Apparatuses, methods, and systems directed to efficient cooling of data centers. Some embodiments of the invention allow encapsulation of cold rows through an enclosure and allow server fans to draw cold air from the cold row encapsulation structure to cool servers installed on the server racks. In other particular embodiments, the systems disclosed can be used to mix outside cool air into the cold row encapsulation structure to cool the servers. In some embodiments, the present invention involves using multiple cold row encapsulation structures to cool the servers installed on the racks.

Cold row encapsulation for server farm cooling system

Energy storage - in the form of UPS units - in a datacenter has been primarily used to fail-over to diesel generators upon power outages. There has been recent interest in using these Energy Storage Devices (ESDs) for demand-response (DR) to either shift peak demand away from high tariff periods, or to shave demand allowing aggressive under-provisioning of the power infrastructure. All such prior work has only considered a single/specific type of ESD (typically re-chargeable lead-acid batteries), and has only employed them at a single level of the power delivery network. Continuing technological advances have provided us a plethora of competitive ESD options ranging from ultra-capacitors, to different kinds of batteries, flywheels and even compressed air-based storage. These ESDs offer very different trade-offs between their power and energy costs, densities, lifetimes, and energy efficiency, among other factors, suggesting that employing hybrid combinations of these may allow more effective DR than with a single technology. Furthermore, ESDs can be placed at different, and possibly multiple, levels of the power delivery hierarchy with different associated trade-offs. To our knowledge, no prior work has studied the extensive design space involving multiple ESD technology provisioning and placement options. This paper intends to fill this critical void, by presenting a theoretical framework for capturing important characteristics of different ESD technologies, the trade-offs of placing them at different levels of the power hierarchy, and quantifying the resulting cost-benefit trade-offs as a function of workload properties.

http://www.cse.psu.edu/~buu1/papers/ps/sigmetrics12.pdf

Energy storage in datacenters: what, where, and how much?

Datacenters spend $10-25 per watt in provisioning their power infrastructure, regardless of the watts actually consumed. Since peak power needs arise rarely, provisioning power infrastructure for them can be expensive. One can, thus, aggressively under-provision infrastructure assuming that simultaneous peak draw across all equipment will happen rarely. The resulting non-zero probability of emergency events where power needs exceed provisioned capacity, however small, mandates graceful reaction mechanisms to cap the power draw instead of leaving it to disruptive circuit breakers/fuses. Existing strategies for power capping use temporal knobs local to a server that throttle the rate of execution (using power modes), and/or spatial knobs that redirect/migrate excess load to regions of the datacenter with more power headroom. We show these mechanisms to have performance degrading ramifications, and propose an entirely orthogonal solution that leverages existing UPS batteries to temporarily augment the utility supply during emergencies. We build an experimental prototype to demonstrate such power capping on a cluster of 8 servers, each with an individual battery, and implement several online heuristics in the context of different datacenter workloads to evaluate their effectiveness in handling power emergencies. We show that: (i) our battery-based solution can handle emergencies of short duration on its own, (ii) supplement existing reaction mechanisms to enhance their efficacy for longer emergencies, and (iii) battery even provide feasible options when other knobs do not suffice.

http://www.cse.psu.edu/~buu1/papers/ps/asplos12.pdf

Leveraging stored energy for handling power emergencies in aggressively provisioned datacenters

In recent years, a new breed of non-uniform memory access (NUMA) systems has emerged: multi-socket servers of multicores. This paper makes the case that data management systems need to employ designs that take into consideration the characteristics of modern NUMA hardware. To prove our point, we focus on a primitive that is used as the building block of numerous data management operations: data shuffling. We perform a comparison of different data shuffling algorithms and show that a na¨ove data shuffling algorithm can be up to 3× slower than the highest performing, NUMAaware one. To achieve the highest performance, we employ a combination of thread binding, NUMA-aware thread allocation, and relaxed global coordination among threads. The importance of such NUMA-aware algorithm designs will only increase, as future server systems are expected to feature ever larger numbers of sockets and increasingly complicated memory subsystems.

