GSI Technology Announces Production Release of Two Gemini-I® APU Servers

Apr 02, 2024

SUNNYVALE, Calif., April 02, 2024 (GLOBE NEWSWIRE) -- GSI Technology, Inc. (Nasdaq: GSIT), the inventor of the Associative Processing Unit (APU), a paradigm shift in artificial intelligence (AI) and high-performance compute (HPC) processing providing true compute-in-memory technology, today announced the release in production of two high capacity, low power server products.  The Leda-E and Leda-S boards incorporate the Gemini-I® APU, capable of 1.2 POPs at 80W and 800 TOPs at 40W, respectively. The new server offerings are available as a single 2U server of eight Leda-E™ boards at 10 POPs, and a 1U server of sixteen Leda-S™ boards, totaling 13 POPs.

GSI's true compute-in-memory technology has proven differentiated capabilities in large (billion item) database search and high-performance compute applications, such as facial recognition, drug discovery, Opensearch, and object detection and re-identification. The low 1.2kW full server power of these formidable systems unlocks the potential for the 1U server to be used in mobile or remote applications.

Single card versions of the Gemini-I have shown six-fold performance improvement over the Xeon® CPU for seed location filtering in DNA read mapping by Cornell University1. Similarly, the Gemini-I® has shown equivalent performance to the A100 for cryptography applications, such as high throughput hash-based security protocols, as demonstrated by Northern Arizona University (NAU)2. NAU found that an 8-card APU server would be compelling for further evaluation due to its smaller form factor in a 2U server compared to an equivalent A100 GPU system.

Server systems based on GSI’s production Gemini-I APU chip come pre-bundled with an application development compiler and can be further customized with Real-Time Fast Back Projection software or Fast Vector Search support.

Systems are shipping from stock. Please contact for more information.


1 Cornell: Accelerating Seed Location Filtering in DNA Read Mapping Using a Commercial Compute-in-SRAM Architecture
Authors: Courtney Golden, Dan Ilan, Nicholas Cebry, Christopher Batten
Link: 2401.11685.pdf (

2 NAU: Evaluating Accelerators for a High-Throughput Hash-Based Security Protocol.
Authors: Kaitlyn Lee, Brian Donnelly, Tomer Sery, Dan Ilan, Bertrand Cambou, and Michael Gowanlock.
Link: Evaluating Accelerators for a High-Throughput Hash-Based Security Protocol (


GSI Technology is at the forefront of the AI revolution with our groundbreaking APU technology, designed for unparalleled efficiency in billion-item database searches and high-performance computing. GSI’s innovations, Gemini-I® and Gemini-II®, offer scalable, low-power, high-capacity computing solutions that redefine edge computing capabilities. GSI Technology is not just advancing technology; we're shaping a smarter, faster, and more efficient future.

Founded in 1995 and headquartered in Sunnyvale, California, GSI Technology has 158 employees and over 125 granted patents. 

For more information, please visit

Forward-Looking Statements

The statements contained in this press release that are not purely historical are forward-looking statements within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, including statements regarding GSI Technology's expectations, beliefs, intentions, or strategies regarding the future. All forward-looking statements included in this press release are based upon information available to GSI Technology as of the date hereof, and GSI Technology assumes no obligation to update any such forward-looking statements. Forward-looking statements involve a variety of risks and uncertainties, which could cause actual results to differ materially from those projected. Examples of risks that could affect our current expectations include the use of our in-place associative computing products in commercial applications and the achievement of specific performance improvements by our products over competing products. Further information regarding these and other risks relating to GSI Technology's business is contained in the Company's filings with the Securities and Exchange Commission, including those factors discussed under the caption "Risk Factors" in such filings.


Investor Relations
Hayden IR
Kim Rogers

Media Relations
Finn Partners for GSI Technology
Ricca Silverio
(415) 348-2724

GSI Technology, Inc.
Douglas M. Schirle
Chief Financial Officer

Primary Logo

Source: GSI Technology, Inc.