Fundamentals 9 min read

2022 Computing Strengthening Action Product Catalog (2023): Evaluation of Server Hardware for Energy Efficiency and Performance

The 2022 Computing Strengthening Action product catalog presents a comprehensive evaluation of seven server models, detailing their energy‑saving technologies, performance under CPU, memory and storage loads, and classifying them as excellent, recommended, or outstanding products to guide data‑center hardware selection.

Architects' Tech Alliance
Architects' Tech Alliance
Architects' Tech Alliance
2022 Computing Strengthening Action Product Catalog (2023): Evaluation of Server Hardware for Energy Efficiency and Performance

To implement the requirements of the "Three‑Year Action Plan for New‑type Data Center Development (2021‑2023)" and accelerate the development of high‑tech, high‑computing‑power, high‑energy‑efficiency, and high‑security data centers, the China Academy of Information and Communications Technology (CAICT) jointly conducted the "Computing Strengthening Action" with the Open Data Center Committee and the National Market Supervision Key Laboratory, evaluating data‑center products across four dimensions: basic functionality, computing efficiency, energy‑saving and carbon‑reduction, and customization.

The testing system, illustrated below, covers CPU, memory, network, and storage categories, selecting exemplary servers, accelerator cards, switches, storage arrays, and distributed storage that demonstrate technical innovation and guide equipment R&D.

The action defines a testing framework that integrates basic capability, computing efficiency, energy‑saving, carbon‑reduction, and customized testing; products are classified into three tiers: "Excellent", "Recommended", and "Outstanding". This article focuses on the computing (server) section.

(1) Huawei Server K22R‑02 – A 2U dual‑node rack server featuring dynamic energy‑management technologies such as power‑capping, dual‑power supply, energy‑saving fan speed control, high‑efficiency titanium PSU, and component sleep modes, reducing power consumption while maintaining high performance. It uses Kunpeng920_7260 CPUs, 32 GB × 16 memory, runs openEuler 20.03, and is tested under CPU, memory, and storage I/O loads with power meters to record average power and CO₂ emissions.

(2) FusionServer 2288H V6 – A 2U dual‑node rack server equipped with dual Intel Xeon Platinum 8380 CPUs, up to 32 DDR4 memory modules, 14 PCIe slots, and large local storage, suitable for cloud, virtualization, database, and big‑data workloads. Energy‑saving measures include component sleep, PID fan control, redundant power, and advanced VC heat‑sink, enabling operation up to 45 °C and significant cooling cost reduction.

(3) H3C UniServer R4900LC G5 – Supports up to two Intel Xeon third‑generation processors (max 270 W), 32 × DDR4 modules (up to 3200 MT/s), and BPS persistent memory. It employs advanced CPU liquid‑cooling plates to lower CPU temperature, achieve ultra‑low noise, and extend server lifespan.

(4) EVOC Adam Server R322N6+ – Tested for memory read/write, disk I/O, OS comprehensive performance, and computing efficiency. It achieved a peak of 2335.84 Gflops in computing‑efficiency tests and was awarded the "Outstanding Product" label for its high‑capacity storage, high‑performance computing, and flexible stability.

(5) FitServer R2280 V6 – Demonstrated excellent performance in memory read/write, disk I/O, and OS comprehensive tests, earning the "Excellent Product" designation.

(6) H3C UniServer R4950 G6 – Tested for memory read/write, disk I/O, OS comprehensive performance, and computing efficiency.

(7) H3C UniServer R5350 G6 (no GPU) – Under seven CPU workloads, two memory workloads, and two storage I/O workloads, the server recorded an average power consumption of 926.96 W and a projected five‑year carbon emission of 27 640.96 kg.

These results, together with evaluations of accelerator cards, switches, storage arrays, and distributed storage, are compiled in the "2022 Computing Strengthening Action Product Catalog (2023)" for reference by users selecting data‑center equipment.

Source: Open Data Center Committee (ODCC) and Data Center & IT Equipment Performance & Energy Efficiency Laboratory (TOPE).

data centerenergy efficiencyserver performancecomputing powerhardware testing
Architects' Tech Alliance
Written by

Architects' Tech Alliance

Sharing project experiences, insights into cutting-edge architectures, focusing on cloud computing, microservices, big data, hyper-convergence, storage, data protection, artificial intelligence, industry practices and solutions.

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.