MSI Gaming GeForce GTX 1660 Ti

Memory Clock 1500 MHz
Base Clock 1500 MHz
Boost Clock 1770 MHz
GPU TU116

When it comes to selecting the best graphics card from among the many options, I confess that I prefer more features to have things cut—you might not use the extra features right away, but at least they're there. However, predicting the future is a daunting challenge, and the gamers have spoken.

The GTX 1660 Ti is compatible with GTX 1060 cards, let alone the more expensive RTX cards. When you move to a complete mainstream build, the GTX 1660 Ti looks a lot better.

The GTX 1660 Ti beats AMD's RX 590 by a large margin in competition, averaging a 20-25 percent advantage. In no game does AMD's similarly priced card come out on top, but it does come close in a few others (Anthem, Battlefield 5, and Strange Brigade).

Real-time ray tracing will sound good, but we'll be fine without it if it degrades results. So, six months after announcing its Turing architecture, Nvidia releases the mainstream options, excluding ray tracing and deep learning. It's everything Turing fans expected, with none of the extra nonsense. The GTX 1660 Ti is here to greet you.

The GTX 1660 Ti's performance against older GPUs is perhaps more significant. It's about 35% faster than the GTX 1060 6GB, and it's also a hair faster than the GTX 1070—though there are certain games where 1070 comes out on top.

Let's talk about overclocking a little more. While the official boost clock is 1890MHz, peak clock speeds approached 2025MHz when we received the Asus card. 

We'll have to dig a little deeper in the coming days to see how much headroom there is, but we used MSI Afterburner to overclock the GPU. First, we increased the voltage to +100, increased the power limit to its maximum (120 percent), and set GPU fans to a constant 80 percent.

The fans aren't even that noisy at this pace, so the large cooler and triple fans are doing their job. Temperatures remained cool during benchmarking (50C max—this is one excellent fruit), and the fans aren't at all that loud at this speed, so the large cooler and better triple fans are doing their job. Then we started checking the GPU core and memory clocks to see how far they could be moved.

It's not to close the gap with an RTX 2060, and I'm interested to see how well the reference-clocked cards overclock, but it's a start. Meanwhile, the memory has plenty of legs—a 25% boost in bandwidth is very impressive.

The good news is that the GTX 1660 Ti performs admirably at 1080p, averaging well over 60 frames per second in most games.

The 1660 Ti is 65 percent quicker than AMD's R9 390 (or the similarly performing R9 290X, which is now over five years old), and it does so while halving graphics card power consumption.

Even so, if you're running anything like an HD 7970 / R9 280X or lower, it'll be a much more noticeable upgrade.

Finally, the GTX 1660 Ti is near the top of the midrange graphics card table in terms of value—depending on the model you're looking at, of course.

Pros
  • Great performance at 1920 x 1080
  • 120W board power compares favorably to AMD competition
  • Latest architecture and NVENC engine
Cons
  • No RT/Tensor cores mean you won't be able to try ray tracing or DLSS
  • No RTX ray tracing capabilities

XFX RX 570

Base Frequency 926-1168 MHz
Boost Frequency 1206-1244 MHz
Max Memory Size 8 GB
Connectivity 1.4 HDR

AMD's RX 570 graphics card is a fantastic deal, and, like the Lords of the Sith, there are always two of them, a master and an apprentice. The Radeon RX 570 came out in tandem with the Radeon RX 580, the apprentice GPU coveting its dark lord's strength. They're two of the best graphics cards you can buy today, with the RX 570, in particular, being a true budget gaming hero.

The RX 570 is of the same silicon as the RX 580, and both are made from the same star stuff that made the RX 470 and RX 480 in the previous generation. This is the Sapphire Nitro+ edition, which has been factory overclocked.

The RX 570 was unquestionably born in the shadow of a GPU rebrand, with AMD's genuinely new, "high-end" RX Vega cards attempting to compete with Nvidia's best. With the Polaris 20 GPU name and a fresh coat of paint, there must be something that distinguishes this card from the previous one.

