Sizing Up Servers: Intel's Skylake-SP Xeon versus AMD's EPYC 7000 - The Server CPU Battle of the Decade?
by Johan De Gelas & Ian Cutress on July 11, 2017 12:15 PM EST- Posted in
- CPUs
- AMD
- Intel
- Xeon
- Enterprise
- Skylake
- Zen
- Naples
- Skylake-SP
- EPYC
Apache Spark 2.1 Benchmarking
Apache Spark is the poster child of Big Data processing. Speeding up Big Data applications is the top priority project at the university lab I work for (Sizing Servers Lab of the University College of West-Flanders), so we produced a benchmark that uses many of the Spark features and is based upon real world usage.
The test is described in the graph above. We first start with 300 GB of compressed data gathered from the CommonCrawl. These compressed files are a large amount of web archives. We decompress the data on the fly to avoid a long wait that is mostly storage related. We then extract the meaningful text data out of the archives by using the Java library "BoilerPipe". Using the Stanford CoreNLP Natural Language Processing Toolkit, we extract entities ("words that mean something") out of the text, and then count which URLs have the highest occurrence of these entities. The Alternating Least Square algorithm is then used to recommend which URLs are the most interesting for a certain subject.
In previous articles, we tested with Spark 1.5 in standalone mode (non-clustered). That worked out well enough, but we saw diminishing returns as core counts went up. In hindsight, just dumping 300 GB of compressed data in one JVM was not optimal for 30+ core systems. The high core counts of the Xeon 8176 and EPYC 7601 caused serious performance issues when we first continued to test this way. The 64 core EPYC 7601 performed like a 16-core Xeon, the Skylake-SP system with 56 cores was hardly better than a 24-core Xeon E5 v4.
So we decided to turn our newest servers into virtual clusters. Our first attempt is to run with 4 executors. Researcher Esli Heyvaert also upgraded our Spark benchmark so it could run on the latest and greatest version: Apache Spark 2.1.1.
Here are the results:
If you wonder who needs such server behemoths besides the people who virtualize a few dozen virtual machines, the answer is Big Data. Big Data crunching has an unsatisfiable hunger for – mostly integer – processing power. Even on our fastest machine, this test needs about 4 hours to finish. It is nothing less than a killer app.
Our Spark benchmark needs about 120 GB of RAM to run. The time spent on storage I/O is negligible. Data processing is very parallel, but the shuffle phases require a lot of memory interaction. The ALS phase does not scale well over many threads, but is less than 4% of the total testing time.
Given the higher clockspeed in lightly threaded and single threaded parts, the faster shuffle phase probably gives the Intel chip an edge of only about 5%.
Facebook
Twitter
RSS
219 Comments
View All Comments
psychobriggsy - Tuesday, July 11, 2017 - link
Indeed it is a ridiculous comment, and puts the earlier crying about the older Ubuntu and GCC into context - just an Intel Fanboy.In fact Intel's core architecture is older, and GCC has been tweaked a lot for it over the years - a slightly old GCC might not get the best out of Skylake, but it will get a lot. Zen is a new core, and GCC has only recently got optimisations for it.
EasyListening - Wednesday, July 12, 2017 - link
I thought he was joking, but I didn't find it funny. So dumb.... makes me sad.blublub - Tuesday, July 11, 2017 - link
I kinda miss Infinity Fabric on my Haswell CPU and it seems to only have on die - so why is that missing on Haswell wehen Ryzen is an exact copy?blublub - Tuesday, July 11, 2017 - link
Your actually sound similar to JuanRGA at SAKevin G - Wednesday, July 12, 2017 - link
@CajunArson The cache hierarchy is radically different between these designs as well as the port arrangement for dispatch. Scheduling on Ryzen is split between execution resources where as Intel favors a unified approach.bill.rookard - Tuesday, July 11, 2017 - link
Well, that is something that could be figured out if they (anandtech) had more time with the servers. Remember, they only had a week with the AMD system, and much like many of the games and such, optimizing is a matter of run test, measure, examine results, tweak settings, rinse and repeat. Considering one of the tests took 4 hours to run, having only a week to do this testing means much of the optimization is probably left out.They went with a 'generic' set of relative optimizations in the interest of time, and these are the (very interesting) results.
CoachAub - Wednesday, July 12, 2017 - link
Benchmarks just need to be run on as level as a field as possible. Intel has controlled the market so long, software leans their way. Who was optimizing for Opteron chips in 2016-17? ;)theeldest - Tuesday, July 11, 2017 - link
The compiler used isn't meant to be the the most optimized, but instead it's trying to be representative of actual customer workloads.Most customer applications in normal datacenters (not google, aws, azure, etc) are running binaries that are many years behind on optimizations.
So, yes, they can get better performance. But using those optimizations is not representative of the market they're trying to show numbers for.
CajunArson - Tuesday, July 11, 2017 - link
That might make a tiny bit of sense if most of the benchmarks run were real-world workloads and not C-Ray or POV-Ray.The most real-world benchmark in the whole setup was the database benchmark.
coder543 - Tuesday, July 11, 2017 - link
The one benchmark that favors Intel is the "most real-world"? Absolutely, I want AnandTech to do further testing, but your comments do not sound unbiased.