XFX Radeon RX 590

Memory Clock 2000 MHz
Boost Clock 1545 MHz
TDP 175 W
Base Clock 1469 MHz

The Radeon RX 590 is much bigger, with double-digit percentage gains in our testing over the Nvidia GeForce GTX 1060, which stubbornly sticks in the same price range.

Key Features and Performance

AMD can now squeeze even more efficiency out of the Polaris architecture thanks to the newer method. The Radeon RX 590 has a core clock of 1,469MHz, unrivaled by the Radeon RX 580. This is a 12 percent improvement over the Radeon RX 580's 1,257MHz clock and a 31% increase over the Radeon RX 480's 1,120MHz clock.

The Radeon RX 590's truly immersive resolutions are 1080p and 1440p; a 4K resolution is usually out of control in today's AAA games without lowering the detail settings significantly.

The RX 580's TDP hovers about 180W, and the 590 can maintain that range thanks to its higher clock frequencies. The 36 CUs (compute units aka shader clusters) x 64 shader processors per CU = 2304 shader processors are still present in the RX 590. The GPU has been clocked at a boost frequency of 1580 MHz by PowerColor.

The ram partition shows 8GB of 256-bit GDDR5 memory with an efficient data rate of 8 Gbps. It can play the new games while still providing a reasonable memory capacity to price ratio in the 1920x1080 and even 2560x1440 monitor resolutions. The Devil May Cry 5, Division 2, and Resident Evil 2 games will be included with the 590 cards.

Design and Cooling

The Fatboy's main heatsink is much thicker than the RX 580 GTS ones, measuring 25.96mm vs. 16.96mm. The heatsink also has a gross surface area of 368,866mm2, which is almost 125,000mm2 more than the RX 580 GTS. The heavy heatsink, which sits atop a unibody VRM heatsink that enables improved heat transfer to the primary cooling apparatus, is laced with thick composite copper tubing.

This is a significant cooler with a triple-slot graphics card, as anticipated. XFX used a branded backplate with the Radeon RX 590 Fatboy to protect all that heavy metal from sagging and to help dissipate heat.

The Radeon RX 590 Fatboy is fitted with three DisplayPorts, an HDMI port, and DVI-D. To control the graphics card, you'll need an 8-pin and 6-pin power connector, and XFX recommends a 600-watt power supply, up from the 550W specified for XFX's GTS RX 580 GPUs.

Final Verdict

The AMD Radeon RX 590 makes a strong case for value in the mid-range graphics card industry. Across all of our synthetic and real-world benchmarks, it outperformed the Nvidia GeForce GTX 1060 6GB by an average of 14%. In some games, the advantage was greater than 30%.

Pros
  • Fast enough for Full HD
  • 8 GB of GDDR5
  • Suitable for CrossFireX
Cons
  • Higher power consumption than comparable Nvidia chips
  • Limited overclocking potential

Gigabyte GeForce GTX 1660 Ti

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

The GeForce GTX 1660 Ti Gaming OC 6G from Gigabyte is the company's highest model, with overbuilt ventilation, a generously high power ceiling, and RGB lighting.

The GeForce GTX 1660 Ti from Nvidia was the first Turing graphics card to be released without the architecture's signature RT and Tensor cores. To compete with AMD's Radeon RX 590 and the previous-generation GeForce GTX 1070, it solely relied on consistency and quality enhancements in today's sports.

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 strongest GPU for budget gaming. It is, after all, Nvidia's Turing line's fastest-growing card. How did Nvidia manage to deliver such high performance for such a low price?

They achieved this by that the Turing architecture to its bare bones, removing the RT and Tensor cores that are responsible for RTX and DLSS (Deep Learning Super Sampling) technologies. As a result, you get the same basic results as RTX graphics cards, but without the bells and whistles and at a lower cost. It achieves the ideal combination between cost and efficiency.

The Nvidia GeForce GTX 1660 Ti is not only suitable for someone looking to get the most out of Nvidia Turing on a budget PC, but it still has enough strength to be considered one of the best graphics cards on the market. The Gigabyte GeForce GTX 1660 Ti Windforce OC is the most comparable GeForce GTX 1660 Ti available. It just has a dual-fan cooler that is in the center of the bag.

According to Tom's Hardware, Gigabyte compensates for this by using 3D-shaped curves on the fan blades for highly efficient airflow, which is further enhanced by the graphics card's fans rotating in opposite directions. A 1,845MHz boost clock completes the package. This card unquestionably provides better value at a lower price.

The perfect cure for addressing turbulent airflow induced by neighboring fans is GIGABYTE's proprietary "Alternate Spinning." Because neighboring fans spin in the same direction, the airflow direction between them becomes reversed, creating chaotic airflow and lowering heat-dissipation performance. GIGABYTE reverses the direction of neighboring fans so that the airflow direction between them is the same, eliminating noise and increasing airflow pressure.