There is very little difference between the RX 470 and RX 570. It's still Polaris, a GPU architecture proven to be highly durable, as shown by the 12nm RX 590. However, with recent price reductions, the RX 570 has emerged as one of the best budget GPUs available.

AMD has given the Polaris 20 GPU inside the RX 570 a little more juice in terms of usability. The higher clock speeds now consume more power and are rated at 150W thermal design point (TDP), compared to 120W for the RX 470. The memory device is the only other significant improvement.

The RX 570 is still available in 4GB and 8GB configurations, but the GDDR5 memory reference clocks have been increased to 7GHz, giving the new card a higher peak memory bandwidth score than the RX 470. The 500-series card has a transfer rate of 224GB/s, while the 400-series card has a transfer rate of 211GB/s.

It's a card that's still a little more costly than this impressive little Polaris GPU, but it's also much slower. If the RX 570 can match the RX 480 and RX 580, it can comfortably outperform the nearest GeForce GPU. With that level of results, it's also competitive with the GTX 1060.

At the maximum 1080p levels, the RX 570 can offer 60fps or more in a large number of games. For 1440p, it's also not a bad little card. You can also get fantastic 1440p gaming results by lowering your settings to high or medium.

Although overclocked SKUs like this Sapphire card is more expensive, a 4GB standard RX 570 will still offer excellent 1080p gaming performance.

Pros
  • Superb 1080p gaming performance
  • Competitive Pricing
  • Premium build quality
Cons
  • Lags behind Nvidia GPUs in power efficiency
  • Same Architecture As Previous-Gen

Asus GeForce GTX 1050 Ti

NVIDIA CUDA Cores 768
Frame Buffer 4 GB GDDR5
Memory Speed 7 Gbps
Boost Clock 1.3x

The GTX 1050 Ti was the first genuinely affordable Pascal graphics card, with the rallying cry of "PC gaming for everyone!" It's aimed squarely at the segment of the industry that AMD has been attempting to monopolize.

The GTX 1050 Ti's main selling point is its low power consumption, making it a simple upgrade for those with a spare PCIe slot. While using precisely the same graphics architecture as previous Pascal cards, the GTX 1050 Ti is a very different beast.

The GP107 GPU in the GTX 1050 Ti does not use Nvidia's previous 16nm FinFET transistor lithography. It has moved to Samsung's smaller 14nm node instead of the TSMC-manufactured process. Due to self-immolation, GTX 1050 Ti cards are not allowed on flights, which is hilarious. Though this card runs cool enough that you might chill it passively.

The disparity between the 14nm and 16nm lithographies would be more noticeable on the specs sheet than in real-world results.

The other cards' general computing clusters GPCs of five blocks of streaming microprocessors (SMs), each with 128 CUDA cores; the GP107's GPCs only have three SMs.

The GTX 1050 Ti has 768 CUDA cores, distributed between those six SM modules due to the different GPU configurations. With 48 texture units and 32 ROPs, this card has the same design as the GTX 950 it is supposed to replace. The stock clock speed of the GP107 GPU in the GTX 1050 Ti is 1,290MHz, with a boost clock of 1,392MHz.

This overclocked MSI version is rated at 1,342MHz and 1,455MHz, respectively, but both the stock and OC cards can easily surpass those clock speeds once you start playing GPU-intensive games.

The power configuration is one of the most significant aspects of the GTX 1050 Ti. It has a thermal design power (TDP) of just 75W, which means it can get all of its frequency from the motherboard's PCIe slot and doesn't need any additional power connections from the PSU.

When it comes to 1440p resolution, the GTX 1050 Ti is also very capable. With all the bells and whistles turned up to eleven, you won't be able to reach 60fps. Still, with some judicious use of GeForce Experience's optimizations, you'll be able to get better performance at a sharper resolution.

On the overclocking front, I was able to get our MSI card up to around 1,900MHz, though the 75W power limit imposed by the PCIe slot was frequently throttling the GPU to around 1,850MHz. You'll get a few extra frames per second on average, but you won't be able to drive the GP107 silicon to its maximum capacity without an additional PCIe power connector.

