Before we move on, it’s worth pausing briefly at what we actually understand about game performance and how it makes sense to evaluate it. There are roughly three basic approaches:
- The test takes place on the most powerful graphics card and the game set to the lowest possible (eg 640 × 480, if it allows) resolution with the minimum possible details. This maximally eliminates the effects of the graphics card and the result is very close to the theoretical potential of the processor, shows a shift in architecture.
- The test runs at a relatively lower but practically extended resolution (often 1080p) in combination with the most powerful graphics card, often with medium details. The aim is to test the settings that are used in practice. The weakness is the simple fact that the combination of settings and hardware does not meet in practice (few people buy a GeForce RTX 3090 or Radeon RX 6900 XT to play in 1080p).
- The resolution and details are chosen so that the game runs in the usual playable FPS (eg 50-90 FPS). Different processors are then tested in a given configuration. This variant appears the least in reviews, but tells the most about practical behavior.
It also depends on whether the user selects the processor for an assembly that will be equipped with one graphics card for the entire period of its use (and the entire assembly will be replaced during the upgrade), or whether it will be a “chassis” for a replaceable graphics card. That is, a report where the basis should remain and where at least one upgrade of the graphics card is planned in the future.
From the point of view of the final assembly (ie the situation when the graphics card in the assembly goes “for life”), the choice of processor according to type 1 and 2 tests is meaningless, as it leads to complete oversizing of the processor for the needs of the assembly. If the user purchases a higher mainstream graphics card for 1440p, performance is more limited by the graphics card than in benchmark (a) built on the high end and (b) tested in 1080p. It makes sense to choose according to the type 3 test.
From the point of view of the assembly, where the graphics cards will be changed, it is good to have some reserve on the processor side. But even there it may not be essential.
The editors of the HardwareUnboxed website prepared tests for four generations of mainstream six-core Ryzen processors. They are all Ryzen 5 – 600X, but it may be worth noting that the last one – the Ryzen 5 5600X – is compared to the previous 65W model, while the previous ones were 95W.
We are the first to have results measured in 1080p. High detail on the left, lower details on the right. 1080p resolution can be considered relevant for the Radeon RX 5600 XT and partially the Radeon RX 5700 XT. It doesn’t matter which of the processors we use, even the Ryzen 5 1600X does not limit performance.
In 1440p, for which the results of Radeon RX 5700 XT and GeForce RTX 3070 are relevant. While in the case of the first card the performance limit is completely on its side, in the case of the second, which is more powerful, the processor is already starting to affect. Given the price of the graphics card, it would be worthwhile (when choosing between these four processors) to choose at least the Ryzen 5 2600X and, if a future upgrade plan is possible, the Ryzen 5 3600X.
4K resolution is the domain of GeForce RTX 3090 from the tested products. It is clear from the graph that even here the Ryzen 5 2600X would not slow down in any way, although in terms of minima it would be better to choose at least Ryzen 5 3600X. However, given the price of the graphics card (where the user pays almost twice as much as ~ 10% slower model), a surcharge would be paid for the Ryzen 5 5600X to make a certain margin. It could be used for a possible future upgrade to a future GPU up to ~ 50% faster than the tested GeForce RTX 3090. For multiple upgrades or upgrades to a ~ 2 × more powerful future product, an eight-core model would not be a thing of the past. For background, it is necessary to go to the previous review HardwareUnboxed, which offers a comparison of Ryzens with different numbers of cores:
For a hypothetical card 2x more powerful than the GeForce RTX 3090, 4k resolution rendering will be a comparable load as for the GeForce RTX 3090 “half-than-4k” resolution rendering, ie 1440p (4k resolution is about 8 megapixels, 1440p corresponds to less than 4 megapixels, ie half). There is already a minimal difference between six and more than six cores in current games. It’s not big, but it can grow slightly and we’re mainly in a segment where 10% higher graphics performance can cost $ 750 more. So in this context, it is not a mistake to pay $ 150 per processor for 2% of performance.
In any case, it can be said objectively that the vast majority of players do not need more than a six-core processor, and 90% in realistically set-up games do not know the difference between a three-year-old and a current model. The dogmatic approach “only a fool would buy something other than a high-end processor for high-end graphics” is unfounded, dogmatic in terms of real tests. It pays for the average player not to pay extra $ 200 on the processor, but to add it to the graphics card (if a more powerful one is available for the price of $ 200 higher). The practical impact will be more significant.
Of course, another situation may concern a player who, for example, is streaming, or a user who is running some other (processor) load on the set in addition to the games. However, a classic player does not have to worry too much about the processor, high-ened surcharges will definitely not pay off.
Source: Diit.cz by diit.cz.
*The article has been translated based on the content of Diit.cz by diit.cz. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!
*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.
*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!