The new Opteron 6300: Finally Tested!
by Johan De Gelas on February 20, 2013 12:03 AM ESTLS-DYNA Power Consumption
For HPC buyers, peak power tends to be a very important metric. As HPC systems are run at close to or equal to 100% CPU load, the energy consumption is at its peak for a long time. Peak power thus also determines the cooling and energy requirements. This is in sharp contrast with most other servers, where calculating the power and amps based on the peak load of a complete rack is considered wasteful as it is highly unlikely that all servers will hit 100% CPU load at the same time. We took the 95th percentile of our power numbers.
Note that the Xeon E5 numbers are not directly comparable to the Opteron numbers as the CPUs are tested in servers with different form factors. We will tackle that in the next test. Let us focus on the Opteron results for now.
AMD has made some real progres here. At the same clock, the total power consumption is 6% lower. Even at a 200MHz higher clock the peak power is very slightly—but consistently—lower (2%).
Of course, we also want to compare the AMD and Intel CPUs directly. To do this, we always run the fans at maximum speed. That way, the fans always consume the same amount of power. We then test with one and two CPUs, while keeping the amount of memory (64GB) the same. This way we measure how much extra power you consume at the wall when you add a second CPU. This number thus includes the voltage regulators (which can amount to up to 10% of the total server power) and the PSU inefficiency.
The Intel Xeon has a TDP of 95W, but even with a very FP intensive application it does not get anywhere near that number. About 75W out of those 94W are consumed by the CPU, as measured by our Hardware Monitoring Software that reads out the MSRs. We are still working on our version for the AMD platform (AMD's documentation is a bit late), but we estimate that the Opteron 6376 consumes about 110W and the Opteron 6380 needs about 120W. That means that AMD's top CPUs are probably consuming a bit more than their TDP indicates if you push the FP unit hard.
We also tried to measure idle power. Take the numbers with a grain of salt, but we measured about 19-20W for the Opteron 6380 (p-states disabled), 17-18W for the Opteron 6376 and 16-17W for the Xeon.
Facebook
Twitter
RSS
55 Comments
View All Comments
coder543 - Wednesday, February 20, 2013 - link
You realize that we have no trouble recognizing that you've posted about fifty comments that are essentially incompetent racism against AMD, right?AMD's processors aren't prefect, but neither are Intel's. And also, AMD, much to your dismay, never announced they were planning to get out of the x86 server market. They'll be joining the ARM server market, but not exclusively. I'm honestly just ready for x86 as a whole to be gone, completely and utterly. It's a horrible CPU architecture, but so much money has been poured into it that it has good performance for now.
Duwelon - Thursday, February 21, 2013 - link
x86 is fine, just fine.coder543 - Wednesday, February 20, 2013 - link
totes, ain't nobody got time for AMD. they is teh failzor.(yeah, that's what I heard when I read your highly misinformed argument.)
quiksilvr - Wednesday, February 20, 2013 - link
Obvious trolling aside, looking at the numbers and its pretty grim. Keep in mind that these are SERVER CPUs. Not only is Intel doing the job faster, its using less energy, and paying a mere $100-$300 more per CPU to cut off on average 20 watts is a no-brainer. These are expected to run 24 hours a day, 7 days a week with no stopping. That power adds up and if AMD has any chance to make any dent in the high end enterprise datacenters they need to push even more.Beenthere - Wednesday, February 20, 2013 - link
You must be kidding. TCO is what enterprise looks at and $100-$300 more per CPU in addition to the increased cost of Intel based hardware is precisely why AMD is recovering server market share.If you do the math you'll find that most servers get upgraded long before the difference in power consumption between an Intel and AMD CPU would pay for itself. The total wattage per CPU is not the actual wattage used under normal operations and AMD has as good or better power saving options in their FX based CPUs as Intel has in IB. The bottom line is those who write the checks are buying AMD again and that's what really counts, in spite of the trolling.
Rory Read has actually done a decent job so far even though it's not over and it has been painful, especially to see some talent and loyal AMD engineers and execs part ways with the company. This happens in most large company reorganizations and it's unfortunate but unavoidable. Those remaining at AMD seem up for the challenge and some of the fruits of their labor are starting to show with the Jaguar cores. When the Steamroller cores debut later this year, AMD will take another step forward in servers and desktops.
Cotita - Wednesday, February 20, 2013 - link
Most servers have a long life. You'll probably upgrade memory and storage, but CPU is rarely upgraded.Guspaz - Wednesday, February 20, 2013 - link
Let's assume $0.10 per kilowatt hour. A $100 price difference at 20W would take 1000 kWh, which would take 50,000 hours to produce. The price difference would pay for itself (at $100) in about 6 years.So yes, the power savings aren't really enough to justify the cost increase. The higher IPC on the Intel chips, however, might.
bsd228 - Wednesday, February 20, 2013 - link
You're only getting part of the equation here. That extra 20w of power consumed mostly turns into heat, which now must be cooled (requiring more power and more AC infrastructure). Each rack can have over 20 2U servers with two processors each, which means nearly an extra kilowatt per rack, and the corresponding extra heat.Also, power costs can vary considerably. I was at a company paying 16-17cents in Oakland, CA. 11 cents in Sacramento, but only 2 cents in Central Washington (hydropower).
JonnyDough - Wednesday, February 20, 2013 - link
+as many as I could give. Best post!Tams80 - Wednesday, February 20, 2013 - link
I wouldn't even ask the NYSE for the time day.