$400 AMD Radeon RX Vega 56 Beats $600 NVIDIA RTX 2070 Using Registry Mod

    0
    2309

    Despite manufacturers’ best attempts, enthusiasts and overclockers have always found a way to go past the factory set power and voltage levels to achieve unprecedented performance figures. Although, there are risks involved in subjecting your chips to these levels of…stress, the gains are usually worth it.

    Before we begin, please note that these mods will reduce the life of your hardware, but GPUs are often made to last for at least 4-5 years, and most of us get rid of them after just 2-3. So unless you plan to preserve your old hardware for your grandkids, these mods are quite safe. Still it’s better to go through guides and other users’ experience before attempting, just to be on the safer side.

    AMD Radeon RX Vega 56

    This testing was originally done by Gamersnexus, along with Buildzoid. They used a PowerColor RX Vega 56 Red Dragon and with the help of registry entries were successful in increasing the power limit to +242%. They compared the massively overclocked Vega 56 to a stock NVIDIA RTX 2070.

    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070: Benchmarks

    Sniper Elite 4 (DX12/4K)

    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070

    Initially, the RX Vega 56 is slower than both the GTX 1070 Ti as well as the the stock RTX 2070. But once you push the power limit all the way up to 240%, it manages to beat both the GeForce cards as well as the Vega 64. It is still marginally slower than the GTX 1080, but considering the $100 price gap between them, I’d call that fair.

    GTA V (4K/1440p)

    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070
    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070

    Rockstar’s GTA V is known to favor NVIDIA hardware and that doesn’t change in this test either. Before the overclock, the Vega 56 lags behind the GTX 980 Ti, but once the mod kicks in, the card goes all the way from 69 FPS to 79 FPS (at 1440p). However it fails to beat the RTX 2070 or even the GTX 1070 Ti. Not quite as effective as in the previous test, but I suppose it’s good enough.

    Far Cry 5 (4K/1440p)

    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070
    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070

    Far Cry’s Dunia engine is relatively vendor neutral, and consequently we see favorable results in this benchmark. Prior to the overclock, the RX Vega 56 roughly levels with NVIDIA’s GTX 1070 Ti, netting 41 and 74 FPS at 4K and 1440, respectively.

    With the registry hack in place, the tables turn in team red’s favor with the  Vega 56 exchanging blows with the GTX 1080, yielding 49 and 86 FPS at 4K and 1440p, respectively. This is a massive gain of approximately 20%, akin to what the Maxwell cards with BIOS mods would achieve.

    Shadow of  the Tomb Raider (DX12/4K/1440)

    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070
    AMD Radeon RX Vega 56 Beats NVIDIA RTX 2070

    Shadow of the Tomb Raider doesn’t really play well with the Radeon RX Vega cards and the results are less exciting. Although, the Vega 56 does get a decent uplift with the overclock, it isn’t enough to match NVIDIA’s fleet. Whether that is because this is an NVIDIA title or the game is shader bound is hard to say. At 4K, the Vega 56 fails to beat even its elder sibling, the Vega 64 and as for the NVIDIA cards, they are just too damn fast in this game.

    Conclusion: Not Bad At All

    While this is not something I’d recommend to everyone, it is worth a try if you have an old Vega 56 card that is giving you trouble in newer games. This registry mod manages to yield a good 15-20% performance boost, propelling the RX Vega 56 ahead of all its rivals, including the RX Vega 64, the GTX 1070s as well as the RTX 2070 in some titles.

    All in all, a decent bargain and it should allow you to squeeze out another year out of your old Vega 56, and if you’re lucky even more. Then again, if you were lucky enough (or rich), you’d have an RTX card or simply a faster GTX 10-series card.

    Further reading:

    *Benchmark credits: GamersNexus

    Leave a Reply