Author page description
Today
- Auras Shows Next-Gen Intel Oak Stream Cold Plates at Computex 2025
➀ Auras showcased its next-gen Intel Oak Stream cold plates for upcoming Xeon CPUs at Computex 2025, designed to support high thermal loads;
➁ The cold plates measure 156.0×107.5×24.2mm, larger than AMD’s SP7 counterparts, and feature a revised four-post retention mechanism;
➂ These designs highlight the growing adoption of liquid cooling in data centers, particularly for accelerated servers and high-density computing environments.
- Auras AMD SP7 Liquid Cooling Cold Plate for 600W CPUs Shown at Computex 2025
➀ Auras showcased its AMD SP7 liquid cooling cold plate for 600W CPUs at Computex 2025, targeting next-gen high-performance server processors;
➁ The compact cold plate measures 120.1×100.6×22.3mm and supports dual-socket designs with six mounting points, outperforming Intel's upcoming Oak Stream counterpart in size;
➂ AMD's SP7 platform, though not yet detailed publicly, is expected to power next-generation servers with 16-channel memory and higher core counts for advanced computing workloads.
May 27
- This is the ASPEED AST2700 Next-Gen BMC
➀ ASPEED unveiled its next-gen AST2700 BMC at Computex 2025, featuring upgraded quad Arm Cortex-A35 and dual Cortex-M4 cores for enhanced server management capabilities;
➁ The BMC integrates new 1Gbps LTPI interface, PCIe Gen4, DDR5 support, and Caliptra-based silicon root of trust security;
➂ Shown earlier than expected on Pegatron's OCP DC-SCM 2.1 module, it aims to counter emerging BMC competitors like Axiado with higher performance and integration.
April 1
- Supermicro SYS-112C-TN Review 1P Intel Xeon 6 1U Server
➀ The Supermicro SYS-112C-TN is a 1U single-socket server designed for high I/O connectivity and performance.
➁ The server features an Intel Xeon 6700P performance core series CPU and offers eight 2.5" U.2 NVMe drive bays.
➂ The review highlights the server's design, hardware specifications, and performance capabilities.
March 30
- STH Q1 2025 Letter from the Editor Re-calibration and Expansion
➀ This letter discusses the changes in the server industry, particularly the rise of AI servers and Intel's pullback;
➁ The challenges faced in reviewing high-end systems, including customs delays and logistical issues;
➂ The team expansion at STH, including new hires and the transition of roles;
➃ The growth of the STH YouTube channel and plans for a one million subscriber audience;
➄ The launch of the Axautik Group and its focus on financial community content;
➅ The updates on the STH Labs shorts channel and the importance of newsletter and YouTube subscriptions.
March 29
- The Crucial 128GB DDR5-5600 SODIMM Kit We Are Using in Mini PCs
➀ The Crucial 128GB DDR5-5600 SODIMM kit includes two 64GB SODIMMs, enabling 128GB of memory in mini PCs with dual SODIMM slots;
➁ The kit has been tested in systems like the Lenovo ThinkCentre M75q Tiny Gen5 and Acemagic F3A with AMD Ryzen AI 370 HX, as well as an upcoming Intel platform;
➂ The kit is expected to be featured in several systems over the next few months.
- The Key Feature for Server CPUs is Still Two Threads Per Core
➀ This article discusses the continued prevalence of two threads per core in server CPUs;
➁ It highlights the performance benefits of simultaneous multi-threading (SMT);
➂ The author uses AMD EPYC 9005 platforms to demonstrate the concept.
March 28
- The NVIDIA Rubin NVL576 Kyber Midplane is Huge
➀ NVIDIA has a new midplane design for the Rubin NVL576 generation;
➁ The design allows for higher compute density and removes the cable cartridge;
➂ The midplane is large and optimized for liquid cooling, enabling higher density without airflow considerations.
March 27
- Mapping Licensing for Virtualization is Cool Now
➀ The article discusses the importance of virtualization licensing in server hardware costs.
➁ It focuses on AMD's EPYC processors and their optimization for virtualization applications.
➂ The author explores how newer processors offer performance improvements, which can lead to significant cost savings in licensing.
March 26
- NVIDIA DGX Station GB300 Edition Launched Without a GPU
➀ NVIDIA has launched the DGX Station GB300 edition, a new AI platform without an integrated GPU;
➁ The system is based on the NVIDIA Grace Arm CPU and requires an additional GPU for video output;
➂ The platform is designed for AI researchers and aims to be a powerful development tool.
