Benchmarking Packet.com's Bare Metal Intel Xeon / AMD EPYC Cloud

Written by Michael Larabel in Computers on 15 November 2018 at 07:40 AM EST. Page 4 of 4. 2 Comments.
AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks

While four of the Xeon options outperformed the Packet EPYC option, on a performance-per-dollar basis for Tensorflow it came in right behind the c1.small instance for the best value.

AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks

The OpenMP-based Rodinia fluid dynamics solver saw the EPYC option offer almost twice as much performance-per-dollar as the next closest Xeon option.

AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks

The M-Queens performance was the best of the tested options on AMD EPYC.

AMD EPYC vs. Intel Xeon Packet Benchmarks
AMD EPYC vs. Intel Xeon Packet Benchmarks

The BRL-CAD open-source solid modeling performance also offered great value on the AMD EPYC server.

With this week being the first time I have tested Packet.com's cloud/server options, it remains quite fascinating the value offered for their bare metal access to these variety of Intel Xeon and AMD EPYC options for less than $2.00 USD per hour. With the earlier 16-way EPYC cloud benchmarks you can see the value offered by Packet with their $1 per hour option to tap a 24-core / 48-thread EPYC 7401P server. In this six-way Packet server comparison, the EPYC 7401P indeed offered the best value / performance-per-dollar overall. In only a subset of the workloads was the EPYC 7401P value unseated by the 40-cent-per-hour Xeon 4-core / 8-thread option albeit with much lower raw performance. To get at least a 24-core option from Packet with an Intel Xeon, it's at least $1.70 per hour.

Overall it was quite an interesting experience testing the Packet servers and hopefully this three-year-old cloud/server company will offer a greater selection of EPYC options. In my limited experience thus far testing Packet, the service was very reliable and did not run into any issues with the immediate deployment of these servers, no performance variation issues thanks to being bare metal access rather than just shared resources in a public cloud, and was interesting enough that I will likely benchmark any new offerings they rollout in the future.

If you enjoy these kinds of benchmarks on Phoronix, consider showing your support by ensuring you are not viewing this web-site with an ad-blocker or otherwise joining Phoronix Premium for ad-free access to the site that also provides multi-page articles on a single page, among other benefits, while allowing the resources to provide for continued Linux hardware benchmarking and expenses like this for public cloud/server benchmarks.

If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.


Related Articles
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.