With the computational systems of even embedded devices becoming ever more powerful, there is a need for more effective and pro-active methods of dynamic power management. The work presented in this paper demonstrates the effectiveness of a reinforcement-learning based dynamic power manager placed in a software framework. This combination of Q-learning for determining policy and the software abstractions provide many of the benefits of co-design, namely, good performance, responsiveness and application guidance, with the flexibility of easily changing policies or platforms. The Q-learning based Quality of Service Manager (2QoSM) is implemented on an autonomous robot built on a complex, powerful embedded single-board computer (SBC) and a high-resolution path-planning algorithm. We find that the 2QoSM reduces power consumption up to 42% compared to the Linux on-demand governor and 10.2% over a state-of-the-art situation aware governor. Moreover, the performance as measured by path error is improved by up to 6.1%, all while saving power.
The immense hardware complexity of modern computers, both mobile phones and datacenter servers, is a seemingly endless source of bugs and vulnerabilities in system software. Classical OSes cannot address this, since they only run on a small subset of the machine. The issue is interactions within the entire ensemble of firmware blobs, co-processors, and CPUs that we term the de facto OS. The current "whac-a-mole" approach will not solve this problem, nor will clean-slate redesign: it is simply not possible to replace some firmware components and the engineering effort is too great. Our response, instead, is to build a high-level model of exactly what a given real hardware and software platform consists of, and captures for the first time the necessary and assumed trust relationships between the software contexts executing on different components (CPUs, devices, etc.). This principled but pragmatic approach allows us to make rigorous statements about the hodgepodge of soft- and firmware at the heart of modern computers. We expect these statements to be, at first, depressingly weak, but it may be the only way to identify changes that provably increase the trustworthiness of a real system, and quantify the benefits of these changes.
SUMMARY: Gaze stabilization reflexes reduce motion blur and simplify the processing of visual information by keeping the eyes level. These reflexes typically depend on estimates of the rotational motion of the body, head, and eyes, acquired by visual or mechanosensory systems. During rapid movements, there can be insufficient time for sensory feedback systems to estimate rotational motion, requiring additional mechanisms. Solutions to this common problem are likely to be adapted to an animal’s behavioral repertoire. Here, we examine gaze stabilization in three families of dipteran flies, each with distinctly different flight behaviors. Through frequency response analysis based on tethered-flight experiments, we demonstrate that fast roll oscillations of the body lead to a stable gaze in hoverflies, whereas the reflex breaks down at the same speeds in blowflies and horseflies. Surprisingly, the high-speed gaze stabilization of hoverflies does not require sensory input from the halteres, their low-latency balance organs. Instead, we show how the behavior is explained by a hybrid control system that combines a sensory-driven, active stabilization component mediated by neck muscles, and a passive component which exploits physical properties of the animal’s anatomy—the mass and inertia of its head. This adaptation requires hoverflies to have specializations of the head-neck joint that can be employed during flight. Our comparative study highlights how species-specific control strategies have evolved to support different visually-guided flight behaviors. SIGNIFICANCE STATEMENT: Across the animal kingdom, reflexes are found which stabilize the eyes to reduce the impact of motion blur on vision—analogous to the image stabilization functions found in modern cameras. These reflexes can be complex, often combining predictions about planned movements with information from multiple sensory systems which continually measure self-motion and provide feedback. The processing of this information in the nervous system incurs time delays which impose limits on performance when fast stabilization is required. Hoverflies overcome the limitations of sensory-driven stabilization reflexes by exploiting the passive stability provided by the head during roll perturbations with particularly high rotational kine-matics. Integrating passive and active mechanisms thus extends the useful range of vision and likely facilitates distinctive aspects of hoverfly flight.
The trend towards system specialization is leading to a proliferation of accelerators, exposing interconnects as serious bottlenecks, both in functionality and performance. As a result, several alternative approaches have been proposed which promise to expand the coherence domain beyond homogeneous sockets to rack scale heterogeneous systems. In parallel, GPU vendors have developed their own high bandwidth interconnects also aiming for heterogeneous coherence beyond the CPU. This expansion of the coherency domain raises many questions that remain unanswered, in particular, how devices other than CPUs will interact with the coherence protocol and whether applications can take advantage of these expanded domains. As protocols such as CXL are still evolving, it is important to explore alternative designs that go beyond what the commercial specifications dictate. For this purpose, we developed CCKit, an open-source, server-class toolkit comprising a complete cache coherency stack on reconfigurable accelerators. CCKit is more flexible than commercial products and its performance is highly competitive with hardware-based implementations, thus enabling important and novel application use-cases for expanded coherence domains. Experimental data from real workloads provide the ability to influence and expand future interconnects, protocols, and applications.