March 25
- Micron SOCAMM Memory Powers Next-Gen NVIDIA Grace GB300 Servers
➀ Micron's SOCAMM memory is being used to power the next generation of NVIDIA Grace GB300 servers;
➁ The new SOCAMM memory form factor offers modular LPDDR5X memory for NVIDIA Grace CPUs;
➂ The modular design allows for easier upgrades and changes in memory configuration during the manufacturing process.
March 24
- This is the Solidigm Liquid-Coolable NVMe SSD Design
➀ Solidigm showcased its liquid-cooled NVMe SSDs at NVIDIA GTC 2025;
➁ The new Solidigm D7-PS1010 E1.S 9.5mm SSD features a cooling design that can cool both sides of the PCB on the drive;
➂ This design will be important for future AI servers due to increasing power density and the need for more efficient cooling solutions.
March 22
- Kioxia LC9 122.88TB PCIe Gen5 NVMe SSD at NVIDIA GTC 2025
➀ At NVIDIA GTC 2025, Kioxia showcased the LC9 122.88TB PCIe Gen5 NVMe SSD, which features a 2.5" form factor and 122.88TB of storage capacity.
➁ The drive utilizes Kioxia's 2TB QLC BiCS FLASH generation 8 3D flash memory and offers a 0.3DWPD endurance rating.
➂ It is compatible with PCIe Gen5 interface, supporting either a single x4 for standard servers or x2x2 for dual-ported high-availability deployments.
March 21
- Samsung 9100 PRO 2TB PCIe Gen5 NVMe SSD Review
➀ Samsung has released their new 9100 PRO 2TB Gen5 NVMe SSD;
➁ The drive is vertically integrated with Samsung's own controller and 236-layer TLC NAND;
➂ The 2TB model offers 14700 MB/s sequential read and 13400 MB/s write speeds, targeting the top-end of Gen5 performance.
- The NVIDIA DGX Spark is a Tiny 128GB AI Mini PC Made for Scale-Out Clustering
➀ The NVIDIA DGX Spark is a compact AI mini PC designed for scale-out clustering;
➁ It features 20 Arm cores, 128GB of LPDDR5X memory, and a Blackwell GPU;
➂ Priced at $3999, it offers 200GbE NVIDIA ConnectX-7 networking and is compatible with clustering configurations.
March 20
- This is the ASUS Ascent GX10 a NVIDIA GB10 Mini PC with 128GB of Memory and 200GbE
➀ The ASUS Ascent GX10 is a NVIDIA GB10 Mini PC featuring 128GB of memory and 200GbE connectivity.
➁ It is designed to be portable and compact, suitable for various applications.
➂ The system includes an NVIDIA ConnectX-7 NIC for 200GbE clustering and a 10GbE NIC for basic networking.
- Softbank to Acquire Ampere for $6.5 Billion in Cash
➀ Softbank has announced the planned acquisition of Ampere Computing for $6.5 billion in cash;
➁ Ampere Computing will retain its name and product lines such as AmpereOne and Ampere Altra;
➂ The acquisition is expected to close in the second half of 2025, with major investors selling their stakes.
March 19
- NVIDIA GTC 2025 Keynote Live Coverage
➀ We are covering the NVIDIA GTC 2025 keynote live, expecting a number of announcements in GPUs, AI, networking, and robotics.
➁ The event is held at the SAP Center, San Jose, due to the large number of attendees.
➂ NVIDIA has seen massive growth with the AI boom, becoming a leader in major data center categories.
March 18
- Acemagic F3A an AMD Ryzen AI 9 HX 370 Mini PC with up to 128GB of RAM
➀ The Acemagic F3A mini PC is powered by the AMD Ryzen AI 9 HX 370 CPU, offering high performance and expandability with up to 128GB of RAM.
➁ The system features RGB lighting and various connectivity options, including USB4, HDMI, and DisplayPort.
➂ It is designed for AI and other demanding applications, making it a versatile choice for users needing a compact and powerful system.
March 17
- ASUS ESC8000A-E13P Review an 8x PCIe GPU Server
➀ The ASUS ESC8000A-E13P is an 8-GPU AI server designed with two AMD EPYC 9004/9005 processors and support for high-speed networking.
