With the launch of 3000 series cards from NVIDIA, I have been asked a question which is better Intel or AMD. The answer is… yes. In this article we will compare the two and see which one comes on top. If you are new to this channel, please consider subscribing as we have a lot of content coming your way.
If you have any specific questions regarding the cards we are testing today, we already have done full reviews on them (RTX 3080 and RTX 3090). Now jumping into equipment used for the tests. We are using three different cards: RTX 3090 GamingPRO OC from Palit, RTX3080 TUF Gaming OC from ASUS and RTX2080 Super Strix also from ASUS.
For the Intel Bench – we used i7 10700k running on ASUS Z490 Maximus II Extreme motherboard.
On the AMD side we have Ryzen 7 3800XT running on ASUS ROG STRIX B550-E GAMING motherboard.
The rest of the components are the same.
Both of these CPUs are running at stock with performance enhancement turned off in bios to keep things fair. The reason we have chosen these chips is because they are the only ones that match core for core when it comes to high end while both having high boost speeds.
In Horizon Zero Dawn at 1080p we are seeing a great scaling from 2080 super to RTX3090, but when we get to the top of the chart it is Intel that is ahead by 10% on average FPS and 5 to 10% higher on the 1 percentiles.
Moving on to 1440p, the difference at RTX3080 level is only a few percent on average frame rates and about 10% at one percentiles. When we look at the RTX3090 the difference is still 10% on both average and 1 percentiles.
Moving on to 4k and AMD has caught up at the RTX3080 level and is actually 7% faster on both average and 1 percentiles. However, at the high end it’s about 7% slower than 10700k.
To those who have not read our RTX3090 review and noticing the 40% uplift in performance from 3080 to 3090 – this is mostly because of VRAM bottleneck. This game requires about 11 to 12GB of VRAM with maximum settings at 4k hence the massive jump in framerates.
In Shadow of the Tomb Raider at 1080p resolution Intel CPU is ahead with both RTX 3080 and 3090 by about 20% on the average scores and 23% on the 1 percentiles. It’s very obvious that Intel CPU is heavily bottlenecked when using the new cards. Surprisingly AMD chip is in fact bottlenecked even when using RTX2080 super, this is actually the trend with the rest of the games so we will leave out 1080p results moving forward.
At 1440p Intel is yet again ahead. The lead at RTX3080 level is about 6% on average FPS and 20% on 1 percentiles and at RTX3090 level the difference grows to 13% on average FPS and 18% on the 1 percentiles.
Moving on to 4k where we have a similar situation like in the previous game, at RTX 3080 level the 3800XT is 2% ahead on the average FPS and 6% lead on the 1 percentiles. When we look at the RTX3090 performance here we hit a hard wall on the graphics card and CPU speed no longer matters.
Next we have Total War Three Kingdoms and here at 1440p we see Intel and AMD are within a few percent of each other across all cards and ultimately heavily bottlenecked by the GPU.
In 4k the results are very similar yet again, just at this point there is a large jump between Intel and AMD on RTX3090 on 1 percentile frame rates. To be fair both of these results are not worth the trouble and would not have a smooth game play anyway.
Lastly we have Metro Exodus. First let’s start with ray tracing and DLSS turne d off. At 1440p resolution we have Intel in a very slight lead, but only when using RTX3090.
When we switch over to 4k, the results even out but yet again with this low FPS rate the game would be unplayable.
When we enable Ray tracing, the frame rates naturally drop but stay even between AMD and Intel in both 1440p and 4k resolutions.
When we enable DLSS, the frame rates increase but yet again at a steady rate.
The results here are very uninspiring and that is due to GPU doing all the heavy lifting, so if you are the kind of person who wants Ray tracing then realistically you would need to run it with DLSS and both AMD and Intel would be just fine. Additionally (at least in this title) you are probably better off aiming for RTX3080 and expect reasonable 1440p resolution.
With gaming out of the way, let’s check out the performance in productivity, starting with Blender BMW test. Interestingly enough we have completely even render times between both AMD and Intel.
Moving on to a longer Classroom test and here in all three scenarios the AMD chip is a whole one second faster, this must be a very happy day for all the hard core AMD fans!
Next we have a test that best targets all video creators which can then be applied to all software in relative terms: Davinci Resolve studio benchmark by Puget Systems and we yet again see AMD performing better than Intel. At RTX3080 level AMD has a 12% lead on overall score and up to 19% improvement on things like temporal noise reduction. When we look at RTX3090 level, AMD is still in the lead but now only by 7% with the highest score change being 4k ProRes 422 Codec at 17% improvement over Intel.
So what does all of this mean and which is better? Well, there is a reason why we used Intel i7 10700k with a mild overclock to 5.1 GHz when we did our RTX 3080 and 3090 reviews. And I mean it was mild as that chip can be pushed much harder with adequate cooling. On the other hand with AMD if you overclock you have to do it across all cores and then you lose boost which is very important in gaming. As it stands right now Intel will get you more FPS but will likely be slightly behind when it comes to productivity when at stock settings. If you overclock it then it will certainly even out.
AMD as a whole system can be slightly better value but you have a much smaller overclock potential and if you do both gaming and creative work I would advise against overclocking as it is more paramount to have a stable system when doing important client work.
If you are building a PC and can’t decide which to pick, I would strongly recommend to wait a few more weeks as AMD is launching their next generation of CPUs which will do two things – challenge Intel and reduce the price for the existing offerings. If you are not in a rush, then wait a bit longer until AMD releases their big Navi graphics cards which we will surely be covering as they come out.
Affiliate disclosure: as an Amazon Associate, we may earn commissions from qualifying purchases from Amazon.