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
Testing Notes
For the EPYC launch, AMD sent us their best SKU: the EPYC 7601. Meanwhile Intel gave us a choice between the top bin Xeon 8180 and the Xeon 8176. Considering that the latter had 165-173W TDP, similar to AMD's best EPYC, we felt that the Xeon 8176 was the best choice.
Unfortunately, our time testing the two platforms has been limited. In particular, we only received AMD's EPYC system last week, and the company did not put an embargo on the results. This means that we can release the data now, in time to compare it to the new Skylake-SP Xeons, however it also means that we've only had a handful of days to work with the platform before writing all of this up for today's embargo. We're confident in the data, but it means that we haven't had a chance to tease out the nuances of EPYC quite yet, and that will have to be something we get to in a future article.
Meanwhile we should note that we've had to retire the bulk of our historical benchmark data, as we upgraded both our compiler and OS (see below). Due to this, we only had a very limited amount of time to run additional systems, and for that reason we've opted include Intel's Xeon E5-2690. The Sandy Bridge-EP processor is about 5 years old, and for customers who aren't upgrading their servers every single generation, it's these servers that we believe are most likely to get upgraded in this round. So for server managers looking at finally buying into new hardware, you can get an idea of much return of investment you get.
Benchmark Configuration and Methodology
All of our testing was conducted on Ubuntu Server "Xenial" 16.04.2 LTS (Linux kernel 4.4.0 64 bit). The compiler that ships with this distribution is GCC 5.4.0.
You will notice that the DRAM capacity varies among our server configurations. The reason is that we had little time left before today's launch embargo. Removing any hardware is always a risk, so we decided to run our tests without significantly changing the internal hardware of the systems we received from AMD and Intel (SSDs were still replaced). As far as we know, all of our tests fit in 128 GB, so DRAM capacity should not have much influence on performance. But it wil have a impact on total energy consumption, which we will discuss.
Last but not least, we want to note how the performance graphs have been color-coded. Orange is AMD's EPYC, dark blue is Intel's best (Skylake-SP), and light blue is the previous generation Xeons (Xeon E5-v4) . Gray has been used for the soon-to-be-replaced Xeon v1.
Intel's Xeon "Purley" Server – S2P2SY3Q (2U Chassis)
| CPU | Two Intel Xeon Platinum 8176 (2.1 GHz, 28c, 38.5MB L3, 165W) |
| RAM | 384 GB (12x32 GB) Hynix DDR4-2666 |
| Internal Disks | SAMSUNG MZ7LM240 (bootdisk) Intel SSD3710 800 GB (data) |
| Motherboard | Intel S2600WF (Wolf Pass baseboard) |
| Chipset | Intel Wellsburg B0 |
| BIOS version | 9/02/2017 |
| PSU | 1100W PSU (80+ Platinum) |
The typical BIOS settings can be seen below; we enabled hyperthreading and Intel virtualization.
AMD EPYC 7601 – (2U Chassis)
Five years after our "Piledriver review", a new AMD server arrives in the Sizing Servers Lab.
| CPU | Two EPYC 7601 (2.2 GHz, 32c, 8x8MB L3, 180W) |
| RAM | 512 GB (16x32 GB) Samsung DDR4-2666 @2400 |
| Internal Disks | SAMSUNG MZ7LM240 (bootdisk) Intel SSD3710 800 GB (data) |
| Motherboard | AMD Speedway |
| BIOS version | To check. |
| PSU | 1100W PSU (80+ Platinum) |
Intel's Xeon E5 Server – S2600WT (2U Chassis)
| CPU | Two Intel Xeon processor E5-2699v4 (2.2 GHz, 22c, 55MB L3, 145W) Two Intel Xeon processor E5-2690v3 (2.3 GHz, 14c, 35MB L3, 120W) |
| RAM | 256 GB (16x16GB) Kingston DDR-2400 |
| Internal Disks | SAMSUNG MZ7LM240 (bootdisk) Intel SSD3700 800 GB (data) |
| Motherboard | Intel Server Board Wildcat Pass |
| BIOS version | 1/28/2016 |
| PSU | Delta Electronics 750W DPS-750XB A (80+ Platinum) |
The typical BIOS settings can be seen below.