Pros
  • Extremely efficient
  • Excellent price
  • HDMI 2.0b, DisplayPort 1.4
Cons
  • No SLI support
  • Some games need to be dropped to Medium settings

GeForce GT 1030

Boost Clock Speed 1468 MHz
Base Clock Speed 1227 MHz
Interface PCI Express 3.0 x4
Supported APIs DirectX

The GeForce GT 1030 from Nvidia performs admirably in the competitive gaming titles. Furthermore, the card's low price, power-efficient architecture, and appealing form factor make it open to almost anyone with a PCIe slot.

The GTX 1660 Ti's performance against older GPUs is perhaps more significant. It's about 35% faster than the GTX 1060 6GB, and it's also a hair faster than the GTX 1070—though there are certain games where 1070 comes out on top.

The GPU operates at a frequency of 1228 MHz to 1468 MHz, and the memory runs at 1502 MHz (6 Gbps effective). The NVIDIA GeForce GT 1030 is a single-slot card that does not need an additional power connector and has a maximum power draw of 30 W. 

The GP108 is the GPU that drives the GeForce GT 1030, a brand-new graphics processor with 1.8 billion transistors. It's a tiny object, weighing only 70mm2, and it consists of using the same 14nm FinFET method as the GP107.

The GeForce GT 1030 fits with 2,048 MB GDDR5 memory and a 64-bit memory interface from NVIDIA.

There are two display outputs: one DVI and one HDMI. A PCI-Express 3.0 x4 interface links the GeForce GT 1030 to the rest of the device. The card measures 145 mm x 69 mm x 18 mm and comes with a single-slot cooling solution. In terms of visual performance, the GT 1030 is very similar to the PS4. It offers incredible capacity for its 30W TDP.

It supports a wide range of AAA games, including recent releases. In terms of performance, it's on par with a GTX 750 Ti. On Low to Medium settings in 1080p, you can run games like Far Cry 5, Battlefield 1, Assassin's Creed Origins, and others and get good frame rates above 30 and with optimization up to 900p Medium to High settings.

Older games, such as GTA V and Far Cry 4, benefit significantly from Medium-High settings in 1080p. When playing Fortnite with a GeForce GT 1030, we can see that the FPS drops to 32 frames per second. Those frames are obtainable while playing Fortnite with a 1920x1080 screen resolution and High graphics.

The GeForce GT 1030 is capable of running at 1440p Medium. On the relatively large resolution setting of 2560x1440, it returns a playable 36 frames per second.

You can run Fortnite at 1080p screen resolution to get a good frame rate. Any resolution higher than 1080p might be less than ideal. Low: 61 FPS, Medium: 48 FPS, High: 32 FPS, Ultra: 23 FPS are the frame rates at 1080p on the various graphics settings.

Pros
  • Very affordable
  • Excellent performance in eSports and DirectX9/11-based games
Cons
  • Trails AMD's Radeon RX 550 in DX12/Vulkan-based games

XFX Radeon RX 560

Compute Units 14/16
Base Frequency 1175 MHz
Memory Speed 7 Gbps
Max Memory Size 4 GB

AMD's Radeon RX 560 is a mid-range graphics card, based on the Polaris 21 graphics processor and built on a 14 nm process, the menu supports DirectX 12 in its Polaris 21 XT version. The RX 560 has nearly double the hardware specifications, resulting in a significant performance boost. The Radeon RX 560 can compete directly with the GTX 1050.

When combined with a 6% boost clock speed increase from 1200MHz to 1275MHz, a minimum increase anticipates the overall effective speed of 10% over the 460. In terms of price and performance, the RX 560 is a new competitor to NVIDIA's GTX 1050.

All modern games will run on the Radeon RX 560 as a result of this. One thousand twenty-four shading units, 64 texture mapping units, and includes 16 ROPs. The Radeon RX 560 has 4 GB of GDDR5 memory connected through a 128-bit memory interface.

The GPU runs at 1175 MHz, with the potential to be boosted to 1275 MHz, and the memory runs at 1750 MHz (7 Gbps effective). The AMD Radeon RX 560 does not need an additional power connector because it is a dual-slot card, and its maximum power draw is 75 W. 1x DVI, 1x HDMI, and 1x DisplayPort are available as view outputs.

