ASUS GeForce GTX 1660 Ti STRIX OC 6 GB

Interface Type PCI Express 3.0 x16
Max Resolution 7680 x 4320
Interfaces 2 x DisplayPort
Bus Type PCI Express 3.0 x16

The most efficient Asus ROG Strix GTX 1660 Ti OC 6 GB is a versatile graphics card that will only improve with use; this is a mid-range GPU that can handle the heat from intense overclocking thanks to its passive cooling system.

It has higher efficiency and includes all of the latest RTX and Tensor core-powered features that aren't available on any of Nvidia's recent GTX cards. When it comes to budget GPUs, the Nvidia GeForce GTX 1660 Ti is unrivaled. Although the GTX 1660 Super is a strong competitor, the Ti is the best GPU for budget gaming. It is, after all, Nvidia's Turing line's fastest-growing card.

Nvidia manages to deliver such good performance for such a low price. They achieved this by that the Turing architecture to its bare bones, removing the RT and Tensor cores responsible for RTX and DLSS (Deep Learning Super Sampling) technologies.

As a result, you get the same primary output as RTX graphics cards, but without the bells and whistles and at a lower cost. It achieves the ideal combination of price and efficiency.

The Nvidia GeForce GTX 1660 Ti is not only ideal for someone seeking to get the most out of Nvidia Turing on a budget PC, but it also has sufficient strength to be regarded as the best graphics cards on the market. It's an asset in our book, particularly since it has the highest factory overclock of any mid-range graphics card we've seen so far.

Furthermore, thanks to its massive heat sink and triple axial fans, this card runs 10 degrees cooler than most of its competitors.

ASUS' Republic of gamers card, which features custom cooling and a redesigned custom PCB. The board is powered by a single 8-pin power connector and contains a few surprises. One is a BIOS with a silent mode that runs at slightly higher temperatures but provides similar performance to the 'performance' mode switch. Even though the GDDR6 memory hasn't been updated, the ICs are stock 12 GHz (effective clock-rate) but can be boosted to 1415 GHz with a flick of your fingers.

To deal with the poor RTX sales, NVIDIA pulled a rabbit out of its hat. The GeForce GTX 1660 Ti and lower SKUs (which will be available) are intriguing. ASUS has a great deal to offer with the latest STRIX revision of the GeForce GTX 1660 Ti regarding the AIB product. This 6GB card delivers excellent output right out of the box, boosting to 1860 MHz. As a result, it has a 4% advantage over the standard requirement.

We assume that the new GeForce GTX 1660 Ti series is just what the industry requires. Because of the high sales price of the range, most consumers have put RTX cards on hold. The GeForce GTX 1660 Ti does a great job of addressing the need for a more affordable product.

Pros
  • It offers great performance for a good price
  • It aims for the highest video quality in 1080p
Cons
  • Does not offer support for ray tracing
  • It does not bundle any game

Nvidia GeForce GTX 1070

GPU Variant GP104-200-A1
Memory Clock 2002 MHz
Boost Clock 1683 MHz
Memory Speed 8 GB GDDR5

The GTX 1070 is a fantastic topmost choice that introduces previous-generation frontrunner output at a discounted rate. It's on par with 1080, and you might also say that 1070 is a superior product due to its lower cost per frame ratio. For those who prefer 1440p gaming, this is the best choice.

The Nvidia GeForce GTX 1070 is the successor to the GTX 970, which was the most excellent pick for your buck graphics card of its type at the time. The 4GB (or 3.5GB if you're pedantic) GTX 970 had 1,664 CUDA cores and a boost clock of 1,178 MHz, making it more than capable of powering the best PC games at 1080p – and it still is. Quick forward to the Nvidia GeForce GTX 1070, and it has dethroned its budget-friendly predecessor.

The GTX 1070 has 256 additional CUDA cores, taking the average to 1,920 and a higher boost clock of 1,683MHz. Plus, there's an extra 4.5GB of GDDR5 VRAM. The Nvidia GeForce GTX 1070, of course, is developed on Pascal's mighty 16nm FinFET production line, which allows for 6.5 TFLOPs of overall output. That leads it on par with some of the best graphics cards on the market. For example, Titan X, but at a fraction of the cost.

Nvidia's Pascal architecture helps them load more transistors into a smaller piece of silicon, improving performance while lowering power consumption to only 150W.

Pascal also includes Nvidia's solution for asynchronous computing. AMD's async compute technology allows a GPU to simultaneously operate on both graphics and computing tasks, allowing a card to complete tasks faster. Nvidia's variant of the technology is known as "pre-emption," It functions somewhat differently: it enables the GPU to pick which processes to evaluate more intelligently.

