Search

This thesis concerns the way technology was conceived in Britain during the period c.1870-1914, when the word 'technology' was not yet available. Just as historians of technology have rarely focused on the conceptual and semantic histories of the word 'technology', intellectual historians have not focused on how technology was conceived during this period. The thesis addresses this gap by focusing on 'ideas of the machine', posited as antecedent to the later concept of technology. This thesis argues that ideas about technological change more characteristic of the early nineteenth century (and known as the 'machinery question') returned in new forms during the period c.1870-1914. It claims that this new understanding of machines consisted in relating them to readings of historical change at large, and so was temporal in nature. During this period, the first historical accounts of the Industrial Revolution of c.1800 were compiled in different settings, producing visions of what is termed here the 'machine past'. As part of their growing ambition, scientists at the British Association produced histories which sought to take credit for the success of new technologies. The historians Arnold Toynbee, William Cunningham and WJ Ashley developed accounts of the Industrial Revolution which presented machines as complex agents of change. Their accounts were built upon by the economist JA Hobson, who is shown to have theorized technology extensively, an aspect of his work which has been hitherto ignored. Contemporaneous with these visions of the 'machine past' were future-oriented works which offered a critical assessment of technological trends. In a range of writings, HG Wells, GK Chesterton and others debated the shape of the 'machine future'. This thesis provides a close reading of influential, competing visions of the machine and demonstrates how projections of machinery (past and future) served as critical commentaries on the growing significance of technology for human life. It concludes that the increasing specialization of intellectual inquiry during this period became a barrier to the holistic investigation of technology, a process described here as a 'technological settlement'. 3

Top high performance computing (HPC) data centers recently entered the era of exascale computing, requiring up to tens of megawatts for a computing facility to meet its users’ computing needs. The massive capacity for power at a single site comes with challenges in power management. Poorly managed power may result in unnecessarily high demand on costly energy or may cause the system to under-perform. An HPC data center may have many types of users and workloads with non-trivial power requirements, making it difficult to select a one-size-fits-all policy. But the high power capacity also offers opportunities for data centers to be key players enabling greater adoption of renewable energy across a power grid. Data centers can adjust their demand through software power management policies to help smart grids balance against nature’s time-varying green energy supply. This thesis claims that multi-tiered power management methods are essential for data centers to implement site-wide power management policies that accurately respond to changing power constraints at a higher cluster-level tier while reacting to application-specific performance impacts at a lower job-level tier. Through investigations over site, cluster, job, and server characteristics, we demonstrate that a feedback-driven multi-tiered power management approach meets power management objectives more effectively than siloed solutions. We design a cluster power management policy that distributes power across jobs using knowledge about job power-performance properties, demonstrating up to 7% reduction in system time dedicated to jobs and up to 11% savings in energy, compared to a policy without job awareness. We provide a power management framework that enables accurate, dynamic cluster power control while reacting to incomplete or inaccurate prior knowledge about job power and performance properties. We add a site-wide power model to a cluster power management policy that offers regulation services in a smart grid, showing 1.3x cost savings compared to a policy that is unaware of site-wide power consumption. We introduce a job power management policy that integrates job performance awareness with knowledge of hardware power-performance trade-offs, demonstrating up to 40% energy reduction and 17% execution time reduction in an imbalanced, compute-bound benchmark compared to a policy without frequency throttling.