Current methods for continuous respiration monitoring such as respiratory inductive or optoelectronic plethysmography are limited to clinical research settings; most wearable systems reported only measures rate. Here we introduce a sensor capable of simultaneously measuring both rate and volume with high fidelity. Our disposable Band-Aid© like formfactor can measure by simply the local strain ribcage abdomen during breathing. We demonstrate that metrics highly correlated measurements from...

Wrinkled gold thin films on elastomeric substrates are used as robust parallel plate electrodes for soft capacitive pressure sensors. The wrinkled structures create a integration with the polymer, allowing repeated normal force to deform film without failure. By incorporating microridged that support counter air cavities within dielectric layer, sensitivity is further increased 0.148 kPa-1 over wide dynamic range of up 10 kPa. and sensor allow consistent measurements exerted by radial artery...

Carbon nanotubes (CNTs) on shape memory polymers result in densified, nano- to microscale wrinkles upon heat-induced shrinkage. These wrinkled CNT bundles can be transferred into soft materials similar the human epidermis for extremely highly stretchable skin mountable strain sensors with a dynamic range of over 700%. As service our authors and readers, this journal provides supporting information supplied by authors. Such are peer reviewed may re-organized online delivery, but not...

High performance parallel programs are currently difficult to write and debug. One major source of difficulty is protecting concurrent accesses shared data with an appropriate synchronization mechanism. Locks the most common mechanism but they have a number disadvantages, including possibly unnecessary serialization, possible deadlock. Transactional memory alternative that makes programming easier. With transactional memory, transaction provides atomic serializable operations on arbitrary...

Abstract Highly sensitive and flexible components are essential for applications in wearable electronics. Using low‐cost rapid prototyping methods, piezoresistive pressure sensors fabricated using shrink‐film, a shape memory polymer that retracts upon heat, to introduce wrinkling carbon nanotube thin films, which improves both elasticity sensitivity. The wrinkles not only provide strain relief, but it also sensitivity by 12 800 fold with response time of less than 20 ms. improved is due the...

This work presents a wrinkled Platinum (wPt) strain sensor with tunable sensitivity for applications in wearable health monitoring. These stretchable sensors show dynamic range of up to 185% and gauge factor (GF) 42. is believed be the highest reported GF any metal thin film over physiologically relevant date. Importantly, are application by adjusting wPt thickness. Performance reliable 1000 cycles low hysteresis after conditioning. The possibility using such real-time respiratory monitoring...

Instruction set customization is an effective way to improve processor performance. Critical portions of applicationdata-flow graphs are collapsed for accelerated execution on specialized hardware. Collapsing dataflow subgraphs will compress the latency along critical paths and reduces number intermediate results stored in register file. While custom instructions can be effective, time cost designing a new each application immense. To overcome this roadblock, paper proposes flexible...

Instruction set customization is an effective way to improve processor performance. Critical portions of application data-flow graphs are collapsed for accelerated execution on specialized hardware. Collapsing dataflow subgraphs will compress the latency along critical paths and reduces number intermediate results stored in register file. While custom instructions can be effective, time cost designing a new each immense. To overcome this roadblock, paper proposes flexible architectural...

With the growing prominence of wearable electronic technology, there is a need to improve mechanical reliability electronics for more demanding applications. Conductive wires represent vital component present in all electronics. Unlike traditional planar and rigid electronics, these new electrical components must conform curvilinear surfaces, stretch with body, remain unobtrusive low profile. In this paper, piezoresistive response shrink induced wrinkled gold thin films under strain...

The unique antiwetting properties of superhydrophobic (SH) surfaces prevent the adhesion water and bodily fluids, including blood, urine, saliva. While typical manufacturable approaches to create SH rely on chemical structural modifications, such are expensive, require postprocessing, often not biocompatible. By contrast, it is demonstrated that purely features easily formed using high throughput roll‐to‐roll (R2R) manufacturing by shrinking a prestressed thermoplastic with thin, stiff layer...

Clustered architectures are a solution to the bottleneck of centralized register files in superscalar and VLIW processors. The main challenge associated with clustered is compiler support effectively partition operations across available resources on each cluster. In this work, we present novel technique for clustering based graph partitioning methods. Our approach incorporates new methods assigning weights nodes edges within dataflow guide partitioner. Nodes assigned reflect their resource...

We demonstrate a novel bonding process that allows for facile multilayer microfluidic chip fabrication and is compatible with cell micropatterning.