A PCI-Express 3.0 x8 interface links the Radeon RX 560 to the rest of the device. The card has a dual-slot cooling solution and measures 170 mm in length. It cost 99 dollars when it first came out.

We'll look at the RX 560, which bridges the performance gap between eSports and modern AAA games that demand every ounce of strength. While the RX 550 is on the top of eSports gamers.

AMD FreeSync Technology, HEVC 4K decoding, and AMD Eyefinity Technology are all supported by the RX 500 series.

For higher resolutions up to 1920x1080, the Radeon RX 560 4GB is the best option. For optimum performance, we recommend a Performance Processor and 8GB of RAM.

The AMD RX 560 has a score range of just 7.54 percent (95th to 5th percentile). This is a relatively small selection, suggesting that the AMD RX 560 performs admirably under several real-world conditions.

Pros
  • Compelling 1080p performance at reduced detail settings
  • Manageable heat/power consumption
  • Radeon RX 560 4GB is reasonably priced if you find a $120 card
Cons
  • Great eSports performance
  • Can play AAA games at Medium settings

GeForce 1660 Super

Memory Clock 1750 MHz
Boost Clock 1785 MHz
Base Clock 1530 MHz
TDP 125 W

The new GeForce GTX 1660 Super strategically positioned entry to Nvidia's swelling portfolio of popular graphics offerings as a budget-focused graphics card designed for 1080p gaming. It's just a minor upgrade from the non-Super GTX 1660, but it's powerful enough to hold esports fans and AAA gamers playing late into the night. For the time being, it comes similar sufficient to the next-up GeForce GTX 1660 Ti in terms of price to earn our Editors' Choice award for 1080p gaming cards.

Key Features and Performance

If you're on a budget, the Nvidia GeForce GTX 1660 Super is an excellent option. It improves on the GTX 1660 to take it to the next stage in terms of efficiency.

The GTX 1660 Super replaces the original GTX 1660's GDDR5 memory with GDDR6, almost doubling its clock speed from 8 to 14 Gbps and increasing bandwidth from 192 to 336 Gbps.

It has more video memory and higher speeds than the original 1660.

Ray tracing is theoretically possible; it will turn any game you attempt to allow it into a slideshow – more so than any other graphics card.

The GTX 1660 Super dominated the less taxing Middle Earth: Shadow of War, averaging 79 frames per second at 1080p with Ultra quality settings.

It has 1,408 CUDA cores, the same as the original 1660. The GTX 1660 Super was around 5% slower in Time Spy than the GTX 1660 Ti, which costs considerably more, and a staggering 14% faster than the GTX 1660 it effectively replaces. The gap was also smaller in Fire Strike, where it was just 2% slower than the GTX 1660 Ti.

Design and Cooling

The GTX 1660 Super can withstand some heat even at 1440p, but cracks begin to appear. In Middle Earth: Shadow of War, the graphics card manages a decent 54 frames per second.

With a black plastic shroud that houses two fans within a housing that measures 6.8 inches long, this double-slot-width card looks like every other card in the Zotac Twin Fan family. By most card standards, that's a slim card, and it's small enough to fit into several lightweight builds. Three DisplayPort 1.4b ports and one HDMI 2.0 port are located on the back of the card.

Final Verdict

If you've been debating whether or not to purchase a standard video card for 1080p gaming, the GeForce GTX 1660 Super could be the catalyst you need: It's a good-looking, reasonably priced alternative to the GTX 1660 and GTX 1660 Ti.

In a nutshell, it turns out that 1080p is the most common resolution for PC gaming, and most people would spend their money on graphics cards in this price range. And, before the AMD Radeon RX 5500 arrives in November, Nvidia has a stranglehold on this segment of the market, and it's easy to see why.

Pros
  • Good 1080p performance
  • Better value than 1660 Ti
  • Doesn't get too hot
Cons
  • Missing RTX features
  • Only 1 HDMI and DisplayPort