancements in the world of technology. Specifically with the new RTX 40 series from NVIDIA. During the September 20th event, a GPU Technology Conference was help by Nvidia about the new cards and what we can expect from them.

A good bit of new information was released regarding the RTX 4090 and the RTX 4080 graphics cards, and their chips subsequently. The 4090 will be released on October 12th, being $100 dollars more than its previous generation’s counterpart, the 3090, at $1599. As for the 4080, we know it will come in two variations. A 12 and 16 Gigabyte card, with the 12gb model’s price at $899, and the 16gb model’s price being unreleased

The price may be high, but with the excessive price point comes bountiful tech. At $1599 you will be expecting a new chip series, higher CUDA Core count, and even higher frames. The new Ampere Next chip architecture will be used for the whole 40 series, being the newest chip released from NVIDIA. The Ampere Next chip is built with TSMC’s 4nm Process, which is a step up from the 5nm process used on the 30 series almost 2 years ago. TSMC (Taiwan Semiconductor Manufacturing Company) has been used for nearly 20 years by NVIDIA, with one exception being made for the 3090 on Samsung’s press, it’s no surprise that the chipset’s manufacturer will stay the same. However, with the Ampere Next chip, it will have nearly 2 times the amount of CUDA cores, and twice the amount of VRAM. Though there may be no released base clock speed or boost clock speed, going off of the released 4090 specs, we can expect twice the amount of the 3090 in the 4090 Ti.

The 4080 Card and its variants will have easily twice the amount of Clock speed, and about a 500 MHz increase in the Boost Clock Speed. However, it will expectedly have a decrease in VRAM from the 3090, with a similar decrease in memory speed. Though they are the same card, the difference in the size of the VRAM creates a clear difference between the two. With the CUDA Core count of the 12gb model being almost 2000 less than the 16gb model, and each of the respective clock speeds varying. These are not the only differences, however, they are the differences that matter most to the buyer upon researching.

Though the new 40 series is coming out soon, with the 4090 coming as soon as October, and the 4080 variants coming in November, this does not mean they will take too much attention from the 30 series. NVIDIA plans to keep pushing out the 30 series until they have run the market dry and see a clear decrease in buyers, before making the 40 series their new toy on display. And with the blessing of the Crypto market crash, many buyers will have smiles on their faces come Black Friday, with a near $400 cut in prices from the 30 series.

To the question of “Is it worth it?”, the answer I would give, is that it depends. There is a small, yet profitable, market for these high-end graphics cards. If you’re really looking for a high-frame rate, 1440p experience, I would say that the 4080 is more than enough. However, if you are working on high-end photo/video editing, and you really want to have a smooth experience, then you might just want to spring for that 4090 and reap the benefits of virtually no lag, and beautiful video.