Mobile data storage device with power management

<p num="0001"> Overview The mobile data storage device may have at least a rotating data storage medium and controller according to various embodiments. The controller can be configured to change the rotational speed of the data storage medium in response to the expected changes in the command queue.</p>

0002<figref num="1">FIG. 3 is a block diagram of an exemplary data storage system configured and operating according to some embodiments.</figref><figref num="2A">A block diagram and drawn motion data from an exemplary data storage device are shown.</figref><figref num="2B">A block diagram and drawn motion data from an exemplary data storage device are shown.</figref><figref num="3">FIG. 3 is a block diagram of some of the exemplary mobile computing devices configured and operating according to various embodiments.</figref><figref num="4">An exemplary command queue profile routine is provided, according to some embodiments.</figref><figref num="5">Map exemplary command logic that is performed according to various embodiments.</figref><figref num="6">Illustrative power management schemes that can be utilized according to various embodiments are shown.</figref><figref num="7">It is a drawing of an exemplary power saving management routine performed according to various embodiments.</figref>

0003Detailed explanation The proliferation of semiconductor design and manufacturing has made computing devices more mobile and more powerful. Previously physically oversized, fixed, form factor technologies are now packaged in mobile electronics that provide high-bandwidth communication connections as well as large amounts of data generation such as high-resolution video recordings. Can be done. Reducing the form factor of mobile computing devices such as smartphones, tablet computers, watches and laptop computers has reduced the available space for batteries. Battery technology continues to advance and alternative power sources such as solar power can charge the battery, but replenishing the battery is slower for mobile computing devices than saving power usage. ,weak.

0004Therefore, the mobile data storage device may have at least a rotating data storage medium and a controller that changes the rotation speed of the data storage medium in response to a predicted change in the command queue. Aggressively changing the rotation speed of the medium can reduce unnecessary power consumption while enabling the function of the system to respond when the user accesses it. That is, the rotation speed of the medium can be changed to a power-saving state and returned to a fully operational state without significantly degrading the performance of the mobile data storage device.

0005Mobile data storage devices can be used in a variety of non-limiting systems and environments, but various embodiments include mobile data storage devices within an exemplary data storage system 100, shown as the block representation in FIG. Use. The data storage system 100 may employ power management techniques according to several embodiments by including one or more local or remote data storage devices 102. One or more data storage devices 102 may have one or more conversion heads 104 suspended on selected data bits 108 and data track 110 to program and sense data through the air bearing 112. Can be provided.

0006The storage medium 106 is fitted with one or more spindle motors 112 that rotate the medium 108 to generate an air bearing 114 on which the conversion head 104 is suspended to access a predetermined portion of the medium 106. be able to. In this way, one or more local processors 116 and remote hosts 118 provide controlled motion of the transform head 104 and spindle 112 to tune the transform head 104 and align it to the selected data bits 108. can do. The advent of network computing has allowed remote hosts 118 and storage arrays 120 to access controller 122 through network 124 via appropriate protocols.

0007The remote host 118 and the local processor 116 continuously or sporadically read the operating conditions of the data storage medium 106 such as vibration and temperature, as well as the operating conditions of the spindle 112 such as rotational speed and power consumption, one or more. Can act independently or simultaneously to monitor and control the sensor 126 of the. The local processor 116 and the remote host 118 are also as volatile and non-volatile memory cells to provide temporary storage of data and data information reserved for execution by the data storage device 102 and controller 122. Command requests in the configurable memory buffer 128 may be submitted, organized, and executed.

0008The continuous minimization of the physical dimensions of data storage components such as the transform head 104 enables implementation in mobile electronic devices that continuously demand smaller form factors and greater computing capabilities. Can be done. Figures 2A and 2B show different representations of the exemplary mobile computing system 140 that can be used in the data storage system of Figure 1, respectively. As shown in FIG. 2A, the mobile computing system 140 includes one or more mobile computing devices 142 capable of communicating data to static and virtual mobile devices over wired and wireless routes. It may be. For example, mobile computing device 142 is a network protocol that allows wireless connectivity to virtual cloud nodes, servers, and other mobile computing devices, as well as serial buses that allow wired connectivity to fixed desktops and servers. Can have.

0009Although not required or limited, the mobile computing device 142 can consist of a battery 144 that is powered and rechargeable or not. The cache memory 146 can be processed by the processor 148 and displayed graphically on the display 150 to provide short-term storage of data that may be moved to the hard disk drive 152 for long-term storage. The mobile computing device 142 can function without the means to cool the components, but the operation of the various mobile computing components individually heats through the consumption of the power supplied by the battery 144. And can work to generate collectively.