/pdf/numa-aware-algorithms-the-case-of-data-shuffling-1lhneyzki3.pdf

NUMA-aware algorithms: the case of data shuffling.

A data center system can include a mobile support structure; one or more enclosures for removable electronic equipment where the enclosures are housed by the support structure; a cooling system in fluid communication with the enclosures for cooling of the electronic equipment where the cooling system is housed by the support structure; and a power system operably connected to the electronic equipment and the cooling system for supplying power thereto where the power system comprises a generator housed by the support system. The mobile data center can be moved to remote locations, and the electronic equipment can be placed in communication with a network when at the remote location Other embodiments are disclosed.

System and method of providing computer resources

A datacenter cooling apparatus includes a portable housing having lifting and transporting structures for moving the apparatus, opposed sides in the housing, at least one of the opposed sides defining one or more air passage openings arranged to capture warmed air from rack-mounted electronics, opposed ends in the housing, at least one of the opposed ends defining one or more air passage openings positioned to allow lateral passage of captured air into and out of the housing, and one or more cooling coils associated with the housing to receive and cool the captured warm air, and provide the cooled air for circulation into a datacenter workspace.

Modular Data Center Cooling

To avoid the need to operate in-chassis fans to cool rack-mounted servers in a data center, the data center is arranged into a hot aisle and a cold aisle. The cold aisle is adjacent to a first side of the rack mounted servers and receives cold air from a cold air supply unit. The hot aisle is adjacent to a second side of the rack-mounted servers and has a lower pressure than the cold aisle. Because of the pressure difference between the cold aisle and the hot aisle, cold air flows through the rack-mounted servers, cooling electronic equipment therein, into the hot aisle. Control systems are used to obtain sufficient cooling.

Cooling servers in a data center using fans external to servers

The need for an uninterrupted power supply (UPS) in a data center is obviated by tying a DC voltage from a backup power source directly to the motherboards of multiple servers in the data center. AC power received from a power utility service is converted into a lower voltage by a site transformer and then provided to one or more power distribution units at a site. The power distribution units supply power to a plurality of servers, which include power supplies that convert the AC electrical power to DC electrical power for use by the servers' motherboards. In the event of a failure of power from the utility service, the backup power source provides DC electrical power to the motherboards, e.g., for sufficient time to start up a generator to provide power in place of the utility service.

Direct tie-in of a backup power source to motherboards in a server system

Large-scale data centers consume megawatts in power and cost hundreds of millions of dollars to equip. Reducing the energy and cost footprint of servers can therefore have substantial impact. Web, Grid, and cloud servers in particular can be hard to optimize, since they are expected to operate under a wide range of workloads. For our upcoming data center, we set out to significantly improve its power efficiency, cost, reliability, serviceability, and environmental footprint. To this end, we redesigned many dimensions of the data center and servers in conjunction. This paper focuses on our new server design, combining aspects of power, motherboard, thermal, and mechanical design. We calculate and confirm experimentally that our custom-designed servers can reduce power consumption across the entire load spectrum while at the same time lower acquisition and maintenance costs. Importantly, our design does not reduce the servers' performance or portability, which would otherwise limit its applicability.

/pdf/high-efficiency-server-design-406a0jprmd.pdf

High-efficiency server design

A server or other computing system contains a plurality of heat-generating electrical components and one or more cooling fans to dissipate heat from those components during operation. The fans are controlled by one or more fan controllers, which are programmed with an adaptive control algorithm to change the speed of each system cooling fan based on the needs of the components. The adaptive control algorithm is selected based on a detected configuration of the system's components, which may be detected upon startup of the system. During operation, a fan controller receives the component temperatures from one or more sensor and uses the algorithm to determine the fan speed. Thus, the system variably controls the speed of the fans based on the configuration of components housed in the server unit and on the current temperature of the components.

Amir Meir Michael

Papers

Modular Data Center Cooling

Cooling servers in a data center using fans external to servers

Direct tie-in of a backup power source to motherboards in a server system

High-efficiency server design

Adaptive fan control based on server configuration