Power & Sites · United States

AI Data Centers Drive Demand for SiC, GaN Power Semiconductors

Credited to TheElec · thelec.net · Just now

Power consumption is rising with the spread of artificial intelligence (AI) data centers, highlighting the importance of compound power semiconductors such as silicon carbide (SiC) and gallium nitride (GaN). Data centers are adopting high-voltage power networks to deliver more electricity efficiently. Conventional silicon (Si) power semiconductors have limitations in reducing power losses.

Bae Sung-beom, a senior principal researcher at the Electronics and Telecommunications Research Institute (ETRI), said Oct. 7 at the Korea Power Device Industry Association's Power Semiconductor Industry-Academia-Research K-Forum at Hotel Nongshim in Busan that the growth of the AI industry is causing power shortages worldwide. “Because it is difficult to expand power infrastructure at the pace of actual demand growth, it is becoming increasingly important to use limited energy efficiently,” Bae said.

Data centers operate large-scale computing resources such as graphics processing units (GPUs) around the clock. As AI models grow larger, the amount of GPUs and storage capacity required by data centers increases, driving up power consumption. How efficiently electricity supplied from external power grids can be delivered to servers with minimal losses is also becoming increasingly important.

Improving power efficiency requires reducing losses that occur when electricity from outside is converted to voltages that servers or GPUs can handle. NVIDIA's 800-volt direct-current (DC) data center power architecture has emerged against this backdrop. The 800V DC system distributes electricity inside data centers at a higher 800-volt DC voltage than conventional systems. It simplifies the existing power-supply architecture and transmits electricity at higher voltage, reducing losses during power conversion and wiring.

As power-supply voltages rise, the role of power semiconductors becomes more important. Improving power efficiency in high-voltage and high-power environments has limitations because designing devices to withstand higher voltages makes it more difficult to reduce power losses and heat generation.

Compound power semiconductors such as SiC and GaN are being considered as alternatives. They can operate at higher voltages and temperatures than Si and reduce losses during power conversion. Their ability to switch at high speeds also helps improve the efficiency of power-conversion equipment and enables smaller components.

Applications vary depending on voltage levels and power-conversion stages. SiC is suited to high-voltage, high-power environments. It can be used in equipment that converts electricity supplied from external power grids at thousands to tens of thousands of volts into power that can be used inside data centers. GaN offers fast switching performance at relatively lower voltage levels. It can help reduce power losses when voltage is stepped down inside servers to levels required by GPUs and other computing devices.

Reliability issues need to be addressed for commercialization. Compound semiconductors are manufactured by combining different materials to form devices. Differences in the coefficients of thermal expansion between materials can cause wafers to warp or crack. The thermal conductivity of substrates also affects device output and efficiency. The industry is researching ways to address these issues using new substrate materials and vertical device structures.

Power & SitesUnited States

HD Hyundai Infracore Wins 390 Billion Won Data Center Engine Order in US

Credited to TheElec · thelec.netJust now

Useful

HD Hyundai Infracore has won a 390 billion won order to supply gas generator engines for power generation at data centers in the US.

HD Hyundai Infracore stated on October 7 that it signed a contract with ERock, a US power infrastructure solutions company, to supply gas generator engine long blocks. The contract runs for two years from 2027 through 2028.

The products are high-efficiency gas generator engine long blocks in the 22-liter class. A long block is the main engine assembly excluding fuel and ignition systems. ERock will use the engines to build generator sets that will supply power to hyperscale data centers operated by US big tech companies.

The company plans to expand its data center generator engine business. It will establish a production system for large generator engines ranging from 1 to 3 megawatts (MW), centered on its new Gunsan engine plant, which is scheduled for completion later this month.

HD Hyundai Infracore plans to launch a 1.5 MW engine with a 37-liter displacement in 2027, followed by a 3 MW engine with a 77-liter displacement in 2028.

Original · TheElec

Front Method

Power & Sites · United States

Google Secures More Than 6 GW in Power Deals for AI Data Centers

Credited to TheElec · thelec.net · 4h ago

Google announced a series of power agreements totaling more than 6 gigawatts (GW) as it moves to secure electricity for its artificial intelligence (AI) data centers.

Google signed a long-term power agreement totaling 3,590 megawatts (MW) with Constellation Energy. It also reached an agreement with Black Hills to secure 2.7 GW of power resources, including reserves, to supply a data center in Wyoming. About a quarter of the capacity covered by the Constellation agreement, or 890 MW, will be newly added through uprates at existing nuclear plants.

According to Constellation Energy and Black Hills, Google signed a 20-year power purchase agreement (PPA) with Constellation to secure 890 MW of new carbon-free generation capacity by increasing the output of 11 existing nuclear units. The first uprate is scheduled for 2028, and Constellation will invest heavily to support the project.

Separately, Google and Constellation signed a 15-year energy supply agreement covering 2.7 GW of existing generation assets within the PJM Interconnection, the U.S. East Coast power-grid operator. By committing to purchase power and generation capacity over the long term, the agreement provides Constellation with a stable revenue base to keep its existing generation assets supplying the PJM power market.

Along with securing power, the two companies will apply AI directly to power-plant and grid operations. Constellation plans to use Google Cloud and Gemini Enterprise to automate power-generation equipment monitoring, maintenance, dispatch optimization, site selection, power-flow analysis, permitting, and interconnection planning. Google Cloud will also be used to strengthen security for operational technology (OT) networks at nuclear plants that are separated from external networks. The strategy expands AI beyond a technology that increases data-center power consumption into "AI for energy" aimed at improving power generation and operating efficiency.

Black Hills signed an agreement to supply power to a data center that Google plans to build in Cheyenne, Wyoming. The agreement runs through 2048, with total power resources for the data center reaching 2.7 GW including reserves. Power service is scheduled to begin in late 2027, with the project expected to reach peak load in 2030.

Black Hills will provide up to 590 MW of the total power resources through its own generation facilities and purchases from the power market. To do so, the company plans to build 564 MW of new natural-gas generation at its existing Cheyenne Prairie Generating Station. The remaining 26 MW will be supplied through power-market purchases and retail electric service.

The remaining roughly 2.1 GW will consist of generation resources contracted from third parties in Wyoming. Black Hills will integrate and manage the output of those resources through a private microgrid, coordinating grid operations, reliability, and power dispatch. The agreement was structured so that costs associated with serving the Google data center are not shifted to existing retail customers. Google will bear the costs associated with supplying the data center, while Black Hills will earn returns on its generation investment and fees for managing the microgrid.

The agreements show that Google’s strategy for securing power for AI data centers is shifting toward long-term acquisition and structuring of generation infrastructure itself. The company is combining multiple approaches, including investment in nuclear uprates, maintaining existing generation assets, new natural-gas generation, and third-party resource management, effectively linking nuclear power, gas generation, power markets, and microgrids into a single supply chain to meet the massive electricity demand of AI data centers.

​HD Hyundai Infracore Wins 390 Billion Won Data Center Engine Order in US

HD Hyundai Infracore Wins 390 Billion Won Data Center Engine Order in US