The 1070's clock speed has been increased to 1,506MHz, which is a significant improvement over the 970's core clock speed of 1,050MHz; the 1080's clock speed is marginally higher at 1,607MHz. The VR aspect of the argument, in our opinion, helps us to overlook the lack of SLI choices. This is especially certain when you consider how much work Nvidia has put into optimizing 1070 for virtual reality. The Lens Matched Shading, Simultaneous Multi-Projection technology used in 1080, was also used in 1070.

Pros
  • Very quiet
  • Fans turn off in idle
Cons
  • High price
  • Older GDDR5 memory

XFX RX 580 8GB

Memory Clock 2000 MHz
Boost Clock 1340 MHz
Base Clock 1257 MHz
TDP 185 W

The XFX RX 580 8GB is the latest Radeon lineup's wise elder statesman. The RX 580 has managed to live up to AMD's pledge of "fine wine" in all of its silicon. It is now the graphics card that we would suggest for your gaming machine. It's a powerful, reasonably priced graphics card with a sizable pool of video memory and the ability to work with modern graphics APIs.

In terms of overall gaming efficiency, the RX 580 is just a few frames per second behind the newer RX 590, and the gap usually is only a few frames per second. Given the considerably higher clock speeds provided by the new 12nm cards, this is surprising.

The 'enhanced' part comes from the fact that the 14nm FinFET technology and the 4th Gen GCN architecture used in the new AMD Radeon cards had a whole year mature after the initial Polaris update. As a result, the manufacturing process and yields improved, and the resulting GPUs became more stable. That's why XFX became able to release RX 580 cards with a higher base clock speed than the RX 480 cards available at launch. The initial RX 480's base and boost clocks were 1,120MHz and 1,266MHz, respectively, while the RX 580's Polaris 20 chip's reference spec is 1,257MHz and 1,340MHz.

Aside from the look at the critical speed bump, the RX 580 is identical to the RX 480 in GPU performance. The core configuration remains the same – the 14nm Polaris 20 in the new card has 36 compute units (CUs), and it also contains 2,304 stream processors shared among them. It also includes the same 144 texture units and 32 ROPs as before.

The memory system is identical, with 8GB of GDDR5 providing a total memory bandwidth of 256GB/s. The RX 580, like the 400 series cards, comes in 4GB and 8GB variants, as well as the RX 570.

Overall, the XFX Graphics Core Next architecture outperforms the competition in terms of gaming efficiency. There are a few games where the Nvidia GPU has the upper hand, but they are all last-gen DirectX 11 titles. The AMD silicon outperforms the GeForce card in more modern APIs, such as DirectX 12.

When you raise the gaming resolution, the XFX card's enhanced memory subsystem shines through. The RX 580 in 8GB trim can better cope with high-res textures' physical demands and the extra total pixel count thanks to an additional 2GB of GDDR5 memory and a wider 256-bit memory bus.

The RX 580 is today's best graphics card for the money on the market right now. With the mining boom's price concerns long gone, the RX 580 8GB cards can now be had for as little as $190 (£196), making them a relative bargain.

Pros
  • Ideal for 1080/1440p gaming
  • Very competitive performance
Cons
  • Efficiency

MSI GeForce RTX 3060 Gaming X Trio

Memory Clock 1875 MHz
GPU GA106
Transistors 13250M
Memory 12 GB, GDDR6, 192-bit

The new MSI GeForce RTX 3060 Gaming X Trio is the company's premium interpretation of NVIDIA's new GeForce RTX 3060 "Ampere" graphics card. Cards in this category are quick enough to max out a game at standard resolutions or play at higher resolutions with lower settings.

The MSI RTX 3060 Gaming X Trio looked impressive when overclocked. In real-life performance, we improved by almost 10%. Overclocking was more substantial than on any other RTX 3060, thanks to the low temperatures and excellent VRM circuitry.

The RTX 3060 has 3,584 CUDA cores, 112 Tensor cores (3rd generation), 28 Ampere RT cores, 112 TMUs, and 48 ROPs. NVIDIA sweetens the deal by increasing the memory to 12 GB from 8 GB on the RTX 2060. The 192-bit GDDR6 memory bus width and type remain unchanged. The memory clock is marginally increased to 15 Gbps.

It's the first time the latest 8 nm "GA106" silicon has been seen on a desktop platform, giving NVIDIA's custom board partners a lot of leeway to either bulk up the chip, like this MSI Gaming X Trio card does, or keep costs down to come up with easy designs that don't break the bank.