0010Regardless of the type, size, and performance of the data storage device within the mobile computing device 142, the heat generated from operation can jeopardize the performance of the mobile computing device 142. In other words, heat is generated from solid-state memory arrays and hard disk drives, and such heat can reduce the ability of these data storage means to accurately read, write, and output data. Illustrative relationships over time between the temperature of mobile computing device 146 and the power consumption of device 146 are provided by solid lines 154 and dashed lines 156. It can be understood that heat is retained within the mobile computing device, at least by the behavior of the components, the environmental conditions, and the interaction with the user.

0011FIG. 3 is a block diagram of a portion of an exemplary mobile computing device 160, arranged according to various embodiments. The mobile computing device 160 can have at least one data storage device 162 having one or more dedicated or distributed controllers 164 that provide a range of computing performance by executing drivers. Various embodiments use at least one driver to communicate with peripheral components to provide the functionality of data caching 166, fall protection 168, power management 170, and thermal management 172. These features can work exclusively to optimize the performance of the data storage device 162, and as a result, the mobile computing device 160, redundantly and collectively.

0012Various functions are available for any data storage device 162, which is part of the mobile computing device 164, with or without cooling means, but the controller 164 lacks cooling means. It is intended to selectively monitor the conditions of the data storage device 162 in order to utilize one or more functions to balance heat retention and power consumption in the mobile computing device. For example, the heat 172 and power 170 management schemes may operate simultaneously before data caching 166 is performed. As another non-limiting example, the fall protection 168 feature moves data to solid-state memory for temporary storage, performing predictive and reactive data caching 166, while power 170 and heat 172 management schemes. Can be stopped.

0013The ability to take advantage of a variety of different features can be retrofitted to data storage devices 162 that previously did not have a controller that could take advantage of such features. However, the mobile enablement kit can optionally be preloaded into the data storage device 152 during manufacturing and before end-user data is stored and various features can be enabled at any time. For intended use, a data storage device 162, such as a hard disk drive, can be used on a fixed computer with cooling means and then installed on a mobile computing tablet without cooling means to utilize dynamic data drivers. Through, the mobile enablement kit recognizes the lack of cooling and establishes a given function. Dynamic data drivers can be configured to establish communication between controller 164 and peripheral components to perform various functions. In this way, the mobile enablement kit can optimize the implementation of features with minimal need for additional software updates.

0014With or without a mobile enablement kit installed on the data storage device 162, the controller 164 is designed to establish and maintain optimal performance despite the device 162 consuming power and generating heat. In addition, continuous, sporadic, and routine measures can be taken positively. Figure 4 shows that at least one controller can actively maintain optimized mobile computing device performance while conditions such as changes in device temperature and power consumption change. It conveys an exemplary command queue profile routine 180, which may be performed according to various embodiments. Routine 180 can begin in step 182 by recording at least one command queue activity over time. Recording activity in this way can be done locally and remotely in temporary or permanent storage locations.

0015Decision 184 then evaluates one or more recorded command queue activities from step 182 to determine if a known activity profile exists. That is, decision 184 can evaluate the timing, status, and sequence of recorded activity to determine if a known activity profile is applicable. If the known profile does not match the recorded activity, step 186 initiates a new profile, which may be alone or later incorporated into another profile. If the known profile matches the recorded activity, step 188 updates the known profile with the recorded event, which may or may not be modified.

0016Registering recorded command queue activity in a new or known profile Step 190 allows step 190 to predict future command queue activity based on the activity observed in step 182. As a non-limiting example, one or more algorithms can identify trends and situations from activity profiles with a high probability of recurrence, which in prediction step 190, reducing the amount of command queue activity or Shown in the increase. The command queue may have a constant execution rate, and step 190 may be unexecuted or partially executed such as data read, servo data maintenance, metadata update, cache storage maintenance, and data write. Note that you can predict the amount of commands you have.

0017Advanced and simple algorithms can be adopted in step 190, but unexpected and previously unencountered activities can occur. Decision 192 determines whether the command queue activity predicted in step 190 is correct in an effort to validate, evolve, and maintain the accuracy of the activity profile as well as the algorithms used to predict future activity. Correct activity prediction triggers step 194 to record command queue activity and activity timing so that the profile and associated algorithms can then predict other future command queue events. If the predicted activity is wrong, decision 192 drives step 190 to ask to predict new activity and effectively remove the wrong prediction from what was included in the activity profile or prediction algorithm. ..

