The Nvidia GeForce RTX 2080 is finally a real, tangible GPU and with six times the power of Pascal-based GTX GPUs it is a beast. But how does Nvidia's best graphic card ever come close to its eternal rival AMD, with its Radeon RX Vega 64?
Well, it will never be apples-to-apples comparisons when it comes to products from different companies. However, because they both try to generate incredible images for your favorite PC games, we put them on paper before they were taken to the lab later this year.
That said, from their design to their projected execution, to how much that performance you cost, let's see which graphics option is better for you: the Nvidia GeForce RTX 2080 versus the AMD Radeon RX Vega 64.
The designs of these two graphics cards are different: the AMD Radeon RX Vega 64 uses a traditional, air-based cooling mechanism with a single fan that draws in cool air and pushes out the rear vents while the RTX 2080 is used. a dual fan approach involving a heatsink in the process of spreading hot air. (Of course the idea here is for better temperature control, but we will not see that bear fruit until a complete assessment.)
AMD has a small lead over Nvidia in producing an official liquid-cooled version of the RX Vega 64, while Nvidia has yet to publicly discuss such an option for its newest card. We will see if Nvidia & # 39; s air-cooled cooling system with two fans bridges that hole.
The other important important factor of design, connectivity, is where the RTX 2080 goes forward with the ability to accommodate a USB-C port, in addition to the usual DisplayPort and HDMI ports. This is standard on the Founders Edition version of Nvidia, and is likely available for manufacturers to implement if they want to.
The RX Vega 64 does not have such an option, but also includes all the latest connectivity standards.
This distinction is important, because in the near future USB-C will become the de facto official connection for virtual reality (VR) hardware and applications. The RTX 2080 is ready for this next phase in simplifying VR, and the RX Vega 64 is not.
Of course, there are still benchmarks on the RTX 2080, because this card is fresh from the press. However, we can compare the two graphics cards on paper using their assessments for different basic performance statistics.
Before we start the RTX 2080, we first show what the RX Vega 64 can do first. This graphics card operates at a 1.247 MHz processor frequency, or clock speed, which is capable of up to 1,546 MHz with the help of basic tools.
The RX Vega 64 GPU itself contains 4,096 stream processors and contains 8 GB of RAM (High-Bandwidth Memory 2) (HBM2) that can handle up to 484.3 gigabytes of data per second or GB / sec, with a speed of 1.89 gigabits per second, or Gbps.
Now the RTX 2080 runs with a basic clock speed of 1.515MHz, only a hair that is shy of the increased speed of the RX Vega 64 and 1.710MHz when it is amplified. Although the CUDA cores of Nvidia and the stream processors of AMD are different, they generally have the same task: rendering pixels and performing other computer tasks.
For the 4,096 stream processors of the RX Vega 64, the RTX 2080 only has 2,944 CUDA cores. However, the CUDA cores are more versatile than the stream processors, which can perform a wider range of compute tasks, while stream processors are more specialized in efficiency.
Back to more meaningful comparisons, the RTX 2080 also has 8 GB of video memory, but with the new GDDR6 standard – a successor to GDDR5X instead of AMD & # 39; s HBM2. This memory actually has slightly less bandwidth than the RX Vega 64, capable of processing up to 448 GB / s, but with an exponentially higher speed of 14 Gbps.
The RTX 2080 does all this with 215 watt (W) power, while the RX Vega 64 requires 295W of the power supply of your system.
Ultimately, the RTX 2080, based on processor and memory speed alone, and power consumption, seems to be a clear winner here. However, this does not take into account differences in processor design and other factors, so only a full set of benchmarks will contain the true answer.
Now, for the ultimate deciding factor: how much these things cost.
AMD's current retail price for its RX Vega 64 is $ 499 (£ 549, about AU $ 630), but third-party manufacturers, such as Gigabyte and Asus, still sell the card more than that due to the increased demand from cryptocurrency miners .
Meanwhile, the RTX 2080 is considerably more expensive, with the Founders Edition costing $ 799 (AU $ 1,199, around £ 602). That is a much higher price than the GTX 1080 of yesteryear, only $ 599 (£ 600, AU $ 925) in comparison.
Of course you also have to take into account the implicit costs of these graphic cards, largely due to their power draw. The RTX 2080 requires less power than the RX Vega 64, which means you need a heavier power supply to support the latter – a price that is incurred if your system is not properly equipped for the job.
Which card is the better choice?
However expensive it may be, the RTX 2080 offers functions in games that are simply not possible with the RX Vega 64 – namely real-time ray tracing for more realistic lighting and shadows in games. That fact alone, not to mention the price, makes this versus a difficult one.
Of course, on paper the RTX 2080 wins in almost every respect. But that is not good for what you really need from a graphics card.
If you are looking for the absolutely most up-to-date features of your graphics card, the RTX 2080 – or even the slightly cheaper $ 599 (AU $ 899, around £ 451) RTX 2070 – is your ticket for the latest and greatest.
The RTX 2080 looks like it's burning through games with Ultra settings and a 4K resolution, let alone that they look better with ray tracing. However, if you just want to play games with a resolution of 1080p or even 1440p with the settings all the way up, the RX Vega 64 is by far a more cost-effective way.
So even in 2018 the classic dichotomy between Nvidia and AMD graphics cards will remain: splurge for Nvidia if you absolutely need the latest and greatest, but AMD's best will serve more mainstream gamers just as well for far less.