Navitas shares surge 5% on Nvidia AI infrastructure partnership

Navitas Semiconductor (NVTS) launched a 24 kW DC-DC power board with NVIDIA (NVDA) for AI data centers, causing a 5% after-hours stock gain. The company's P/S ratio is 61.67, much higher than its historical median, indicating high growth expectations. Its GF Score is 63/100, showing moderate financial health with strengths in stability and momentum but weaknesses in profitability. Insiders sold $150.7M in shares over the past year, while five gurus hold positions, with four adding to them.

Navitas Semiconductor (NVTS) shares rose 6.21% after hours on Monday. The company unveiled an 800 V-to-12 V power delivery board, developed with Nvidia (NVDA), for AI data centers. The board uses GaNFast technology and aims for 97% efficiency at full load.

Intel (INTC) is expanding into the GPU market with its new D-Series processors for AI inferencing, competing with Nvidia (NVDA). The chips are deployed via PCIe cards and Intel is partnering with OEMs like ASUS, HPE (HPE), and Dell (DELL). Intel's stock has surged 183% YTD, driven by CPU demand for AI. AMD (AMD), Arm (ARM), and Qualcomm (QCOM) are also competing in the data center CPU market.

Arista Networks (ANET) shares fell 2.7% to $211.01 on October 12 as it announced plans to expand Ethernet use in AI computing, aiming to reduce customer dependency on proprietary networks. The company introduced the Etherlink SU-144 architecture and liquid-cooled 7060XE7 switches to address data bottlenecks in AI clusters.

Amazon (AMZN) is reportedly considering a $7 billion acquisition of AI startup Decart, which develops software for AI workload management and 3D environment simulation. Decart, valued at $4 billion in May, could help AWS compete with Nvidia (NVDA) in AI infrastructure. The deal is not guaranteed, according to The Wall Street Journal.

Microsoft's shares rose 1% after announcing plans to expand AI workloads using its Maia chips. The company introduced Maia 300, aiming for multi-gigawatt capacity and 96 chips per rack. Maia 300 is expected to process up to seven times more AI tokens per rack than its predecessor and deliver 40% more tokens per megawatt than leading GPU systems. Production deployments of Maia 300 are due to start in 2027, with Maia 200 already running production workloads, including Copilot.