It's also intended to provide a significant performance boost over the GeForce GTX 1060 6 GB "Pascal," another standard NVIDIA graphics card. The card is capable of 1440p gaming with moderate information or 1080p gaming with full ray tracing.

The latest GeForce Ampere architecture introduces the second generation of RTX Technology, which includes new Ampere CUDA cores with concurrent FP32+INT32 math performance, 2nd generation RT cores that double intersection performance over the previous generation, hardware for ray-traced motion-blur effects.

MSI's highest level GeForce RTX 3060 Gaming X Trio boosts the RTX 3060's reliability. This card has an over-specced VRM solution and a beefy Tri Frozr cooling solution that's configured for deficient gaming noise levels is loaded with all the RGB bling you might ask for and looks high-end when mounted.

MSI has also set the maximum GPU Boost frequency to 1852 MHz, giving the card its highest processing plant speeds (compared to 1777 MHz reference). MSI hasn't sent us any pricing details for this card, but we expect it to be considerably more expensive than the $329 starting price, which is unlikely to last more than a day or two.

We discovered a 5% gap between the best and worst overclocker, which is significantly more than what you get from any card's factory overclock. Memory over-clocking yielded substantial gains, which is unsurprising considering that these chips are officially rated for 16 Gbps.

Pros
  • Factory Overclocked out of the box
  • Great overclocking potential and performance
  • ZeroFrozr Technology Included on Torx 4.0 fans
Cons
  • No word on pricing
  • Very tall and huge graphics card

MSI GeForce GTX 980 Ti Gaming 6 GB

Memory 7010 MHz
Technologies Microsoft DirectX: 12 OpenGL: 4.5
CUDA Yes
Interface PCI Express 3.0 x16

Testing the latest MSI GeForce GTX 980 Ti Gaming (OC edition) is an excellent product with entirely silent cooling. It comes factory overclocked, and when combined with the default variables, such as the 6 GB graphics memory, this product exudes awe-inspiring gaming performance. There's a significant Maxwell GPU, and it's a phenomenal player with that kind of game rendering power.

The 980 Ti is based on the Large Maxwell GPU, which is the same GPU that drives the Titan X. the product has been reduced in size, but there is still plenty of performance to be had. The GTX 980 Ti has a spacious six Gigabytes of graphics memory, and with these specifications, the GTX 980 Ti can pique the interest of genuine gamers.

The GPU powering the GeForce 980 Ti is whopping, with a massive switching frequency; it's a slightly updated GM200 A1 GPU, which also powers the Titan X. So, it's a somewhat different GM200 iteration.

In contrast to the GeForce 980, this product has a nice 6 GB frame buffer and nearly a third more shader processors, accumulating up to 2816 of them playing the binary game in a GPU whopping 8 Billion transistors (GeForce GTX 980 has 5 Billion).

In terms of recollection, NVIDIA's GeForce GTX 980 Ti features 7 Gbps memory, the fastest GDDR5 memory currently available on a graphics card, at least before the competition releases HBM (stacked memory) soon. When combined with GPU Boost 2.0, this product's dynamic clock is marketed in the 1076 MHz range for reference products. The 980 Ti's reference base clock is 1 GHz. It's not that the card can't go heavier; instead, it's ready to keep the product's energy consumption in check.

Monitor interfaces for the GeForce GTX 980 Ti include DVI, HDMI 2.0, and DisplayPort; however, this will vary slightly with board partner products released after Computex, depending on their development and cooling.

Through a card like the MSI GeForce GTX 980 Ti gaming, you'll be able to play the latest games at a whopping Ultra HD 8.2 Mpixels at a 3840x2160 resolution with only one card, as we'll see in this analysis.

The standard core clock on the MSI GeForce GTX 980 Ti Gaming (OC edition) is 1178 MHz. The MSI GeForce GTX 980 Ti Gaming (OC edition) comes with a newly redesigned V TwinFrozr cooler with slight aesthetic improvements and, hey, still appears passive when the product is idle to keep stuff cold.

When the GPU is under a specific temperature and light GPU load, the fans are turned off. In Call of Duty: Modern Warfare, the 980 Ti spits out over 60 frames during the test, for an average of 81 frames per second. It was also only 7% slower than the GTX 1070 in terms of performance.

Pros
  • Impressive out-of-the-box overclock pushes past Titan X-level performance.
  • Can handle most 4K gaming without the hassle of SLI.
Cons
  • Open-air cooler means case fans need to remove card heat from chassis.
  • Slightly pricier than stock GeForce GTX 980 Ti.