0018With the ability to predict future command queue activity, such as command capacity and urgency, the controller can take steps to save power on mobile computing devices. The command logic 200 of FIG. 5 illustrates how one or more controllers can respond to predicted and actual command queue activity according to various embodiments. It should be noted that the command logic 200 is not exclusive and may be partially executed during the execution of other control functions such as thermal management scheme 180, data caching 166, and fall protection 168. Returning to FIG. 5, the prediction of the actual command arrival in step 202 triggers the determination of whether the temperature exceeds the threshold in the evaluation and determination 204 of the temperature of the data storage device. The temperature above the limit causes the controller to slow down the rotation speed of the data storage medium while suspending the execution of at least one command queue in step 206. Reductions in medium speed and pauses in commands can dissipate heat quickly while reducing the amount of power consumed.

0019When the medium rotates at a reduced speed for a predetermined time, such as 30 seconds, decision 204 is revisited to determine if further reduction in rotation speed is needed. If so, the data storage device can eventually experience a step of rotational speed that can power off the data storage device. If the determination 204 determines that the temperature of the device is safe for executing the command, then step 208, and the derived heat that predicts how the execution of the command affects the temperature of the device. Calculate the command window based at least partially on the profile.

0020As a result of step 208, the time, power consumption, or temperature fluctuation window executes all commands received or contained in the command queue in step 202 as step 210 increases the rotation speed of the data storage device. Allowed to do or not, step 212 executes at least one instruction. Note that step 212 may execute the time delay imposed by the actual data access command as well as the thermal management scheme, but such execution is not essential. Completion of step 212 of the specified command window from step 208 advances logic 200 back to decision 204 where another assessment of the temperature of the data storage device is made. Through a cyclical return to decision 204, Logic 200 can continuously focus on what the temperature of the data storage device is, and data caching can serve user requests for a short period of time. Therefore, actions can be taken to reduce the temperature and power consumption of the device without compromising the user experience.

0021Logic 200 in Figure 5 is responsible for strategically adjusting the temperature of a data storage device through one or more intervals, such as a reduction of 500, 1000, or 5000 rpm, by reducing the number of revolutions of the data storage medium. Can bear. Data storage devices can also be powered off simply in the presence of high temperatures, but such behavior is because spinning up media requires more power than saving from spindown. Will be harmful to power consumption. That is, the logic 200 provides a balance between power consumption and heat dissipation by gradually spinning down the data storage medium as the temperature rises. During high-capacity command queue conditions such as operating system loads, Logic 200 can provide an optimized heat and power balance that maintains system performance. However, in low or sporadic command queue conditions that can accommodate the use of mobile computing devices, Logic 200 also dissipates heat and consumes power by keeping the device rotation and heat below a certain threshold. You don't have to balance.

0022FIG. 6 is an exemplary power management scheme that can be performed by a controller in a mobile data storage environment according to various embodiments to more aggressively save power, which can be particularly useful for low processing times. Represents 220. Although not required, scheme 220 is a system idle state, such as 1 minute, where the data storage device is powered off by stopping the conversion head and stopping the rotation of the data storage medium, regardless of the device temperature. Can be triggered according to a predetermined time of. In addition, scheme 220 can optionally be executed continuously, individually or in combination with other methods and logic.

0023The start of the power management scheme 220 begins at step 222 of receiving at least one read command from the host. Step 224 proceeds to check the local or network cache for the requested data. If the data is found in the cache, scheme 220 can return the data to the host without changing the state of the spinning data storage device. At step 224, if the retrieval of the requested data fails, the determination of whether the data storage device is rotating is achieved, making decision 226. The spinning data storage medium results in executing at least one command from the command queue in the next available command window in step 228, such as the command window established in step 206 of logic 200. If the medium is stationary, step 230 then holds the command in step 232 until the medium is rotated to a predetermined speed, such as 9000 rpm. Completion of step 232 allows step 228 to execute the command in the next available command window.

0024At the end of the command window where at least one command is executed in step 228, step 234 writes the local cache and shuts down the conversion head before the data storage device is powered off in step 236. Various embodiments follow a predetermined spin-down profile that allows efficient spin-up when new commands are received, while other embodiments use data storage devices for the purpose of maximizing power savings. It spins down suddenly and turns off the power. It is contemplated that the host can set and adjust the degree of power saving in the power management scheme 220, such as by setting high or low protection settings that correspond to different spindown profiles.

