AMD Ryzen 7 1800X Linux Benchmarks – Paying for speed
Finally, we have the top model of the AMD Ryzen 7 launch series, the AMD Ryzen 7 1800X. Like the two lower end models, the Ryzen 7 1800X has 8 cores / 16 threads and 16MB L3 cache. Realistically it is a 200MHz speed bump from the Ryzen 1700X with a $100 price premium. AMD is targeting these CPUs at overclockers and claims that the 1800X is binned for higher overclocks. We did want to point to our guide on installing Linux kernel 4.10.1 for Ubuntu which is realistically the kernel you need to be using. Even if you need to use Ubuntu 14.04 or 16.04 LTS versions, if you are using Rzyen using older kernels will likely cause crashing. The AMD Ryzen 7 1700 is an extremely interesting part. It has the same 8 cores/ 16 threads as the higher-end CPUs. It can be overclocked with little effort. It also has a much lower TDP at 65W.
For our Ubuntu 14.04 with kernel 4.10.1 benchmarks we standardized on a very simple platform.
- CPU: AMD Ryzen 7 1800X
- Motherboard: ASUS Prime B350-Plus
- RAM: 16GB (4 x 16GB) DDR4-2666
- Boot Media: Intel DC S3710 400GB
- Cooling: Noctua NH-L9x65 SE-AM4
- Operating System: Ubuntu 14.04 LTS with Kernel 4.10.1
We expect most 1800X users will opt for a X370 motherboard. When you are running a server-oriented OS, there is very little reason to splurge on a $200 motherboard. The one item we wish the ASUS Prime B350-Plus had was an Intel 1GbE NIC to make OS installations easier.
We recommend using Ryzen only with Linux kernels 4.10 and later. We linked the Ubuntu upgrade guide above but if you are running Debian, RHEL, or CentOS 7 you will need to upgrade to a newer kernel ASAP. Here is the guide to stop the crashes in CentOS 7 by upgrading the kernel to 4.10.1 there.
AMD Ryzen 7 1800X Linux Benchmarks
For our testing we are using Linux-Bench scripts which help us see cross platform “least common denominator” results. We are using gcc due to its ubiquity as a default compiler. One can see details of each benchmark here. We are likely going to update the Linux-Bench in the near future with a few new tests as well as an even simpler to use/ faster revision, but for now, we are using the legacy version that now has over 100,000 test runs under its belt.
The item to remember here is that any benchmark we are publishing has had at least 10,000 profiling runs on a multitude of different architectures to ensure we get consistent results before we add it to our repertoire. Normally we go one step further and heat soak every machine at 100% load for 24 hours. On multi-node systems we even use a “sandwich” and controlled temperature and humidity to ensure systems are as close to real-world temperatures as possible. Due to the desktop nature of Ryzen chips, we are going to present our linux kernel 4.10.1 based results all with 24-hour heat soak but with no adjoining systems.
Python Linux 4.4.2 Kernel Compile Benchmark
This is one of the most requested benchmarks for STH over the past few years. The task was simple, we have a standard configuration file, the Linux 4.4.2 kernel from kernel.org, and make with every thread in the system. We are expressing results in terms of complies per hour to make the results easier to read.
We wanted to point out that there are a few differences between our Ryzen 7 1700X results we published on launch day and the additional results here. Namely, we are using a different kernel that has many of the necessary patches required to make everything run smoothly.
There are a few ways to look at this graph. First, the Ryzen 7 1800X is an absolute beast of a CPU. If you have a bunch of AWS c4 instances crunching numbers, the payback, including electricity, will be around two months with the AMD Ryzen 7 1800X machines instead.
c-ray 1.1 Performance
We have been using c-ray for our performance testing for years now. It is a ray tracing benchmark that is extremely popular to show differences in processors under multi-threaded workloads.
There are two ways to view this graph. First, the AMD Ryzen 7 systems do extremely well in this type of workload. Second, and perhaps more pertinent, is asking the question of whether the Ryzen 7 1800X is really worth almost a 75% premium over the Ryzen 7 1700 when you include the cooler. If you are not overclocking (a lot) then the Ryzen 7 1700 may be the better option as we surmised in AMD Ryzen 7 1700 Linux Benchmarks – The Zen you should buy.
7-zip Compression Performance
7-zip is a widely used compression/ decompression program that works cross platform. We started using the program during our early days with Windows testing. It is now part of Linux-Bench.
If you look near the bottom of that chart, you will find the 4 core Intel generations that also have ECC enabled and no 10GbE. Performance is not even close.
NAMD is a molecular modeling benchmark developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. More information on the benchmark can be found here.
With plenty of cores and clock speeds, the Ryzen 7 1800X manages solid NAMD performance.
Sysbench CPU test
Sysbench is another one of those widely used Linux benchmarks. We specifically are using the CPU test, not the OLTP test that we use for some storage testing.
At this point, the chart is self-explanatory. It is even besting the higher clock speed Intel chips.
OpenSSL is widely used to secure communications between servers. This is an important protocol in many server stacks. We first look at our sign tests:
And the verify results:
Again, we see the Ryzen 7 1800X offer performance in the same range as the AWS c4.4xlarge instance class on AWS. That is a $582.68/ month AWS instance against a $500 CPU in a $1000 system.
UnixBench Dhrystone 2 and Whetstone Benchmarks
One of our longest running tests is the venerable UnixBench 5.1.3 Dhrystone 2 and Whetstone results. They are certainly aging, however, we constantly get requests for them, and many angry notes when we leave them out. For example on our original AMD Ryzen 7 1700X review where UnixBench was crashing due to the kernel version we were using we left the results out and received many e-mails asking about them. UnixBench is widely used so it is a good comparison point.
Here are the whetstone results:
Intel performs relatively well here but we can see that the Ryzen 7 1800X puts some serious hurt down on the Haswell-E generation Xeon E5-1660 V3 which is an 8 core/ 16 thread 140w TDP offering.
If you are looking to get AMD Ryzen there are a few aspects to choosing the 1800X. First, the software ecosystem is highly optimized for Intel and we see this everywhere. CentOS broke, Ubuntu broke, VMware broke and there AMD SMT challenges on Microsoft Windows as well. With all of that said, we have the guides for Ubuntu and CentOS to get a system working well with Zen. The second aspect is competition within the AMD Ryzen 7 lineup itself. For $100 less you can get a Ryzen 7 1700X that is close in performance. For essentially $216 less on the $500 CPU, you can get the Ryzen 7 1700 with a cooler and a lower TDP. Our take at this point is that the top reason to buy an AMD Ryzen 7 1800X right now is if you simply want to tell people you have one. The software ecosystem desperately needs to mature to take full advantage of the chip. If it takes 18 months to do so, are we in the situation where you could be buying a second-generation Zen part with the money you saved by not getting the 1800X? For those building clusters, the Ryzen 7 1700 is going to be the better option. For those who are buying one system and want the best AMD has to offer, the AMD Ryzen 7 1800X is it.