[ad_1]
My gaming PC dates from 2019, more or less in March of that year. At the time I assembled what I felt was a fairly balanced mid-range platform for play in Full HD: Gigabyte B450 Aorus Pro motherboard, Ryzen 5 2600, GeForce RTX 2060 6GB and 16GB DDR4-3000 RAM (8GB x 2). Over the years, thanks to work and private life, I have done very little gaming (less than I would have liked) and therefore everything has remained unchanged up to the present day.
I played God of War, Marvel’s Spider-Man, DOOM Eternal and others still quietly with that platform, without too many worries, with the intention of carrying on with it for a while longer and thinking only of an improvement in the future or a total makeover. But often things don’t go as planned and, to please a loved one, I got rid of the RTX 2060: I had the possibility to get my hands on a 6800 XTand so I made the switch.
I was well aware that the balance of the platform would be off and that that 6800 XT would turn out braked by the rest of the platform, but at the same time I thought it would be a good opportunity to check how much potential performance I was leaving on the street. In a nutshell, a way to revive the advice to create balanced PCs but with some numbers and data in support. Thus was born this article.
I took the 6800 XT and put it on our video card test platform, where we analyze the latest proposals on the market: ASUS ROG X670E Hero, Ryzen 9 7950X and 32GB of DDR5-6000. I logged data with a dozen video games and then mounted the same card in my 2019 PC, running the same test suite. How many frames per second could I hit but am not getting because of my old PC? Let’s see it together in the tests.
Radeon RX 6800 XT, how much am I “neutering” it?
With Borderlands 3, we immediately see a 35fps difference in Full HD, which is reduced by going up in resolution, i.e. by shifting the load more to the GPU. As I said, for me, who play in Full HD, it’s the classic “Houston we have a problem“, with a 27.2% performance drop for the 6800 XT installed on my PC compared to the test platform. Or, if you want to see it in reverse, I could gain 37.5% performance by keeping the 6800 XT and switching to a latest generation configuration.
Cyberpunk 2077 presents me with an even heavier bill: The 6800 XT paired with my Ryzen 5 2600-powered system loses a whopping 42.3% compared to the performance observed on the test platform. Again, the gap narrows as you go up the resolution. In 4K, for example, I lose only 4-5%.
F1 2022 a game where platform mattersso taking the 6800 XT and transferring it to the test platform earns 101.7% in Full HD, 72.6% in 1440p and 12.1% in 4K.
Far Cry 6 evidently loves a high core count and very high frequencies, because the 6800 XT coupled to my PC’s Ryzen 5 2600 stops at 70 fps against 151 fps on the test platform: the improvement of 116%. In short, notwithstanding the differences between the different games based on the engine and optimization, there is no doubt that my PC is holding the 6800 XT back. The following tests also show this:
To conclude, taking a simple average, the Radeon RX 6800 XT improves its performance on the test platform compared to my PC by:
- 1080p: +74.4%
- 1440p: +39.6%
- 2160p: +8.8%
The not too large gap in 4K shows that the more load moves to the GPU, the less the rest matters. However, if you play at lower resolutions, even at 1440p, building unbalanced systems will cost you performance.
RTX 2060 on a 2019 and a current PC: does that change anything?
Not satisfied with this comparison, and to tell the truth a little disheartened given the results, I decided to check if my PC was indeed a balanced system. So I took an RTX 2060 – not the same one I had, but that’s not important – and tested it both on the test platform and on my PC. NOTE: the 4K result in Far Cry 6 is not there because the 6 GB of memory did not allow the test to run correctly.
As you see, the difference is rather limited, which means that the RTX 2060 was already expressing itself at its maximum potential or almost on the 2019 system. On the other hand the hardware differences are profound, a 6-core Zen+ CPU from the 2000 series and a 16-core Zen 4 from the 7000 series, not to mention memory, so if the RTX 2060 had anything more to give, it would have Done. Here’s how much fps the card gained on the test platform:
- 1080p: +6%
- 1440p: +3%
- 2160p: +2%
Because at the moment I don’t really think about changing the PC entirely, how could I improve system performance and do justice to this 6800 XT? Simple, switching CPUs. Here I have to say thanks to AMD, because socket AM4 stability and continuous BIOS updates for my B450 allow me to install all CPUs up to the Ryzen 5000 series.
Recover “lost” performance thanks to a Ryzen 7 5800X3D
Since this is a PC used only for gaming in my limited free time, the best upgrade I could do right now is that of upgrade to the Ryzen 7 5800X3D. And so, thanks to the availability of that CPU in the editorial office, I decided to install it on the system and see how much performance I could “recover”. Here are the test results compared to what I recorded with the Ryzen 5 2600:
Upgrading to a Ryzen 7 5800X3D would greatly benefit the performance of the 6800 XTespecially at Full HD resolution that I use, while going up the margin is reduced to almost zero. Game performance improvement in Full HD by almost 60% on average, while in 1440p it is close to 30%. In 4K, however, I would earn less than 4%.
What is the gap between my Ryzen 7 5800X3D-enhanced PC and the Ryzen 9 7950X test rig? If I decide to upgrade to a new state-of-the-art system I could gain the following gaming performance:
- 1080p: +11.5%
- 1440p: +8.9%
- 2160p: +5%
The best thing for me would be to buy the Ryzen 7 5800X3D and nothing else, I could go several years playing pretty well, even if I decide to upgrade to a 1440p monitor. Of course, I could have some problems by activating ray tracing, but upscaling technologies could come to my rescue. Switching to a new system created from scratch, also considering the prices in circulation, is out of the question.
Conclusion
This article born for put into numbers and give shape to a concept written many timesor that in the purchase of a component and in the creation of a PC it is necessary to proceed with a careful choice of the various parts, to end up with bottlenecks. Sure, the graphics switch did improve gaming performance, but it was less exciting than it might have been had I combined it with a CPU upgrade. In short, don’t repeat my mistake, if you can, or at least if you have to, at least be aware of it.
.
[ad_2]
Source link