Current approaches in targeted patient treatments often require the rapid isolation of specific rare target cells. Stream-based dielectrophoresis (DEP) based cell sorters have limitation that maximum number sortable types is equivalent to output channels, which makes upscaling a higher different technically challenging. Here, we present microfluidic platform for selective single-cell sorting bypasses this limitation. The consists 10 000 nanoliter wells are placed on top interdigitated...

Patients with relapsed or refractory diffuse large B-cell lymphoma (DLBCL) have limited treatment options, particularly if they are transplantation chimeric antigen receptor (CAR) T-cell ineligible, and novel therapeutics needed. An 86-year-old woman DLBCL received a novel, first-in-class small molecule inhibitor of N-myristoyltransferase (NMT) as the initial patient on phase I dose escalation trial. Daily oral administration 20 mg PCLX-001 tablets produced pharmacokinetic profile suitable...

Clustered architectures are a solution to the bottleneck of centralized register files in superscalar and VLIW processors. The main challenge associated with clustered is compiler support effectively partition operations across available resources on each cluster. In this work, we present novel technique for clustering based graph partitioning methods. Our approach incorporates new methods assigning weights nodes edges within dataflow guide partitioner. Nodes assigned reflect their resource...

Practical wearable applications of soft strain sensors require capable not only detecting subtle physiological signals, but also withstanding large scale deformation from body movement. Encapsulation is one technique to protect both environmental and mechanical stressors. We introduced an encapsulation layer crack-based wrinkled metallic thin film as avenue improve sensor stretchability, linear response, robustness. demonstrate that encapsulated have increased robustness stability,...

We present a rapid, simple, and scalable approach to achieve superhydrophobic (SH) substrates directly in commodity shrink wrap film utilizing Argon (Ar) plasma. Ar plasma treatment creates stiff skin layer on the surface of film. When shrinks, mismatch stiffness between bulk causes formation multiscale hierarchical wrinkles with nano-textured features. Scanning electron microscopy (SEM) images confirm presence these biomimetic structures. Contact angle (CA) contact hysteresis (CAH)...

Register bypass provides additional datapaths to eliminate data hazards in processor pipelines. The difficulty with register is that the cost of network substantial and grows substantially as width or pipeline depth are increased. For a single application, many paths have extremely low utilization. Thus, there an important opportunity design application-specific processors remove large fraction while maintaining performance comparable full bypass. We propose systematic customization process...

This paper presents the design and implementation of a compiler runtime infrastructure for automatic program distribution. We are building research that enables experimentation with various partitioning mapping strategies study distribution's effect on resource consumption (e.g., CPU, memory, communication). Since many optimization techniques faced conflicting targets memory communication), we believe it is important to be able their interaction. present set enable flexible modeling These...

Set-associative caches are traditionally managed using hardware-based lookup and replacement schemes that have high energy overheads. Ideally, the caching strategy should be tailored to application's memory needs, thus enabling optimal use of this on-chip storage maximize performance while minimizing power consumption. However, doing in hardware alone is difficult due complexity, dissipation, overheads dynamic discovery application characteristics, increased likelihood making locally...

Multicluster architectures overcome the scaling problem of centralized resources by distributing datapath, register file, and memory subsystem across multiple clusters connected a communication network. Traditional compiler partitioning algorithms focus solely on operations to maximize instruction-level parallelism. The distribution data objects is generally ignored. In this work, we examine explicit its affects operation partitioning. must consider several factors: object size, access...

On recent GPU architectures, dynamic parallelism, which enables the launching of kernels from without CPU involvement, provides a way to improve performance irregular applications by generating child dynamically reduce workload imbalance and utilization. However, in practice, parallelism does not due high kernel launch overhead low occupancy. Consequently, most existing studies focus on mitigating overhead. As has decreased algorithmic redesigns hardware architectural innovations,...

Abstract While motion sensors such as accelerometers are commonplace now, highly sensitive conformal that can quantify fine with high resolution remain to be realized. Current strain exhibit a trade‐off between dynamic range and sensitivity; these therefore not suitable for applications require both large excellent signal resolution, gesture control virtual augmented reality or feedback prosthetics. Shape‐memory polymers coupled thin films on stretchable elastomeric support substrates have...

Accelerators have emerged as an important component of modern cloud, datacenter, and HPC computing environments. However, launching tasks on remote accelerators across a network remains unwieldy, forcing programmers to send data in large chunks amortize the transfer launch overhead. By combining advances intra-node accelerator unification with one-sided Remote Direct Memory Access (RDMA) communication primitives, it is possible efficiently implement lightweight tasking distributed-memory...