After covering the wave of GeForce Super GPUs at the end of last year and the subsequent popular Radeon Navi-based graphics cards, we revisited several GPUs to see how previous generation products compete at their new reduced prices.
Today we're looking at AMD's Radeon RX Vega 56, originally a $ 400 product that was scheduled to compete against the GTX 1070 in mid-2017. If you are wondering why this is relevant, we would say that the main reason is that same day you can buy a Vega 56 graphics card at a discount. They currently sell in the U.S. for just $ 250. In the same stadium, you'll find the GTX 1660 Super for around $ 230 and AMD's own 5600 XT for $ 280.
However, you should know that the cheapest Vega 56 cards are not good on sale. The $ 250 MSI Air Boost is the cheapest we could find, and it's a hot, noisy fan model. There's also the Gigabyte Gaming OC for $ 280, but it's worse, the failure rate on this model is worryingly high. Good models like the PowerColor Red Dragon or the Sapphire Pulse are sold at much higher prices, which makes them a depreciation.
For testing, we use the Asus Strix model, which we have upgraded with the right VRM thermal pads so that it does not throttle thermally. This way we show you the best out-of-the-box performance of a Vega 56 card, but we will not underestimate or optimize the card in any way. We don't like to test these cards overclocked or under voltage, which is essentially the same, because while some cards have a GPU pin, others have duds and unless you're ready to buy maybe half a dozen ahead of you, win the silicon – Lottery, test with the GPU under tension and behave as it is the norm can be very misleading.
All tests were carried out in our GPU test bench, which was powered by a Core i9-9900K overclocked to 5 GHz with 16 GB DDR4-3400 memory. But before we turn to the fps graphs, let's take a look at power consumption.
Vega 56 was a power pig and Navi has come a long way to fix this. Here we see that the total system utilization with the 5600 XT has decreased by almost 40%. Obviously, this is an enormous energy saving. So it will be interesting to see how they compare in terms of performance.
First we have the Assassins Creed Odyssey and here the 5600 XT was 8% faster than Vega 56 at 1080p and 7% faster at 1440p. This is not a big difference, but given the efficiency of the 5600 XT, this is a big step forward for AMD.
Star Wars Jedi: Fallen Order is another example where the 5600 XT is up to 10% faster than the Vega 56, an impressive win given the different power consumption.
The results of the outside worlds are interesting because AMD traditionally has to struggle with its GCN architecture in games with the Unreal Engine. Here the 5600 XT offers big performance increases of almost 30% when you look at 1% low performance and 18% for the average frame rate, which is a big performance improvement over Vega 56.
When testing with Call of Duty Modern Warfare at 1080p, we again have a performance advantage of 9% in favor of the 5600 XT.
The additional 2 GB of VRAM will help Vega 56 in Red Dead Redemption 2, but the performance is still very close as the 5600 XT only decreases with a margin of 7%.
The control performance is uniform, here the 5600 XT was less than 5% faster at both resolutions tested when comparing the average frame rates.
Performance in World War II is also competitive. We suspect Vega 56 will do a bit better with 1440p due to the additional VRAM, but overall we see a very similar gaming experience with both GPUs.
The results of F1 2019 are very surprising. In the past, Vega fared well in the F1 titles, although we tested with DX11. We expect Vega 56 to work very well with DX12, but at least that's not the case here compared to the 5600 XT. The 5600 XT was a whopping 26% faster at 1080p and 22% faster at 1440p, which means a significant increase for the newer navigation system.
We see an almost identical performance in Rainbow Six Siege with the new Vulkan API, although Vega 56 is only slightly less at 1440p and we again believe that this is due to the additional 2 GB VRAM.
In Fortnite, the 5600 XT can also hold its own against Vega 56 and delivers an average of 8% more frames at 1080p. The performance at 1440p is basically identical.
Next up we have Gears 5 and here the 5600 XT offers strong gains over the Vega 56 and boosts performance at 1080p by 17% and 13% at 1440p.
Shadow of the Tomb Raider's results are competitive. Here the 5600 XT is only 6% faster at 1080p and offers practically identical performance at 1440p.
A more typical 13% increase in performance is observed when testing with Apex Legends at 1080p. As we have seen many times, the margin is reduced at 1440p, this time to 7% in favor of the 5600 XT.
Recently we looked at PUBG, which is usually not a strong title for AMD. Navi makes a significant contribution to improving AMD's position. At both resolutions tested, the 5600 XT was 14% faster than the Vega 56.
Given the size of the Vega 56 chip and the fact that it contains more transistors, a lot more cores and HBM2 memory for a bandwidth increase of over 20%, it is remarkable what AMD has achieved with the much smaller Navi chip . Yes, part of the gains can be attributed to the 7 nm process, but there is also a large part of it that can be attributed to the RDNA architecture. Before we go too far into this conversation, let's look at how Vega 56 and 5600 XT compare head to head in the 32 titles tested.
We can see that the 1080p 5600 XT was 8% faster on average than the Vega 56. The only title where the Vega 56 had a significant performance advantage was Sniper Elite 4, a title optimized for the GCN architecture became what makes sense. Vega was also a bit faster in Red Dead Redemption 2, which we think is more of a VRAM issue. We also saw identical performances in Strange Brigade, World War II, Resident Evil 2, Rainbow Six Siege, Battlefield V, DiRT Rally and Control.
Many other titles favored Navi, including F1 2019, Ghost Recon Breakpoint, The Outer Worlds, Gears 5, The Witcher 3, PUBG, and a number of other titles.
If we switch to 1440p, we will see more of it, although the 5600 XT's lead will decrease a little. At this higher resolution, the newer GPS GPU was 6% faster and overall more consistent performance.
What we have learned
AMD has made a giant leap with the RDNA architecture and the 7 nm manufacturing process. The Radeon 5600 XT consumes about half the power of a Vega 56 GPU while offering a little more power. These results are consistent with AMD's own claims that RDNA performance can be increased by 50% compared to a GCN-based GPU with the same performance. They also claim that ~ 20-25% of this buoyancy comes from 7nm process shrinkage, with the bulk coming from their improvement in performance per cycle.
Given the massive efficiency improvement and the fact that it's ready to go, we think you're much better off with the newer Radeon GPU. In the future, AMD will spend more time optimizing game performance for RDNA. So expect the gap to widen over time.
- AMD Radeon RX 5600 XT on Amazon
- GeForce RTX 2070 Super on Amazon
- GeForce RTX 2060 Super on Amazon
- GeForce RTX 2060 at Amazon
- GeForce GTX 1660 Super on Amazon
- AMD Radeon RX 5700 XT on Amazon
- GeForce RTX 2080 Ti on Amazon