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

NVIDIA GeForce GTX 1070 Ti

GPU Name GP104
Memory Clock 2002 MHz
GPU Variant GP104-300-A1
TDP 180 W

The Nvidia GTX 1070 Ti astounds and impresses us with production figures nearly identical to those of the Nvidia GTX 1080 at a much more affordable price. Nvidia's new GPU, like previous Ti cards, adds more CUDA cores and higher-frequency capacity to the already-amazing GeForce GTX 1070.

Nvidia has steadily raised the status of its x70 card from the celebrated mid-range GTX 670 to the high-end popularity that the GTX 1070 currently enjoys, so a Ti variant was unavoidable.

This GPU is a mix of those two cards, apart from any clever placement on the part of the component manufacturer. It combines the GTX 1070's usable video memory and increases clock speed with the GTX 1080's CUDA core count and base frequency.

In the end, it's a far more approachable high-end GPU than the Nvidia GTX 1080's astronomical price. Furthermore, the Nvidia GTX 1070 Ti matches the performance of its older brother – but more on that later.

This GPU borrows the GeForce GTX 1080's 5-phase dual-FET power supply, raising the card's TDP to 180 Watts. The GTX 1070 Ti has an improved copper vapor chamber, which lets it maintain a cool head when overclocking.

Unsurprisingly, the Nvidia GTX 1070 Ti outperforms the GTX 1070, but it's incredible how closely it matches the GTX 1080. Although our synthetic tests reveal a significant difference between the two, frame rates in games reveal a graphics card that performs nearly as well as a much more expensive component.

In synthetic simulations, the Nvidia GTX 1070 puts up a fair battle against the AMD Radeon RX Vega 64. Similarly, the GTX 1080 Ti still achieves marginally higher temperatures than its predecessor, though 77 degrees. Celsius is still well within the Pascal architecture's healthy operating range.

According to Arstechnica, the Nvidia GTX 1070 Ti is the ideal card to modernize your gaming PC whether you've been stuck with an older graphics card or a lower-tier GPU. It costs a fraction of the price of the Nvidia GTX 1080 while providing virtually identical output. And, most of all, it's just a few bucks more expensive than the standard Nvidia GTX 1070.

Pros
  • Very quiet
  • Fans turn off in idle
  • Beats the Radeon RX Vega 56
Cons
  • High price
  • Additional 6-pin power input barely needed

EVGA RTX 2070 Black Gaming

Graphics Ram Size 8 GB
Chipset Brand NVIDIA
Graphics RAM Type GDDR6
Graphics Co processor NVIDIA GeForce RTX 2070

EVGA RTX 2070 Black Gaming is the most affordable one with real-time ray-tracing in games or at least a semblance of it. The RTX 2070 is being offered to the vast bulk of gamers that play at 1440p resolution or lower. For the RTX 2070, NVIDIA has made some fascinating design decisions.

Compared to all of our comparisons, the EVGA RTX 2070 Black is only 2% sluggish than the RTX 2070 Founders Edition, which is a small difference. According to TechPowerUp, the card is 12% quicker than the GTX 1080 and 12% slower than the GTX 1080 Ti, putting it in the middle of the pack. The RTX 2080 and RTX 2080 Ti, NVIDIA's flagship cards, are 22 percent and 49 percent faster, individually.

Previous generations of this card, such as the GTX 1070 and GTX 970, used the same chips as the SKUs above them, such as the GTX 1080 and GTX 980. The RTX 2070 is based on NVIDIA's third-largest "Turing" chip, the TU106, rather than the TU104.

It's worth noting, though, that the TU106 isn't a chip successor in the same manner that the GP106 or GM206 are. While those two have precisely half the muscle of the GP104 or GM204, the TU106, instead of the TU104, has half the muscle of the top-dog TU102.

NVIDIA RTX is a comprehensive real-time ray-tracing model for consumer gaming that adds real-time ray tracing to 3D gamers.

The TU106 is the third-largest chip built on the "Turing" architecture, and it differs from chips like the GP106 in that it has half the number-crunching machinery of the largest TU102 chip, but not half that of the TU104, as we described earlier. This helps NVIDIA build the RTX 2070 with more than 3/4 the amount of CUDA cores as the RTX 2080 without losing expensive TU104 die by disabling CUDA cores that aren't always fully functional.

At the highest level, the GPU uses PCI-Express 3.0 x16 host networking to link to GDDR6 memory via a 256-bit wide GDDR6 memory bus is the same memory interface RTX 2080 and the TU104 it's built on. As a result, each GPC contains 768 CUDA cores, 96 Tensor cores, and 12 RT cores, totaling 2,304 CUDA cores, 288 Tensor cores, and 36 RT cores around the TU106 silicon.

