AMD confirmed the official specifications for the upcoming Ryzen 5 CPUs last week, which will be available in exactly three weeks. By announcing these specifications, the company has largely left the cat out of its pocket.
We now know a few things that we had suspected for some time: Ryzen 5 is the same physical chip as Ryzen 7, so all models have two CCXs with four physical cores each, although not all are activated. In the six-core models, one core is deactivated from each CCX, while the quad-core parts deactivate two cores per CCX.
Since reviewing Ryzen 7, we have wanted to examine the downcore functionality of the BIOS of all AM4 motherboards. With this setting you can deactivate cores within the CCX modules or deactivate an entire CCX as a whole.
I was also interested in testing the Ryzen 7 eight-core CPUs with a range of GPUs. So far I've only compared game performance to the extreme Titan X Pascal. So I thought it was interesting not only to test with a high-end GPU, but also with something more realistic like the GTX 1070 and maybe even the GTX 1060.
Now that I know exactly how to configure Ryzen 5 CPUs and can mimic these settings, I've pulled a stack of GPUs out of memory and run tests.
Needless to say, we Not We actually still have a Ryzen 5 CPU. We just disable the cores to simulate the specs and performance for this series – or at least for the 1500X and 1600X, which have a full 16MB L3 cache. How close do we get? Pretty close, we think.
As an achievable overclocking for the flagship, I decided to lock the Ryzen 7 1800X at 4 GHz, but I was limited to 3.9 GHz on some 1700X chips and 3.8 GHz on one of my 1700 chips, so 4, 0 GHz is by no means guaranteed for all Ryzen 7 processors. To keep everything the same, I ran all of the Ryzen configurations in this article at 4 GHz.
At this frequency, I was able to simulate the overclocking performance of the Ryzen 5 1600X and the Ryzen 5 1500X, provided both can run all cores at 4 GHz.
For comparison, I also tested with three Kaby Lake CPUs (the dual-core i3-7350K with HyperThreading, the quad-core 7600K and the quad-core 7700K with HT) as they currently represent the best gaming performance. They also have no problems reaching 4.8 GHz, which could be seen as a slight overclocking.
I'm not going to focus on clock-to-clock comparison in this article as we know that Kaby Lake's IPC performance is slightly better – it doesn't need to be discussed again.
All CPUs, including the Ryzen models, were paired with DDR4-3000 memory and GPUs including the GTX 1080 Ti, GTX 1070 and GTX 1060. We have not used AMD graphics cards because none of them can reproduce the performance of the 1080 Ti and I & # 39 ;. d Keep the results consistent using all Nvidia cards rather than swapping the GTX 1060 for an RX 480 to showcase AMD's GPUs.
We have the results of more than 300 benchmark runs from six games with six CPU configurations and three GPUs …
Far Cry Primal
I wanted to include Far Cry Primal as this is a game that Ryzen is really having trouble with, and I also wanted to see how quickly we encounter a GPU bottleneck with less extreme GPUs.
Disabling cores within the CCXs does not improve performance here, although we never expected it to. After this article, I want to test with the second CCX disabled to see what difference it makes in games like Far Cry where Ryzen is having a hard time.
Since all cores are clocked at 4.0 GHz, the Ryzen CPUs are not yet comparable to the Core i3 in Far Cry Primal, and it is worth noting that we saw strong increases when changing from 7350K to 7600K and then back to 7700K .
When testing with the GTX 1070, the Kaby Lake processors all appear concentrated, since this is not a CPU-intensive title. The Ryzen CPUs are not too far back now, although they still lag behind even with a lower GPU.
With the GTX 1060 doing the rendering work, the Ryzen CPUs are pretty much on par with Kaby Lake. It wasn't long ago that this type of performance (which roughly corresponds to the GTX 980) was considered extremely high quality.
Far Cry Primal was developed in a way that just doesn't work well with the Zen architecture, and it's unlikely we'll see a patch for better support given the age of the game. I previously found that disabling SMT in Far Cry Primal is really helpful and increases performance by about 15%.
Hyper-threading can make a big difference in this title. Since I know how much threads love it, I made sure to include Mafia III results.
The Core i7-7700K was 41% faster than the 7600K when comparing the minimum frame rates, while the simulated Ryzen 5 1500X configuration matched the higher clocked 7600K, which was impressive.
Likewise, the 1800X is instantly a little faster than the 7700K in this title, although overclocking would allow Intel's chip to advance around 10 fps. The simulated 1600X was also faster than the 7600K, while the 1500X was a little slower on average, but a little faster at a minimum.
We see again that the Ryzen CPUs seem to perform better than the parts from Intel when shutting down on the GTX 1070. I assume this is due to the fact that they offer better minimum frame rates and these slower GPUs cannot quite reach the same maximums, which ultimately pulls the average down.
In combination with the GTX 1070, both the 1800X and the simulated 1600X are faster than the 7700K at 4.8 GHz, while the 1500X is roughly on the same level as the 7600K – an impressive result since it is clocked 17% lower.
With the installed GTX 1060, all CPU configurations offer almost the same performance, which of course is limited here by the GPU.