Topic guide

Racks

Explore sources on rack layouts, power density and IT equipment interfaces. Review technical notes to understand how mechanical, electrical and cooling requirements meet at the rack level.

Sources
191
Downloadable PDFs
52
Years
2008–2026

All sources

73–96 / 191

Research paperPDF

Applied Sciences (MDPI AG) 2024

Numerical Evaluation of Hot Air Recirculation in Server Rack

Make hot-aisle back pressure (hot-aisle pressure above cold aisle) a design input in contained aisles; server inlet temperature rises once back pressure reaches the fans' free-delivery point.

  • A three-grid GCI study (1.22 to 1.99 million cells) selected a 1,799,871-cell trimmed hexahedral mesh with seven prism layers, near-wall y+ of about 5.
Research paperPDF

Buildings (MDPI AG) 2023

Numerical Investigation of Thermal Performance with Adaptive Terminal Devices for Cold Aisle Containment in Data Centers

For hot-spot removal, before choosing lower supply temperature or higher airflow outright, compare the COP drop (3.2 to 2.6) and the energy difference (20.1% and 4.2%) with a rack-specific air distribution option.

  • A grid study at 260,000 to 1,340,000 cells showed results stabilising at about 560,000 cells.
Technical guidePDF

Open Compute Project Foundation 2022

Open Rack V3 BBU Shelf Specification, Rev 1.1

Model ORV3 in-rack BBU as an alternative to a central UPS and check that generator start and transfer time fit inside the roughly 4-minute window at rated load per shelf.

  • Each BBU shelf holds one PMI module and six 3 kW BBU modules in 5+1 redundancy.
Technical guidePDF

Open Compute Project Foundation 2026

Open Rack V3 High Power (60A) AC WHIP Power Cable Design Specification, Revision 1.1.0 (11 November 2026)

Add a high-power ORV3 feed type (60 A 480 V / 63 A 400 V IEC 60309) to the busway/receptacle catalogue and size each whip at 48 A continuous in North America for power-shelf capacity checks.

  • Assembly must be UL Listed per UL 817 and CE certified; parts to UL 62/UL 758; connectors tested to IEC/EN 61984 and IEC 60512.
Technical guidePDF

Open Compute Project Foundation 2026

Open Rack Wide (ORW) Base Specification, Revision V1.0.1 (effective May 5, 2026)

Add an ORW rack family whose footprint, height and payload come from the chosen design specification rather than fixed defaults, prompt users for those values, and warn that floor-loading and aisle-pitch checks depend on them.

  • The frame accepts air-cooled or liquid-cooled 48 V busbars; busbar mounting interfaces are identical to ORV3 HPR.
Research paperPDF

Frontiers in Energy Research (Frontiers Media SA) 2023

Performance evaluation and modeling of active tile in raised-floor data centers: An empirical study on the single rack

With active tiles, always seal under-rack gaps; if left open, low pressure at the tile surface pulls hot air in under the rack and worsens the lower servers.

  • CRACs were set to 21 °C but delivered 23 °C supply, and thermal load was about 65% of cooling capacity, so one CRAC was off.
Research paperFree to read

arXiv preprint (Microsoft Azure) 2023

POLCA: Power Oversubscription in LLM Cloud Providers

In the power planner, apply different diversity and oversubscription factors by workload: inference rows can be provisioned below sum-of-nameplate (for example peak about 80% of provisioned), while training rows should assume near-coincident peaks with only about 10% headroom.

  • POLCA allows 30% more servers in the same inference GPU cluster while maintaining SLOs.
White paperFree to read

Schneider Electric 2024

Power and Cooling Capacity Management for Data Centers

For every rack, row and upstream device store design maximum supply, design maximum demand (nameplate or as-configured) and measured actual where available; enforce supply >= demand at each hierarchy level including under N-1.

  • Safety margins are typically 10-20%, up to 30% or more with weak change control.
Research paperFree to read

arXiv preprint (Microsoft / OpenAI / NVIDIA) 2025

Power Stabilization for AI Training Datacenters

For AI-training halls, carry a 'synchronous load' flag and capture utility time-domain (MW/s ramp up/down, dynamic range) and frequency-domain (critical band, spectral cap) limits as interconnection inputs.

  • GPU power smoothing with the floor at 90% of TDP gave a total energy overhead of 10.5% in simulation.
Technical guideFree to read

Uptime Institute Journal 2020

Rack Density is Rising (Andy Lawrence)

Ask users for modal and peak rack density separately, not just the average, and size cooling and power for the peak cluster.

  • Mean density was calculated after excluding respondents above 30 kW as outliers.

All content and rights belong to the respective authors and publishers. SoSo CAD provides only the citation, its own summary and a link to the source, and is not responsible for the sources' content.

Monthly newsletter · free

Keep your reading list current

Once a month: new library sources, changes to Turkish regulations and upcoming data centre events.