HW Databases, Where art Thou?

Information about computer hardware is invaluable not only for enterprises and consumers, but also for the research community [2]. A lot of hardware components and systems, such as microprocessors, memory, disks, network links, servers, and consumer devices, appear and disappear. Archives for the technical specications of such hardware allow us to understand the historical trajectory of technological advancement, identify urgent problems we face today, and predict the future by extrapolating from current trends, for example: Are the number of cores really doubling every x years? Are mobile devices evolving faster than workstations? How to make trade-os between DRAM bandwidth, capacity, and latency? Are the assumptions made by a paper still applicable at present, and with a larger dataset?