SHEPARD: Scheduling on heterogeneous platforms using application resource demands

Heterogeneous computing technologies, such as multi-core CPUs, GPUs and FPGAs can provide significant performance improvements. However, developing applications for these technologies often results in coupling applications to specific devices, typically through the use of proprietary tools. This paper presents SHEPARD, a compile time and run-time framework that decouples application development from the target platform and enables run-time allocation of tasks to heterogeneous computing devices. Through the use of special annotated functions, called managed tasks, SHEPARD approximates a task's performance on available devices, and coupled with the approximation of current device demand, decides which device can satisfy the task with the lowest overall execution time. Experiments using a task parallel application, based on an in-memory database, demonstrate the opportunity for automatic run-time task allocation to achieve speed-up over a static allocation to a single specific device. © 2014 IEEE.

Managed Acceleration for In-Memory Database Analytic Workloads

In-Memory Databases (IMDBs), such as SAP HANA, enable new levels of database performance by removing the disk bottleneck and by compressing data in memory. The consequence of this improved performance means that reports and analytic queries can now be processed on demand. Therefore, the goal is now to provide near real-time responses to compute and data intensive analytic queries. To facilitate this, much work has investigated the use of acceleration technologies within the database context. While current research into the application of these technologies has yielded positive results, they have tended to focus on single database tasks or on isolated single user requests. This paper uses SHEPARD, a framework for managing accelerated tasks across shared heterogeneous resources, to introduce acceleration into an IMDB. Results show how, using SHEPARD, multiple simultaneous user queries all receive speed-up by using a shared pool of accelerators. Results also show that offloading analytic tasks onto accelerators can have indirect benefits for other database workloads by reducing contention for CPU resources.

Rasch Analysis of the Daily Living Tasks Dependent on Vision (DLTV)

PURPOSE. To examine internal consistency, refine the response scale, and obtain a linear scoring system for the visual function instrument, the Daily Living Tasks Dependent on Vision (DLTV). METHODS. Data were available from 186 participants with a clinical diagnosis of AMD who completed the 22-item DLTV (DLTV-22) according to four-point ordinal response scale. An independent group of 386 participants with AMD were administered a reduced version of the DLTV with 11 items (DLTV-11), according to a five-point response scale. Rasch analysis was performed on both datasets and used to generate item statistics for measure order, response odds ratios per item and per person, and infit and outfit mean square statistics. The Rasch output from the DLTV-22 was examined to identify redundant items and for factorial validity and person item measure separation reliabilities. RESULTS. The average rating for the DLTV-22 changed monotonically with the magnitude of the latent person trait. The expected versus observed average measures were extremely close, with step calibrations evenly separated for the four-point ordinal scale. In the case of the DLTV-11, step calibrations were not as evenly separated, suggesting that the five-point scale should be reduced to either a four- or three-point scale. Five items in the DLTV-22 were removed, and all 17 remaining items had good infit and outfit mean squares. PCA with residuals from Rasch analysis identified two domains containing 7 and 10 items each. The domains had high person separation reliabilities (0.86 and 0.77 for domains 1 and 2, respectively) and item measure reliabilities (0.99 and 0.98 for domains 1 and 2, respectively). CONCLUSIONS. With the improved internal consistency, establishment of the accuracy and precision of the rating scale for the DLTV and the establishment of a valid domain structure we believe that it constitutes a useful instrument for assessing visual function in older adults with age-related macular degeneration.

Training-induced changes in the pattern of triceps to biceps activation during reaching tasks after chronic and severe stroke

This exploratory study was undertaken to investigate the mechanisms that contributed to improvements in upper limb function following a novel training program. Surface electromyography (EMG) was used to examine training-induced changes in the pattern of triceps and biceps activation during reaching tasks in stroke survivors with severe paresis in the chronic stage of recovery. The EMG data were obtained in the context of a single blind randomised clinical trial conducted with 42 stroke survivors with minimal upper limb muscle activity and who were more than 6 months post-stroke. Of the 33 participants who completed the study, 10 received training of reaching using a non-robotic upper limb training device, the SMART Arm, with EMG triggered functional electrical stimulation (EMG-stim), 13 received training of reaching using the SMART Arm alone, and 10 received no intervention. Each intervention group engaged in 12 1-h training sessions over a 4-week period. Clinical and laboratory measures of upper limb function were administered prior to training (0 weeks), at completion (4 weeks) and 2 months (12 weeks) after training. The primary outcome measure was 'upper arm function' which is Item 6 of the Motor Assessment Scale (MAS). Laboratory measures consisted of two multijoint reaching tasks to assess 'maximum isometric force' and 'maximum distance reached'. Surface EMG was used to monitor triceps brachii and biceps brachii during the two reaching tasks. To provide a comparison with normal values, seven healthy adults were tested on one of the reaching tasks according to the same procedure. Study findings demonstrated a statistically significant improvement in upper limb function for stroke participants in the two training groups compared to those who received no training however no difference was found between the two training groups. For the reaching tasks, all stroke participants, when compared to normal healthy adults, exhibited lower triceps and biceps activation and a lower ratio of triceps to biceps activation. Following training, stroke participants demonstrated increased triceps activation and an increased ratio of triceps to biceps activation for the task that was trained. Better performance was associated with greater triceps activation and a higher ratio of triceps to biceps activation. The findings suggest that increased activation of triceps as an agonist and an improved coordination between triceps and biceps could have mediated the observed changes in arm function. The changes in EMG activity were small relative to the changes in arm function indicating that factors, such as the contribution of other muscles of reaching, may also be implicated.

Simultaneous remote haptic collaboration for assembling tasks

Stand-alone virtual environments (VEs) using haptic devices have proved useful for assembly/disassembly simulation of mechanical components. Nowadays, collaborative haptic virtual environments (CHVEs) are also emerging. A new peer-to-peer collaborative haptic assembly simulator (CHAS) has been developed whereby two users can simultaneously carry out assembly tasks using haptic devices. Two major challenges have been addressed: virtual scene synchronization (consistency) and the provision of a reliable and effective haptic feedback. A consistency-maintenance scheme has been designed to solve the challenge of achieving consistency. Results show that consistency is guaranteed. Furthermore, a force-smoothing algorithm has been developed which is shown to improve the quality of force feedback under adverse network conditions. A range of laboratory experiments and several real trials between Labein (Spain) and Queen’s University Belfast (Northern Ireland) have verified that CHAS can provide an adequate haptic interaction when both users perform remote assemblies (assembly of one user’s object with an object grasped by the other user). Moreover, when collisions between grasped objects occur (dependent collisions), the haptic feedback usually provides satisfactory haptic perception. Based on a qualitative study, it is shown that the haptic feedback obtained during remote assemblies with dependent collisions can continue to improve the sense of co-presence between users with regard to only visual feedback.