Intel Developer Finds 50 Watt Power Regression In Linux

[Michael]

How are you going with getting PTS used/sponsored by companies like Intel for doing automated testing to find these regressions? Has there been much take up? It's sounding increasingly useful for finding these kinds of issues and I can understand not wanting to run high power drain systems continuously and out of your own pocket.

PTS is already used by multiple groups within Intel, NVIDIA, AMD, and countless other unnamed companies. I don't go to them seeking funds though for running a public test farm here but assisting them in their custom engineering and deployment needs of which they generally can't make results public.

[wizard69]

It's already fully-automated by PTS. The only thing missing me from routinely monitoring it is only having one WattsUp meter (~$100 per unit) and not running the very high-end systems constantly (but only lower-end Core i3/i5s or Atoms) due to energy costs.

The other thing here is the indication was that this was a forty core monster machine ( monster for this year) so we are talking a watt and a fraction for each core. That really isn't a lot so I wonder if I'm interpreting things right here.

The other thing is if you are going to benchmark power usage you really need to test a few Apple Mavericks based systems. Comparing or judging Linux against Windows is a bit like packing the panel of judges in your favor.

As for systems I'm not convinced a lot of people are interested in the high end systems. I'd rather see good benchmarking on portables and small form factor machines myself. That is machines that are very likely purchased for power savings in the first place.

[curaga]

The other thing here is the indication was that this was a forty core monster machine ( monster for this year) so we are talking a watt and a fraction for each core. That really isn't a lot so I wonder if I'm interpreting things right here.

The other thing is if you are going to benchmark power usage you really need to test a few Apple Mavericks based systems. Comparing or judging Linux against Windows is a bit like packing the panel of judges in your favor.

As for systems I'm not convinced a lot of people are interested in the high end systems. I'd rather see good benchmarking on portables and small form factor machines myself. That is machines that are very likely purchased for power savings in the first place.

In the lkml thread, the user with a Q6600 quad core reported a 25W difference. That's 6.25W per core.

[Serge]

The other thing here is the indication was that this was a forty core monster machine ( monster for this year) so we are talking a watt and a fraction for each core. That really isn't a lot so I wonder if I'm interpreting things right here.

The other thing is if you are going to benchmark power usage you really need to test a few Apple Mavericks based systems. Comparing or judging Linux against Windows is a bit like packing the panel of judges in your favor.

As for systems I'm not convinced a lot of people are interested in the high end systems. I'd rather see good benchmarking on portables and small form factor machines myself. That is machines that are very likely purchased for power savings in the first place.

I disagree completely. I think that the people who purchase and operate such "monster machines" are immensely more concerned about the power consumptions of these machines than the owners of portables and small form factor machines are of theirs.

[mrugiero]

I disagree completely. I think that the people who purchase and operate such "monster machines" are immensely more concerned about the power consumptions of these machines than the owners of portables and small form factor machines are of theirs.

Agree. Those people are usually using it in business and industries, and costs matter a lot. Power consumption is a cost, both in uptime (as it generates more heat) and in money.
However, consumers care about battery life, and it is tightly coupled to power consumption. Probably not as much as the ones with monster machines, but they do, and if the regression hits them (us, I guess), they should care too.

[wizard69]

I disagree completely. I think that the people who purchase and operate such "monster machines" are immensely more concerned about the power consumptions of these machines than the owners of portables and small form factor machines are of theirs.

Well that I have to disagree on. Nothing is driving the industry like getting as much performance, for the lowest power, in your hands as is possible. I do agree that power is important for large installations of machines, I just don't see that concern driving the industry the way portable devices are. Ask your self this, what sort of chips these days are getting the latest process technologies. Even Intel has let on that desktop and server chips are on the back burner, instead they have put very heavy emphasis on portable chips for the next year.

The other way to look at this is to follow the money and look at what sort of ARM chips are getting the most development dollars. Server chip products are in development but lets face it the intensity isn't the same as similar ARM chips destined for portable devices.

Id be the first to admit that the two customer bases here exhibit a massive gap in understanding the technology. Somebody building a data center or compute cluster will have a technical understanding about power usage that the common smart phone user doesn't have. All the smart phone user understands in that his phone is fast and that the battery lasts a long time. Power usage is only important once you realize that there are some pretty significant difference in how platforms behave when it comes to power usage.

[wizard69]

Agree. Those people are usually using it in business and industries, and costs matter a lot. Power consumption is a cost, both in uptime (as it generates more heat) and in money.
However, consumers care about battery life, and it is tightly coupled to power consumption. Probably not as much as the ones with monster machines, but they do, and if the regression hits them (us, I guess), they should care too.

Power consumption is extremely important to consumers even if they don't know the technical details. All they know is that the battery is dead and thus are frustrated.

In business it is more of a mixed bag of competing factors. Certainly the cost of operating the equipment is an issue in some environments, on the other hand many businesses will take the approach of buying the hardware to get a job done and let the chips fall where they may. The business reality is that you can only buy the technology available at the time of purchase.

I'm still of the opinion that you follow the money to find out what is important at the moment. Server chips and even desktop chips aren't the focus of the big manufactures anymore. There are a couple of manufactures trying to break into the ARM server market but you don't see billions of dollars flowing their way. In a nut shell the big server or compute machinery industry is a niche that is relatively small and frankly of diminishing interest right now. Intel and AMD may have their server hardware but lets face it they haven't show the interest like the have for chips targeted at portables.

[Serge]

Well that I have to disagree on. Nothing is driving the industry like getting as much performance, for the lowest power, in your hands as is possible. I do agree that power is important for large installations of machines, I just don't see that concern driving the industry the way portable devices are. Ask your self this, what sort of chips these days are getting the latest process technologies. Even Intel has let on that desktop and server chips are on the back burner, instead they have put very heavy emphasis on portable chips for the next year.

The other way to look at this is to follow the money and look at what sort of ARM chips are getting the most development dollars. Server chip products are in development but lets face it the intensity isn't the same as similar ARM chips destined for portable devices.

Id be the first to admit that the two customer bases here exhibit a massive gap in understanding the technology. Somebody building a data center or compute cluster will have a technical understanding about power usage that the common smart phone user doesn't have. All the smart phone user understands in that his phone is fast and that the battery lasts a long time. Power usage is only important once you realize that there are some pretty significant difference in how platforms behave when it comes to power usage.

Well, I don't think that looking at what kind of devices sell better is relevant to the question of what kind of computer users are more concerned about power usage. Two completely unrelated issues.

However, I think that debating "concern" isn't entirely relevant when discussing what kind of devices should be benchmarked for power usage. I do believe that those who make purchasing decisions on large machines such as servers, cluster machines, and so on are more likely to take an interest in power benchmarks than those who buy personal devices. I do not have any statistics to back that claim up, however.

[carewolf]

Finding these power regressions, isn't that what we have phoronix for?

[user82]

Well it seems to be fixed with this patch!