HP-G8 (2U Chassis) - Xeon E5-2690
| CPU | Two Intel Xeon processor E5-2690 (2.9GHz, 8c, 20MB L3, 135W) |
| RAM | 512 GB (16x32GB) Samsung DDR-3 LR-DIMM 1866 MHz @ 1333 MHz |
| Internal Disks | SAMSUNG MZ7LM240 (bootdisk) Intel SSD3700 800 GB (data) |
| Motherboard | HP G8 |
| BIOS version | 9/23/2016 |
| PSU | HP 750W (Gold) |
Other Notes
Both servers are fed by a standard European 230V (16 Amps max.) power line. The room temperature is monitored and kept at 23°C by our Airwell CRACs.
Facebook
Twitter
RSS
219 Comments
View All Comments
oldlaptop - Thursday, July 13, 2017 - link
Why on earth is gcc -Ofast being used to mimic "real-world", non-"aggressively optimized"(!) conditions? This is in fact the *most* aggressive optimization setting available; it is very sensitive to the exact program being compiled at best, and generates bloated (low priority on code size) and/or buggy code at worst (possibly even harming performance if the generated code is so big as to harm cache coherency). Most real-world software will be built with -O2 or possibly -Os. I can't help but wonder why questions weren't asked when SPEC complained about this unwisely aggressive optimization setting...peevee - Thursday, July 13, 2017 - link
"added a second full-blown 512 bit AVX-512 unit. "Do you mean "added second 256 ALU, which in combination with the first one implements full 512-bit AVX-512 unit"?
peevee - Thursday, July 13, 2017 - link
"getting data from the right top node to the bottom left node – should demand around 13 cycles. And before you get too concerned with that number, keep in mind that it compares very favorably with any off die communication that has to happen between different dies in (AMD's) Multi Chip Module (MCM), with the Skylake-SP's latency being around one-tenth of EPYC's."1/10th? Asking data from L3 on the chip next to it will take 130 (or even 65 if they are talking about averages) cycles? Does not sound realistic, you can request data from RAM at similar latencies already.
AmericasCup - Friday, July 14, 2017 - link
'For enterprises with a small infrastructure crew and server hardware on premise, spending time on hardware tuning is not an option most of the time.'Conversely, our small crew shop has been tuning AMD (selected for scalar floating point operations performance) for years. The experience and familiarity makes switching less attractive.
Also, you did all this in one week for AMD and two weeks for Intel? Did you ever sleep? KUDOS!
JohanAnandtech - Friday, July 21, 2017 - link
Thanks for appreciating the effort. Luckily, I got some help from Ian on Tuesday. :-)AntonErtl - Friday, July 14, 2017 - link
According to http://www.anandtech.com/show/10158/the-intel-xeon... if you execute just one AVX256 instruction on one core, this slows down the clocks of all E5v4 cores on the same socket for at least 1ms. Somewhere I read that newer Xeons only slow down the core that executes the AVX256 instruction. I expect that it works the same way for AVX512, and yes, this means that if you don't have a load with a heavy proportion of SIMD instructions, you are better off with AVX128 or SSE. The AMD variant of having only 128-bit FPUs and no clock slowdown looks better balanced to me. It might not win Linpack benchmark competitions, but for that one uses GPUs anyway these days.wagoo - Sunday, July 16, 2017 - link
Typo on the CLOSING THOUGHTS page: "dual Silver Xeon solutions" (dual socket)Great read though, thanks! Can finally replace my dual socket shanghai opteron home server soon :)
Chaser - Sunday, July 16, 2017 - link
AMD's CPU future is looking very promising!bongey - Tuesday, July 18, 2017 - link
EPYC power consumption is just wrong. Somehow you are 50W over what everyone else is getting at idle. https://www.servethehome.com/amd-epyc-7601-dual-so...Nenad - Thursday, July 20, 2017 - link
Interesting SPECint2006 results:- Intel in their slide #9 claims that Intel 8160 is 2% faster than EPYC 7601
- Anandtech in article tests that EPYC 7601 is 42% faster than Intel 8176
Those two are quite different, even if we ignore that 8176 should be faster than 8160. In other words, those Intel test results look very suspicious.