EVGA's dual-slot, the dual-fan cooler does a great job keeping the card cool with temperatures of only 64°C during gaming. Gaming noise levels are pretty much identical to the RTX 2070 Founders Edition, which is good.

The RTX 2070 Black, which is 16 percent faster but equally priced, has significantly better power, heat, and noise than AMD's fastest, the Vega 64. Overall, the RTX 2070 is an outstanding option for 1440p gaming with full data, delivering at least 60 frames per second in all recent games. It's also possible to play 1080p games at a high frame rate.

Pros
  • Good performance
  • Better pricing than other models
  • Overclock capability
Cons
  • Not much faster than GTX 1080
  • Waiting for DLSS games

PowerColor 'Red Devil' Radeon RX 590 8GB

Engine Clock Up to 1560 MHz
Stream Processors 2304
Memory Size/Bus 8GB
Memory Clock 8400 MHz

The PowerColor Red Devil RX 590 features 8GB of GDDR5 ram, 2304 stream processors, and a memory clock speed of 2000MHz, all connected by a new high-speed 256-bit memory interface.

GDDR5 memory has the maximum memory bandwidth of any memory technology currently available, allowing for improved GPU performance. AMD App Accelerator produces a "co-processing" environment in which the AMD RadeonTM Graphics processor's compute processing capacity collaborates with the device central processor to accelerate enabled applications. For the best output, the PowerColor RX 590 Red Devil comes with an 8Pin + 6Pin power connector with superior 6+1 multi-phases. Dr.

With breathtaking 3D visual effects, natural lighting, and lifelike visuals, you'll get an intense gameplay experience and unrivaled picture quality. The new Windows10 features and driver support are supported by this comprehensive OS.

Mos is included in the platinum power package to provide greater power performance and reliability. PowerColor Red Devil RX 590 enlarges the surface of the heatsink and equips it with 8mm heat pipes to obtain a stronger thermal solution. This reduces the temperature while increasing efficiency.

One distinction from the Radeon RX 580 and its 14nm process is that AMD's new Polaris graphics card uses GlobalFoundries' 12nm process node. The 2,304 Graphics Core Next (GCN) stream processors, 144 texture modules, 32 ROPs, and 8GB of GDDR5 memory remain largely unchanged. It also has a 256-bit memory bus, with a total memory bandwidth of 256GB/s.

The most noticeable change is most likely the increased clock rates. The Red Devil card from PowerColor will rise to 1,576MHz, compared to 1,380MHz on the Radeon RX 580 edition. This is a 196MHz upgrade, placing the card ahead of the NVIDIA GeForce GTX 1060 6GB.

The larger cooler, which has five heat pipes and a dense fin stack to have more surface space for heat dissipation, accounts for most of its bulk. The card has a TDP of 225W, giving NVIDIA a significant advantage in terms of power consumption.

Superior scalability is given by multi-GPU support. With AMD CrossFireTM technology, we can boost our gaming efficiency by up to 2x, 2.75x for TriFire setups, and 3.75x for QuadFire configurations. Expand the game field of view across all screens by running various displays from a single graphics card.

AMD's combined graphics processing and computing cores have a new architecture that allows for higher usage for better performance and reliability.

4K resolution monitor support and integrated high-speed HDMI performance with 1080p 120Hz 3D Stereoscopic support.

For a quicker GPU, it has double the bandwidth per lane as PCIe Gen 2. Components that are the most solidly built improve stability, ultimate performance, and quality. A power-saving feature that controls frame rate dynamically depending on your in-game movements. When available, RadeonTM Chill will increase power efficiency and lower temperatures for supported products and games.

With Radeon VR Ready Premium solutions, you can immerse yourself in lifelike augmented reality environments. Better security by assisting in protecting the pre-startup—or pre-boot—process from bootkit threats and quicker startup times and hibernation resuming. With AMD's VSR, you can get content that rivals 4K when playing your favorite video games on a 1080p monitor.

Pros
  • Faster than Radeon RX 580 and GeForce GTX 1060 6GB
  • Strong game bundle
  • Ideal for 1080P gaming
Cons
  • Too expensive compared to Radeon RX 580
  • Unable to hold rated core clock rate in a closed case

MSI GeForce RTX 2060

Memory Clock 1750 MHz
Boost Clock 1680 MHz
Base Clock 1365 MHz
TDP 160 W

The new GeForce RTX 2060 surprised reviewers and gamers alike with outstanding results and plenty of custom cards were available on day one, thanks to NVIDIA's latest Turing GPU update.

With its excellent performance, a slew of custom cards (including the RTX 2060 GAMING Z), and other features, NVIDIA had already established the GeForce RTX 2060 as one of the best RTX updates. The GeForce RTX 2060 is capable of 2560 x 1440 gaming also in its Founders Edition version, with over 60% more performance than the previous-gen GeForce GTX 1060.

