GSI Technology's APU Breakthrough Sparks Stock Surge and $50M Offering

GSIT stock skyrockets 200% today as Cornell study confirms company's APU matches GPU performance

GSI Technology (GSIT) stock jumped 200% after Cornell University confirmed its APU can match NVIDIA GPU performance using much less energy. The Gemini-II APU is faster, efficient, and ideal for AI tasks. The company shows strong revenue growth, a solid financial position, and new board leadership, boosting investor confidence and interest in its innovative semiconductor technology. GSI Technology, Inc. is a semiconductor company with a market value of $148 million. Today, GSI stock jumped 200% after news about its technology was released. Researchers at Cornell University tested GSI's Associative Processing Unit (APU) and confirmed it can match the performance of NVIDIA GPUs while using 98% less energy. Over the last six months, GSI stock had already increased by 95% according to InvestingPro data. The Cornell study was presented at the Micro '25 conference. It showed that the Gemini-I APU could handle retrieval-augmented generation (RAG) tasks at the same speed as NVIDIA's A6000 GPU but used much less power, as per the report by Investing.com. The study also found that the Gemini-I APU can perform retrieval tasks several times faster than normal CPUs, reducing processing time by up to 80%. Lee-Lean Shu, Chairman and CEO of GSI Technology, said, "Cornell's independent validation confirms what we've long believed -- compute-in-memory has the potential to disrupt the $100 billion AI inference market". The APU was tested on datasets ranging from 10GB to 200GB. The study also introduced a new way to test compute-in-memory devices. GSI's Gemini-II APU, which is the second generation, is said to deliver 10 times faster throughput, lower delays, and better energy efficiency for AI workloads compared to Gemini-I. The company thinks the APU's combination of speed, energy efficiency, and programmability will help it grow in edge AI, data centers, defense, and other areas where saving energy is important. GSI Technology was founded in 1995 and is based in Sunnyvale, California. It focuses on semiconductor memory solutions, as per InvestingPro. The company has a strong financial position, with a current ratio of 5.79 and more cash than debt. In the first quarter of fiscal year 2026, GSI reported $6.3 million in revenue, a 35% increase from last year. The company's earnings per share showed a loss of $0.08, which matched market expectations. After the annual stockholders' meeting, GSI elected five directors: Elizabeth Cholawsky, Haydn Hsieh, Ruey L. Lu, Lee-Lean Shu, and Ronald R. Steger. The votes in favor ranged from 12,479,186 to 12,618,015, and there were 6,103,899 broker non-votes, Investing.com. Overall, the combination of Cornell's validation, new Gemini-II technology, revenue growth, and strong financials gave investors confidence and caused the 200% jump in GSIT stock. Q1. Why did GSIT stock jump 200% today? GSIT stock soared after Cornell University confirmed its APU matches NVIDIA GPU performance while using much less energy. Q2. What is GSI Technology's Gemini-II APU? Gemini-II is GSI's second-generation APU that is faster, more energy-efficient, and designed for AI tasks in edge, data centers, and defense.

Why Is GSI Technology Stock Falling Tuesday? - GSI Technology (NASDAQ:GSIT)

GSI Technology Inc. (NASDAQ:GSIT) shares fell Monday after the company announced a $50 million registered direct offering with a single institutional investor, just days after unveiling a major AI breakthrough validated by Cornell University. The company will issue 1.51 million shares of common stock at $10 each and 3.49 million pre-funded warrants at $9.99 per warrant. The offering is expected to close on or about Oct. 22, 2025, pending customary closing conditions. Needham & Company is serving as the sole placement agent. GSI said it plans to use the net proceeds, alongside existing cash, for general corporate purposes, including further development of its Gemini Associative Processing Unit (APU) for high-performance and AI computing. Also Read: Wolfspeed's Golden Cross Ignites -- Turning Israel Englander's $2 Shares Into 1,299% Moonshot Cornell Validation of AI Compute-in-Memory Technology The financing comes after Cornell University researchers independently confirmed GSI's compute-in-memory (CIM) design can achieve GPU-class performance on large AI tasks while consuming roughly 98% less power. That breakthrough initially sent GSI shares up nearly 200%, drawing strong retail and institutional interest. CEO Lee-Lean Shu said Cornell's findings reinforced GSI's belief that compute-in-memory could disrupt the $100 billion AI inference market by delivering GPU-level processing at a fraction of the energy cost. The company's Gemini-I APU was shown to complete retrieval-augmented generation workloads faster than traditional CPUs. After last week's sharp rally, GSI Technology shares have since retreated, trading lower on Tuesday as investors digest the offering and evaluate potential dilution. GSIT has a 52-week high of $18.15 and a 52-week low of $1.62. The shares have surged more than 305% year to date and currently trade about 203% above their 50-day simple moving average. Price Action: GSIT shares were trading lower by 9.21% to $11.78 at last check Tuesday. Read Next: Coinbase Capitalizes On Trump's Pro-Crypto Agenda With $375M Echo Deal Photo by YAKOBCHUK V via Shutterstock GSITGSI Technology Inc$11.88-8.40%OverviewMUMicron Technology Inc$204.13-1.28%NVDANVIDIA Corp$182.41-0.12%SMHVanEck Semiconductor ETF$346.71-0.15%SOXXiShares Semiconductor ETF$292.42-0.19%Market News and Data brought to you by Benzinga APIs

Why Is GSI Technology Stock Soaring Monday? - GSI Technology (NASDAQ:GSIT)

Shares of GSI Technology, Inc. (NASDAQ:GSIT) are rising Monday after the company shared a significant technological advancement regarding high-performance computing power. GSIT stock is challenging resistance. See what the experts say here. What To Know: GSI Technology announced the publication of a paper led by researchers at Cornell University. Findings from the research paper show that the company's APU CIM (Compute-In-Memory) architectures can match GPU-level performance for large-scale AI applications with a dramatic reduction in energy consumption. GSI Technology's stock was up nearly 200% following the news, according to Benzinga Pro. GSI Technology said its Gemini-I associative processing unit delivered comparable throughput to Nvidia's A6000 GPU on RAG workloads. The unit delivered over 98% lower energy consumption than a GPU over various large datasets. GSI Technology said the APU's unique design allows it to perform retrieval tasks several times faster than standard CPUs, shortening total processing time by up to 80%. "Cornell's independent validation confirms what we've long believed -- compute-in-memory has the potential to disrupt the $100 billion AI inference market," said Lee-Lean Shu, CEO of GSI Technology. "The APU delivers GPU-class performance at a fraction of the energy cost, thanks to its highly efficient memory-centric architecture." GSIT Analysis: GSI Technology has demonstrated remarkable performance year-to-date, up 358.3%, which underscores its appeal among investors looking for high-growth opportunities. From a technical perspective, the stock is trading approximately 306.8% above its 50-day moving average of $3.67 and about 345.7% above its 200-day moving average of $3.35. This substantial upward movement indicates strong bullish momentum, although its RSI of 63.30 suggests it's nearing overbought territory, which could lead to increased volatility in the near term. GSIT Price Action: GSI Technology shares were up 180.08% at $14.21 at the time of publication on Monday, according to Benzinga Pro. The stock is trading within its 52-week range of $1.62 to $18.14. Read More: Nvidia's Real 'Bottleneck' Isn't Chips, It's Power, Says Beth Kindig, Tech Analyst Shares What It Means For AI Investors Image: Shutterstock GSITGSI Technology Inc$14.13178.2%OverviewMarket News and Data brought to you by Benzinga APIs