➁ The server features an NVIDIA MGX design and can accommodate 8x PCIe GPUs, providing 384GB of VRAM and 384 CPU cores.
➂ The review covers the external hardware overview, including the front and rear of the system, and details on the storage, cooling, and expansion options.
March 15
- STH Weekly Newsletters You Want to Subscribe in Q1 2025
➀ Introduction to STH Weekly Newsletters;
➁ Launch of new Substack by Axautik Group;
➂ Explanation of the format and content of the newsletter, including weekly recap, top five stories, and content preview.
- CoolIT Shows off a 4kW Coldplate for Liquid Cooling
➀ CoolIT has showcased a new 4000W cold plate designed for liquid cooling;
➁ The cold plate is capable of cooling 4000W at a flow rate of 6L per minute;
➂ The technology is expected to extend the longevity of single-phase direct liquid cooling for future-generation accelerators.
March 14
- ASRock Rack 4U8G-TURIN2 A Fresh Take on a Classic AI Server Design
➀ The ASRock Rack 4U8G-TURIN2 presents a new approach to the classic AI server design, utilizing AMD EPYC CPUs and featuring unique server engineering aspects.
➁ The server is equipped with 2.5" drive bays, NVMe SSDs, and offers flexibility with its various connectivity options and expansion slots.
➂ The review highlights the server's design focus on PCIe expansion cards, power supplies, and rear I/O connectivity, emphasizing its capabilities for high-end card support and redundancy.
March 13
- Kioxia LC9 is the 122.88TB PCIe Gen5 NVMe SSD
Kioxia LC9 is a new 122.88TB SSD from Kioxia designed for high-capacity and high-density AI clusters. It supports PCIe Gen5 and offers a 0.3DWPD endurance, making it suitable for large-scale data storage needs.
The LC9 uses Kioxia's 2Tb QLC BiCS FLASH generation 8 3D flash memory and is targeted at reducing power, footprint, and TCO compared to traditional hard drives.
The high capacity and advanced interface of the LC9 make it a compelling option for AI storage arrays.
- Lip-Bu Tan Takes Over as Intel CEO
➀ Lip-Bu Tan has been appointed as the new CEO of Intel, starting March 18, 2025;
➁ The appointment comes after a series of leadership changes at Intel, including the resignation of Pat Gelsinger;
➂ Tan's background as a Silicon Valley royalty and his experience as the CEO of Cadence Design Systems are seen as crucial for Intel's future direction.
March 11
- AMD Ryzen 9 9950X3D Review
➀ The AMD Ryzen 9 9950X3D is a successor to the 7950X3D, featuring Zen 5 architecture and 144MB of L2+L3 cache.
➁ The chip offers a 10-15% performance boost due to the advanced Zen 5 architecture and improved 3D V-Cache stacking.
➂ It is compatible with the AM5 platform and has a 170W TDP, addressing previous limitations of X3D chips.
- Altera Agilex 3 Launch Starting the Path of Lower-end Upgrades
➀ Altera has launched the new Agilex 3 family, bringing the 2nd gen HyperFlex architecture to the lower-end market segment.
➁ The new Agilex 3 promises higher performance and lower power consumption than the Cyclone V.
➂ The Agilex 3 FPGAs are expected to be widely used in various systems over the next few years.
March 9
- Inside a 100G SR4 QSFP28 Module A Quick Teardown
➀ The article provides a quick teardown of a 100G SR4 QSFP28 module;
➁ It discusses the internal components and design of the module;
➂ The author compares the module to a 100G DAC module and highlights the differences in complexity.
- How to Reset a HPE iLO Administrator Password From the OS Using ipmitool
➀ This article provides a guide on how to reset the HPE iLO administrator password using ipmitool from the OS;
➁ The guide includes steps to access a root shell in Linux, finding the admin user, setting the iLO administrator password, and setting user access;
➂ The process is useful for servers with physical access and no iLO password.
March 8
- HPE ProLiant DL145 Gen11 Review An AMD EPYC 8004 Edge Server
➀ The HPE ProLiant DL145 Gen11 is reviewed, highlighting its features and performance as an AMD EPYC 8004 edge server;
➁ The server's design, I/O options, and expandability are discussed;
➂ The review also touches on the server's power efficiency and potential use cases.