6 Matching Annotations
  1. Dec 2021
    1. Section 345.5 of ASME B31.3 addresses the requirements of pneumatic testing.

    2. For pneumatic test, section 345.5.4 of ASME B31.3 stipulates that the test pressure shall be 110% of the design pressure. The test media shall be clean dry air or nitrogen. Any other test media may be proposed by the Engineer as long as it is non-flammable and does not adversely affect the piping system or cause a safety hazard. Oxygen shall never be used a test media. As per ASME B31.3, the pneumatic test pressure is required to be increased gradually until the pressure reaches the lower of 25psi or half the test pressure. Thus, if the design pressure is 100psi, the required test pressure is 110psi. Since half of this value is 55psi, the initial pressurisation would not exceed 25psi. After attaining this pressure, all threaded, bolted and other mechanical joints shall be examined as per para. 34.4.1(a). Some engineering specifications conservatively recommend a lower value of initial pressurisation to 10-15psi and checking the system for major leaks. Subsequently, the test pressure is gradually increased in steps, holding the pressure at each step for sufficient time to allow the piping strains to equalize. The minimum recommended pressure steps are lower of 25psi or 25% of test pressure and a holding time of 5-10 minutes at each step is recommended as well by many specifications. The test pressure is then reduced to the design pressure and all joints are examined for leaks

  2. Sep 2021
    1. Frontier supercomputer at ORNL with HPE/Cray based on AMD's custom Epyc CPUs, Radeon Instinct GPUs and Infinity Fabric consuming 30 - 40MW and will be liquid cooled

    2. Aurora supercomputer at ANL with Intel's 2x Sapphire Rapids CPUs and 6x Xeon Xe GPUs per node

  3. Nov 2020
    1. The combination of increased server efficiencies and greater server virtualization (which reduces the amount of server power required for each compute instance) has enabled a sixfold increase in compute instances with only a 25% increase in global server energy use, whereas the combination of increased storage-drive efficiencies and densities has enabled a 25-fold increase in storage capacity with only a threefold increase in global storage energy use

      A sixfold increase in compute instances with only 25% increase in global server energy use

      How is server energy use calculated?

      A 25-fold increase in storage capacity with only a threefold increase in global storage energy use

      How is storage energy use calculated?

    2. The data leveraged here facilitate a more technology-rich and temporally consistent approach than was available previously. Since 2011, analysts at Cisco have published data and outlooks for worldwide server stocks, data center workloads, server virtualization levels, and storage estimates for traditional, cloud, and, most recently, hyperscale data centers (1). In a series of reports starting in 2016, Lawrence Berkeley National Laboratory has published energy trend analyses of servers, storage devices, and network devices commonly used within data centers (8, 11, 13). Analysts have documented the numbers and locations of hyperscale data centers that represent a substantial fraction of global data center compute instances, and major data center operators are increasingly reporting their PUE

      Cisco (2011-Preent): Worldwide server stocks, data center workloads, server virtualization levels and storage estimates for all types of data centers

      Types of Data Centers: Traditional, Cloud, Hyperscale

      LBNL(2016-Present):Energy trend analysis of servers, storage devices and network devices

      Locations and total count if Hyperscale DC representing a substantial fraction of global DCs

      PUE data reported by major DC operators