This means we're talking about the TU106 GPU, a 12nm chip with 1920 CUDA cores, 48 ROPs, and 6GB of super-fast GDDR6 ram that makes the RTX 2060 punch way above its weight thanks to expanded memory bandwidth.

MSI has overclocked the RTX 2060 GAMING Z by 30MHz on the base GPU clock, and by 150MHz on the boost GPU clock, for a total of 1830MHz. The boost GPU clock of 1830MHz is the best of the custom GeForce RTX 2060 cards so far, but don't worry, I'll be overclocking the card to get even more performance out of it.

Gaming output at 1080p and 1440p, especially in games like Overwatch and PUBG, presents a huge opportunity for monitor owners at those resolutions with high refresh rates.

MSI has chosen TWIN FROZR 7 cooling technology with a chunkier triple-slot cooler to keep the TU106 and 6GB of GDDR6 cool. The RTX 2060 GAMING Z has 3 x DP and 1 x HDMI, but no VirtualLink USB-C connector. MSI is using the MOSFET cooling, the same as on other GAMING Z and TWIN FROZR cards, which is critical considering the power components' need to remain cold.

MSI has used a 5+2 step PCB arrangement and a single 8-pin PCIe power connector to power both the stock and highly overclocked versions of the motherboard.

In our research, Overwatch serves a dual purpose: first, it's a common esports game, and second, it shows us what we'll need to run it at 1440p for 144/165Hz displays, and at 4K on the recent 144Hz HDR G-Sync displays with the RTX 2060.

MSI's tweaks on the RTX 2060 GAMING Z, which has the fastest out-of-the-box GPU boost clocks of any custom RTX 2060 on the market, have enhanced the GeForce RTX 2060 experience.

Pros
  • Excellent aesthetics
  • GDDR6
  • Great thermals
Cons
  • Power limit is quite low
  • No Virtualink Connector

MSI Radeon RX 5700 X

Stream Processors 2304
Boost Clock Speed OC Mode: 1750 MHz Gaming Mode: 1725 MHz
Base Clock Speed 1610 MHz
Interface PCI Express 4.0 x16

We are looking at MSI's Radeon RX 5700 Gaming X revision, which is the non-XT edition. The XT model's younger brother has an improved appearance and reduced noise levels. Fabricated on a 7nm node and capable of dealing with NVIDIA's GeForce RTX 2060 and maybe even 2070, we'll see how it stacks up against the reference card.

With its GAMING X card, MSI delivers a much more refined AMD Radeon RX 5700. It's the coolest running card on our charts right now, and it's also the card that uses the least amount of fuel, with excellent results at 1080p and 1440p. It's not that tough to sneak in any 4K gaming.

On the MSI RX 5700 GAMING X, you'll get fantastic FPS in 1080p and 1440p titles, and you'll have no trouble playing iconic games like Apex Legends and Overwatch on a 21:9 display at 3440x1440.

The Gaming X is equipped with two 90mm Torx (v3.0) fans with a Zero Frozr configuration, which means the fans do not spin below 60 degrees Celsius.

The RX 5700 will be equipped with two 8-pin PCIe power connectors and three DisplayPort outputs, one HDMI port. MSI is very conservative with the clocks, only allowing for a Turbo frequency of 1750 MHz. MSI has prioritized a power-efficient and quiet approach over squeezing an additional 1% to 2% perf out of the card at the expense of increased wattage.

Between the flagship RX 5700 XT GAMING X and the RX 5700 GAMING X, MSI hasn't improved much in terms of cooling technology, with the same fantastic TWIN FROZR 7 cooler used on the latest RX 5700 GAMING X. This means we have ZERO FROZR technology in use here for quiet use while the card isn't being used for games, and it also looks great.

MSI overclocked the Navi 10 XL GPU within the RX 5700 GAMING X to 1750MHz boost, 1725MHz game clocks, and 1610MHz base clocks so that it wouldn't be a GAMING X card without it. The 8GB of GDDR6 runs at the same 14Gbps bandwidth as the rest of the device.

MSI's custom RX 5700 GAMING X measures 297 x 58 x 140mm, so check the device measurements before downloading it. The custom RX 5700 GAMING X can give your mITX (or similar) gaming rig some powerful results if your mITX (or similar) gaming rig can handle it.

First and foremost, MSI equips the RX 5700 GAMING X with an incredibly dense and high-performing heat sink.

The RX 5700 GAMING X is kept cool by MSI's use of a dedicated aluminum layer that protects the card's power supply components. Overall this chip is a better option to go with

Pros
  • Very low temperatures
  • Reasonably priced
Cons
  • Some efficiency lost
  • Memory overclocking limited by adjustment range