10 power-saving myths debunked

Is powering a server off and on bad for the system? We disprove this legend and others

Companies are finding themselves embroiled in a power crisis as they struggle to find ways to rein in soaring energy costs -- as well as do their part to address global climate change. However, how can you be certain that the power-saving strategies your company has adopted are, in fact, the best ones? After all, there are plenty of myths out there about saving energy that are patently false. In this report, we examine 10 such myths and bring the truth to light.

[ For more on how companies are battling soaring energy costs, please read "IT vs. the permanent energy crisis," as well as Ted Samson's Sustainable IT blog. ]

Myth No. 1: Powering a computer or server up and down limits its life span. The extreme temperature and current swings of power cycling can stress electronic components (especially capacitors and diodes) in a machine.
Fact: Power cycling healthy electronics is not a source of stress. The same electrical components that are used in IT equipment are used in complex devices that are routinely subjected to power cycles and temperature extremes, such as factory-floor automation, medical devices, and your car.

There is a kernel of truth in this myth, however: Cycling power on a sick system is going to bring attention to latent component weaknesses that go unnoticed in operation. Power-on diagnostics are brief yet rigorous and can be performed remotely on servers with dedicated management controllers. Power cycling doesn't just save energy. It's a zero-cost aid to maximizing server availability.

Myth No. 2: It takes too long to cold-start servers to react to spikes in demand. If customers are made to wait, they'll go elsewhere.
Fact: Idling servers at zero workload as hot spares is an egregious waste of energy and an administrative burden. If customers need to wait while you spin up cold spares to handle rising workload, brag about it. For a Web site, put up a static page asking users to wait while additional resources are brought online. As for the wait, people will stay on hold if they know their call will be answered. Build power management into your services architecture and make it part of the message that you send to users and customers.

[ For more on the high costs of powering servers, please read "Green IT numbers don't lie." ]

You can also select systems that cold-boot rapidly. Model to model and brand to brand, servers exhibit wide variances in power-up delay. This metric isn't usually measured, but it becomes relevant when you control power consumption by switching off system power. It needn't take long. Servers or blades that boot from a snapshot, a copy of RAM loaded from disk or a SAN can go from power-down mode to work-ready in less than a minute. The most efficient members of a reserve/disaster farm can quiesce in a suspend-to-RAM state rather than be powered down fully so that wake-up does not require BIOS self-test or device querying and cataloging, two major sources of boot delay.

Myth No. 3: The power rating (in watts) of a CPU is a simple measurement of the system's efficiency.
Fact: Efficiency is measured in percentage of power converted, which can range from 50 to 90 percent or more. The AC power not converted to DC is lost as heat, which increases the cooling burden of the system, adding even more to the overall energy loss. Unfortunately, it's often difficult to tell the efficiency of a power supply, and many manufacturers don't publish the number. You can either look for systems with published efficiency numbers or measure the actual power draw of various systems at idle and full load, then make your decisions based on that.

Myth No. 4: It's better to pack one big server with all the RAM, CPUs, and peripherals it can hold rather than to use multiple smaller servers.
Fact: This is only true if the big server is fully utilized, which can be dangerous with critical applications. Multiple smaller servers can be powered off or put in suspend mode when not in use, and they are safer from a redundancy point of view.

Also, populating a system with as many CPU cores and as much RAM as it will hold will result in a system that uses substantially more power than a base configuration of one dual-core CPU and a modest amount of RAM. Tailoring the server configuration to the software you'll be running can save energy without resorting to extreme measures.

Myth No. 5: LCD monitors use a trivial amount of power, so you might as well leave them on. Their colors and backlight brightness improve with warm-up time.
Fact: The average 17-inch LCD monitor consumes 35 watts of electricity. Adding together the hundreds of LCDs in an enterprise, the power used may not be that trivial. Energy Star LCD monitors will power down to sleep mode if the PCs' power management software is set up to tell them to. This saves energy and cash -- between $10 and $40 per year, according to Energy Star -- though not as much as simply turning the monitor off when it isn't in use. Even with the monitor turned off, an LCD's power supply will use between 1 and 3 watts of power. The only way to get it to zero is to unplug the power supply.

As for warm-up times, they are much shorter than they used to be: LCDs with LED backlighting rather than fluorescent don't need any warm-up time at all.

Myth No. 6: A notebook doesn't use any power when it's suspended or sleeping. USB devices charge from the notebook's AC adapter.
Fact: Sleep in Vista and Standby mode in XP save the state of the system to RAM and then maintains the RAM image even though the rest of the system is powered down. Hibernate saves the state of the system to hard disk, which reduces the boot time greatly and allows the system to be shut down. Sleep and Standby continue to draw a small amount of power, between 1 and 3 watts, even though the system appears to be inactive. By comparison, Hibernate draws less than 1 watt. Even over the course of a year, this difference is probably negligible.

[ For more on the benefits of putting PCs into low-power modes, please read "The ROI of PC power management." ]

Powering a laptop off doesn't necessarily reduce power usage to zero. This is easily confirmed by touching the power supply of a laptop that has been powered off for a while; it'll still be warm. Unless you unplug the power supply, it still burns energy.

Myth No. 7: Notebook batteries just wear out. There's not much you can do to make them last longer.
Fact: Many laptops with nickel-cadmium batteries come with a battery-reconditioning utility that drains the battery fully, then brings it back to a full charge. Laptops with lithium-ion batteries aren't afflicted with the same memory problem as those powered by NiCad batteries. However, unlike NiCad batteries, lithium batteries prefer to be only partially discharged: Running them all the way down will shorten their life span. The calibration utility for lithium batteries actually just recalibrates the capacity measurement to reflect the loss of capacity over time; it doesn't affect actual battery life. Battery life for either type of battery can be prolonged greatly by removing the battery when the unit is plugged into AC power. This approach is recommended if your laptop supports it and power outages are infrequent in your area.

Myth No. 8: Flash SSDs (solid-state drives) reduce the amount of power consumed by a laptop.
Fact: You may or may not experience a reduction in power consumption if your system is equipped with an SSD. It will vary greatly depending on the application. Typical office applications that don't constantly access the hard drive will show very little additional battery life with an SSD installed. Software that streams data from the drive constantly, such as video applications, will show greatly increased battery life. Other power savers such as LED backlighting can save more energy in typical applications.

Myth No. 9: Going to DC power will inevitably save energy.
Fact: Going to DC power entails removing the power supplies from a rack of servers or all the servers in a datacenter and consolidating the AC-DC power supply into a single unit for all the systems. Doing this may not actually be more efficient since you lose a lot of power over the even relatively small distances between the consolidated unit and the machines. New servers have 95 percent efficient power supplies, so any power savings you might have gotten by going DC is lost in the transmission process. Your savings will really depend on the relative efficiency of the power supplies in the servers you're buying as well as the one in the consolidated unit.

Myth No. 10: You're bound to save money by rushing out and buying the most energy-efficient equipment as soon as possible.
Fact: Savings realized by more efficient equipment have to be balanced against the cost of running the existing equipment. Replacing $5,000 servers before their end of life to save $30 per year in power is not going to save money.

Instead, look for ways to implement energy-saving strategies that don't require new equipment or user buy-in. For example, applying a policy through Active Directory to shut down systems that aren't in use after business hours doesn't require buying new equipment and will save a lot of money. If you can get user buy-in, other actions such as powering off monitors, PCs, printers, and the like will save lots of power without buying anything.

Mario Apicella, Brian Chee, and Tom Yager contributed to this article.

Join the discussion
Be the first to comment on this article. Our Commenting Policies