Newst GeForce RTX 3080 (Deep Review)

Expectations for the Nvidia GeForce RTX 3080 couldn’t have been higher when it initially launched because it had to dramatically outperform Nvidia’s top-tier graphics card from the Turing era.

But like the rest of the Nvidia Ampere, the RTX 3080 hasn’t just risen to the occasion; offering quick 4K gaming at a reasonable, if not entirely inexpensive, MSRP completely redefines the performance of the top-tier graphics cards. The RTX 3080 appears to represent the biggest generational gain in power we have seen in a very long time compared to the cards it replaces.

It performs 20 to 30 percent better than the card for the Nvidia GeForce RTX 2080 Ti and, more amazingly, 50 to 80 percent better than the card for the Nvidia GeForce RTX 2080. The performance boost and significantly more affordable retail price make this GPU even more appealing. Its price is almost half as much as the GPU from the previous generation.

Newst GeForce RTX 3080 Review

The Nvidia GeForce RTX 3080 lowers the barrier to entry for high-end gaming by enabling the finest PC games to operate at higher resolutions and refresh rates while costing less.

There are less expensive cards with better value propositions, but when it comes to the upper group of graphics cards, the Nvidia GeForce RTX 3080 is the best value for those who have the cash to spend but want the greatest return on their investment. For those who can afford the price of entry, we highly recommend it as the best graphics card to buy.

Features

Based on the Ampere graphics architecture, the Nvidia GeForce RTX 3080 offers significant upgrades over Turing in raw performance and power efficiency. The entire performance profile is considerably above what any Nvidia Turing graphics card was capable of because Nvidia boosted the power budget over the RTX 2080 by such a large margin while improving power efficiency.

The ray tracing (RT) and Tensor cores in the second and third generations have improved, but the rasterization engine may have seen the most advances. By making both of the data paths on each Streaming Multiprocessor (SM) capable of handling Floating Point 32 (FP32) workloads, Nvidia was able to double the number of CUDA cores present on each SM, which is a significant improvement over Turing, where one data path was devoted exclusively to integer workloads.

This practically doubles raw FP32 throughput core for core, but it won’t immediately result in a frame rate increase in many of your favorite PC games. This means that even though the Nvidia GeForce RTX 3080 only has 68 SMs more than the RTX 2080, its CUDA core count has nearly tripled, going from 2,944 to 8,704. This results in a significant generational leap, increasing the theoretical FP32 throughput from around 10 TFLOPs to 29.7 TFLOPs.

Even though it doesn’t quite reach the “2x performance” goal that we’re sure some people were hoping for, gaming performance sees one of the biggest generational jumps in years when you combine the increase in CUDA cores with enormous gains to cache, texture units, and memory bandwidth. On that, however, later.

Since Nvidia has the RTX designation, the RT cores are also back and include significant upgrades. Second-generation RT cores, which will perform similarly to first-generation RT cores but be twice as efficient, are a feature of Nvidia Ampere graphics cards, notably the RTX 3080.

Ray tracing is a novel technique for rendering scenes realistically that involves having an SM cast a light ray into the scene being drawn, and the RT core will take over from there. The RT core will perform all the calculations required to determine the location of that light ray’s bounce and will then communicate that information to the SM.

The SM can now use the ray tracing data to more accurately illuminate the scene while still being left alone to render the rest of the picture. Unfortunately, we haven’t yet got to a stage where ray tracing isn’t utterly detrimental to performance. Maybe one day, the Nvidia RTX 3080 renders impressively, especially when DLSS is enabled (more on that later).

This time around, tensor cores are also two times as powerful, which is why Nvidia decided to only include 4 in each SM as opposed to the 8 in a Turing SM. With the increased number of SMEs, DLSS performance also significantly improves.

Design

Nvidia chose an entirely new cooler design for the actual Founders Edition graphics card, which is far more practical than the company has ever done with a reference design. To make the rear of the card only a heatsink, the business developed a shorter, multi-layered PCB. This allowed Nvidia to attach a fan to the end of the graphics card, drawing cold air through the heatsink and pushing it upward and out of the chassis.

Since this fan blows hot air straight over the CPU and RAM, we were a little concerned that it may affect the temperatures of these components. However, even with our system, which uses an AMD Ryzen 9 3900X processor with a Noctua NH-12UA air cooler, we saw no performance change. We suppose it helps that most PC games, at least as of yet, don’t genuinely stress the GPU and CPU to the same level.

We have mixed thoughts regarding the new 12-pin power connector, but it is undoubtedly there in terms of power supply. The redesigned PCB layout of the Founders Edition card requires this smaller connector to make this new cooler function; thus, we wish Nvidia’s 2 x 8-pin PCIe to 1 x 12-pin adaptor was a little bit longer.

Even though it’s currently difficult to tie it so that it isn’t instantly evident, aftermarket cards won’t be using it right away. However, it’s important to note that Nvidia is making the 12-pin power connector design open to any manufacturer that wants to utilize it, including AMD.

Also advantageous are the Founders Edition’s three DisplayPort and one HDMI 2.1 output for displays. However, we object to Nvidia’s removal of the USB-C production in this case. Creators will continue to employ this potent card, and today’s top USB-C displays are the only ones trusted by industry professionals.