0025The use of the command queue profile routine 180 in FIG. 4, the command logic 200 in FIG. 5, and the power management scheme 220 in FIG. 6 can be used individually and simultaneously to balance temperature control and power consumption. In some embodiments, the controller of the data storage device intelligently executes various schemes and logics to adapt to the method in which the mobile computing device is used. FIG. 7 conveys an exemplary power saving routine 240 that can be implemented according to various embodiments that intelligently perform aggressive and responsive reduction of power consumption in mobile computing devices.

0026It is contemplated that any number of steps and decisions can be made prior to step 242, which predicts the activity of at least one command queue based on the derived profile. Routine 180, for example, is used in step 242 to predict the activity of one or more command queues, such as changing command capacity in a queue, increasing command execution, and establishing a backlog of commands in a queue. Can be done to start or update the profile to be done. Prediction of any kind of command queue activity or system activity, such as high or low throughput, can trigger step 244 to adjust the rotational speed of the data storage medium.

0027As a non-limiting example, step 242 has a low command queue capacity, or a slow queue command that slows down the rotation of the data storage medium at predetermined intervals, such as 500 revolutions or 25% of the operating speed before step 244. Uptake can be predicted. Prediction of command queue activity is made by step 244 slowing down the rotation speed of the data storage medium depending on the command queue that meets certain parameters, such as the number of new or executed commands over a given time period. , Can make it possible to actively save power. Step 244 can instead increase the rotational speed of the data storage medium so that an increased number of commands can be executed in the future, consuming as compared to executing commands gradually over a long period of time. Power can be reduced.

0028The derived profile can be used to improve the accuracy of command queue prediction in step 242, but rotation speed prediction and adjustment can be inaccurate. Decision 246 evaluates whether the positive action by the controller in step 244 is correct. Verification that one or more of the above rotational speed adjustments are correct turns routine 240, where command queue activity is predicted continuously, sporadically, and routinely, to step 242. Misprediction or adjustment of the rotation speed of the data storage medium may correspond to a predetermined time interval, such as 1, 5, 10, 30, 60 seconds, or more, which defines the timeout interval for failed command queue activity. it can.

0029That is, decision 248 can identify whether false command queue predictions and adjustments to the rotation speed of the data storage medium correspond to unnecessary delays in the execution or capacity of commands in the queue. False predictions and / or rotation adjustments do not result in command execution over the timeout interval. Step 250, when it is revisited in step 242, more good for the evolution of the profile that has been derived to predict the activity of not future command queue, to update the derived profile. The command queue delay over the timeout interval causes step 252 to reduce the rotation speed of the data storage medium by a predetermined interval that may be the same as or different from the interval in step 244.

0030An intended but non-essential embodiment of Routine 240 involves predicting slowness or inactivity in the command queue and stepping down the rotation speed of the data storage medium at uniform or different intervals up to low power rotations such as 500 rpm. .. Gradual slowdown grading is a predicted or actual change in user and command queue activity by having a mobile computing device bring the data storage medium to a state of operation such as 5600 or 9000 rpm. Allows you to respond quickly and seamlessly. The power consumed by the mobile computing device performing routine 240 can be optimized by turning off the spindle motor continuously or intermittently to reduce the rotational speed of the data storage medium.

0031The various aspects of Routine 240 are not essential or limiting, and any part of Routine 240 can be added, modified, and removed at will. For example, a step can be inserted to generate the derived profile or completely stop the rotation of the data storage medium together. Through the various embodiments described herein, mobile computing devices can predict command queue activity and reduce power consumption. The ability to selectively respond to command queue activity predicted by reducing the rotational speed of the data storage medium is complemented by a tapering of rotational speed that allows a seamless return to operating rotational speed.

0032It will be appreciated that the techniques described above can be readily utilized in any variety of applications, including computing environments with cooling means. Many features of the various embodiments of the present disclosure have been described in the above description, along with details of the structure and function of the various embodiments, but this detailed description is merely exemplary and in particular, the appended claims. It should be understood that changes may be made with respect to the structure and arrangement within the principles of the present art over the entire range indicated by the broad general meaning of the terms expressed by the scope of. For example, certain elements may be modified for a particular use without departing from the spirit and scope of the present disclosure.