Inference chip
Search documents
Jim Cramer says you can still find stocks to buy on tough days in the market
CNBC· 2026-03-18 22:58
Company Insights - Nvidia is highlighted as a strong investment opportunity despite current market volatility, with Jim Cramer recommending it as a stock to buy due to its independence from geopolitical tensions and inflation concerns [1] - The company has experienced muted stock performance over the past eight months, but upcoming announcements from the GTC developers event, including a new inference chip and significant order expectations, could lead to a breakout [1][2] - Analysts at Cantor Fitzgerald project Nvidia's earnings per share could reach $15 by 2027, suggesting a forward price-to-earnings ratio of about 12 times, compared to the S&P 500's 18 times [3] Market Context - The broader market faced significant declines, with the Dow Jones Industrial Average dropping over 750 points, or 1.6%, due to inflation fears exacerbated by rising oil prices linked to geopolitical events [4] - Federal Reserve Chairman Jerome Powell's comments indicated that inflation is not slowing as expected, although he alleviated concerns about stagflation, stating that unemployment levels are not indicative of a 1970s-style economic crisis [5] - Cramer noted the difficulty in finding good stocks in the current market climate, emphasizing Nvidia's position as a rapidly growing company with low valuations, which makes it an attractive option despite potential short-term pressures from oil prices and Federal Reserve policies [6]
Nvidia’s GTC 2026 Arrives: Six Key Things Investors Are Watching Out For
Yahoo Finance· 2026-03-17 18:30
Core Insights - Nvidia is set to unveil significant updates at the GPU Technology Conference (GTC) 2026, with a keynote from CEO Jensen Huang [1] Group 1: Inference Chip and AI Strategy - Analysts are anticipating an announcement regarding an inference chip developed in partnership with Groq, which Nvidia acquired for $17 billion last year. This chip is aimed at enhancing AI model deployment for tasks like user query responses, marking a shift from model training to practical AI applications [2] - The company plans to start shipping the next-generation Vera Rubin AI system in the second half of this year, with a focus on the roadmap for its successor, Feynman, named after physicist Richard Feynman [2] Group 2: CPU and Market Competition - There is a renewed focus on CPUs as they become increasingly favored for inference tasks. Investors are looking for updates on Nvidia's strategy to compete with Intel and AMD in the CPU market [2] Group 3: Photonics Development - Nvidia is expected to provide insights into its strategy for developing photonics chips, which utilize light for faster data transfer. The company has committed $2 billion each to photonics manufacturers Coherent and Lumentum [2] Group 4: CUDA and AI Applications - Nvidia's Compute Unified Device Architecture (CUDA) is central to the application of its GPUs in AI training and physical AI. The company is likely to discuss plans to leverage this technology amid the growing demand for physical AI solutions [2] Group 5: Export Regulations and Market Dynamics - The company has received approval to export its H200 processors to China under specific conditions, despite concerns from lawmakers regarding national security. Nvidia's market share in China's GPU sector has dramatically decreased from 95% to zero [2]
Nvidia's battle for inference tech
Youtube· 2026-03-16 16:19
Core Insights - Nvidia's traditional model of offering a single chip for all AI tasks is facing challenges as the industry shifts towards inference, which is expected to dominate the AI market by 2030 [2][4] - The company is set to unveil a new inference chip designed to address the limitations of its current GPUs, which are not optimized for energy efficiency and memory capacity needed for real-time AI tasks [4][5] - Demand visibility and updates on back orders are crucial for Nvidia's stock performance, with a previous back order of $500 billion mentioned [6] Company Developments - Nvidia is expected to announce a new inference chip based on technology from Grock, which was acquired for $20 billion, aimed at improving speed and power efficiency [4] - The options market anticipates a modest 4% stock movement following the announcements, indicating that the market is already somewhat aware of the upcoming products [5] - The focus may shift from just product announcements to broader demand visibility and potential updates on future orders [6] Market Context - Competitors like Google, Amazon, and Meta are developing alternatives for inference, putting pressure on Nvidia's market position [3] - Analysts are speculating on Nvidia's potential comments regarding China and the easing of export restrictions for AI chips, which could positively impact stock performance [8][9] - Nvidia's strategy includes forming partnerships within the AI ecosystem, with recent deals indicating a focus on expanding its technological capabilities [10][11] Financial Considerations - Nvidia has significant cash reserves, leading to speculation about potential buybacks or new partnerships, which could be discussed in upcoming analyst Q&A sessions [12][13]
This Billionaire Tesla Bull Just Backed Up the Truck on Nvidia Stock
247Wallst· 2026-03-10 14:43
Group 1 - Nvidia has a market valuation exceeding $4 trillion, with a remarkable 1,300% gain over the past five years and a gross margin above 75% [1] - Billionaire Leo KoGuan has purchased 1 million shares of Nvidia, valued at approximately $180 million, viewing recent stock volatility as an opportunity rather than a concern [1] - Despite Nvidia's stock facing a ceiling, the company's operational performance remains strong, with high margins and continuous innovation under CEO Jensen Huang [1] Group 2 - The AI chip market is becoming increasingly competitive, which may pressure Nvidia's high margins and growth trajectory [1] - Analysts may not be fully pricing in Nvidia's sustainable high margins, suggesting that the stock could be undervalued if these margins persist longer than expected [1] - Nvidia's recent quarterly performance was strong, but market-wide fears may have muted the stock's reaction, presenting a potential buying opportunity for investors like KoGuan [1]
Did NVIDIA Crash Memory Stocks Like Micron and SanDisk Yesterday?
247Wallst· 2026-03-04 14:32
Core Viewpoint - NVIDIA's potential announcement of a new inference chip at the upcoming GTC conference may have contributed to the decline in memory stocks like Micron and SanDisk, raising concerns about a shift in the memory market narrative [1][2] Group 1: Market Reactions - Micron's shares fell after a significant sell-off in Korean memory stocks, with SK Hynix dropping 11.5% and Samsung 9.9% [1] - Despite the broader market weakness, American memory stocks like Micron and SanDisk showed resilience, with Micron up 3.3% and SanDisk up 4.2% before market opening [1] - The KOSPI index fell 12% in the latest trading session, with SK Hynix and Samsung experiencing further declines of 9.6% and 11.7%, respectively [1] Group 2: NVIDIA's New Chip - NVIDIA is reportedly preparing to launch a new chip that utilizes a radically different architecture, focusing on SRAM instead of high-bandwidth memory (HBM) [1] - The new chip is expected to enhance efficiency for inference workloads, potentially allowing NVIDIA to compete more effectively against companies like Broadcom [1] - NVIDIA's acquisition of Groq for $20 billion indicates a strategic shift towards this new architecture, which could impact the demand for HBM and NAND flash memory [1] Group 3: Implications for Memory Stocks - The introduction of NVIDIA's new chip could lead to a re-rating of memory stocks, particularly if it disrupts the current memory supercycle that has benefited companies like Micron and SanDisk [2] - Micron's guidance suggests projected revenue of $18.7 billion for the next quarter, indicating strong performance despite potential long-term threats from NVIDIA's new technology [2] - SanDisk is expected to report historic EPS growth this year, but the market is concerned about the long-term implications of NVIDIA's advancements [2]