The Founders Edition is a beautiful piece of hardware in person, despite the few troubles we had with it and the fact that we initially thought it wasn’t lovely. The RTX 3080 has a sleek, all-black design with silver accents that gives it a high-end hardware appearance.

The ‘GeForce RTX’ branding in white on the side of the graphics card is the only source of illumination, which will no doubt please any anti-RGB users out there. Additionally, third-party cards will provide players who truly want to go all-out with rainbow lights that choice.

Global Rating

4.0 Out Of 5 Stars

Testing And Performance

We knew the RTX 3080 would be a fast graphics card just from Nvidia’s own (over-the-top) marketing, but to label it “quick” is a little bit of an understatement. It has been in our home computer since the minute we got the box, running everything from Final Fantasy XIV to Control, and has only been plugged into our test bench for genuine benchmarks.

We had an Nvidia GeForce RTX 2080 Ti in that system before the 3080, and the difference was instantly noticeable before we even started quantifying performance. One of the games we play the most is Final Fantasy XIV, for example. In that game, especially in the most recent expansion, there were times when the RTX 2080 Ti would fail to maintain 60 frames per second at 4K.

With the RTX 3080, such is not the case. In reality, the game often runs at 75 to 100 frames per second at 4K, with the highest graphical settings, compared to the average RTX 2080 Ti’s 60 frames per second. This is a significant improvement in performance for less than half the cost.

No matter what game we played, this story just kept repeating itself. Has Metro Exodus reached its Ray Tracing and DLSS limits? At 4K, smooth locked 60 fps. How can I control the various ray-tracing effects? Slick as silk.

Even Final Fantasy XV has several odd optional graphical enhancements and runs smoothly at 60 frames per second in 4K. Even though we use the term “accessible” exceptionally loosely in this context, 4K@60 gameplay is here.

Even if the thermals in our benchmarks aren’t particularly interesting, you should remember that they were captured on an outdoor test bench. Temperatures in our closed tower, which had two 240mm fans acting as intake, reached a top of about 60°C, significantly lower than the mid-80°C temperatures we regularly observe with the RTX 2080 Ti Founders Edition.

The Nvidia GeForce RTX 3080 is in a league of its own, easily outperforming the RTX 2080 Ti when examining the real benchmark results. A tremendous generational jump when you consider that the RTX 2080 was only 40% faster than the GTX 1080 in our Nvidia GeForce GTX 1080 review from back in 2018; the RTX 3080 starts by being a remarkable 63% faster than the RTX 2080 and 26% faster than the 2080 Ti in 3DMark Time Spy Extreme.

However, Time Spy Extreme isn’t even the RTX 3080’s best-case scenario. Multi-sample anti-aliasing is highly expensive and not worth it, so we maxed out every option that wasn’t MSAA in Red Dead Redemption 2. We noticed a massive 87% improvement from generation to generation.

This isn’t quite the 2x performance boost promised at the RTX 3080 announcement, but it’s certainly closer than we anticipated. Overall, the RTX 3080 outperforms the RTX 2080 by 50–80%, with Fire Strike Ultra being the lone exception when it only managed a 29% edge.

However, that significant performance disparity only becomes apparent at 4K when the graphics hardware is not constrained. In many games in our testing library, even the AMD Ryzen 9 3950X with 64GB of 3,600MHz RAM fell short of the RTX 3080. This explains why, for example, in Metro Exodus, the RTX 2080 Ti and the RTX 3080 are almost comparable at 1080p but have a 19% performance edge at 4K.

Because of this, we don’t advise anyone to purchase the Nvidia GeForce RTX 3080 unless they want to play at 4K, or perhaps 3,400 x 1,440 if you have the greatest ultrawide monitor available that can fully take use of what the RTX 3080 brings to the table. You would be better off buying the Nvidia GeForce RTX 3070 instead of expecting gains at lower resolutions.

However, the Nvidia GeForce RTX 3080’s performance further widens the high-end performance gap between Nvidia and AMD. One of the top AMD graphics cards, especially in gaming performance, is the AMD Radeon RX 6800 XT. However, it can’t compete with the RTX 3080, which outperformed every artificial benchmark we ran.

Product Comparison

The NVIDIA GeForce RTX 3080 is in high demand because it can play 4K games at top-tier framerates. Games like Battlefield 5, Resident Evil Village, and Cyberpunk 2077 will look fantastic if the settings are at their highest.

This powerful card also works with the latest GDDR6X interface and 10GB of VRAM. RX 5700 Xt from XFX is your one-stop shop if you’re an expert software developer or programmer concentrating mainly on 3D rendering and spending a lot of money.

To Sum Up

The RTX 3080 is one of the greatest graphics cards for handling 4K gaming, especially considering its pricing. 4K gaming is exceedingly challenging to operate. You can play all games at their highest settings at or very close to 60 frames per second at this resolution.

The performance requirements for the upcoming generation of video games are likely to soar. The RTX 3080 is substantially more potent than the GPUs in the PS5 or Xbox Series X (at least on paper). Many people kept their 10-series graphics cards because the generational improvements between Nvidia Pascal and Nvidia Turing devices were quite small. If you have one of these previous cards, the RTX 3080 will bring huge improvements.