For Institutions
About
Goldman Sachs stated that Google/Broadcom's TPU is rapidly narrowing the gap in inference costs with NVIDIA's GPU. The unit token inference cost has decreased by about 70% from TPU v6 to TPU v7, roughly on par with NVIDIA's GB200 NVL72. This does not mean that NVIDIA's position is shaken, but it clearly indicates that the core evaluation system of AI chip competition is shifting from "who computes faster" to "who computes cheaper and more sustainably."
On January 20, Eastern Time, Coreweave options saw a total of 168908 contracts traded, with calls accounting for 49% and puts making up 50%.
On January 20, Eastern Time, NVIDIA options saw a total of 1966189 contracts traded, with calls accounting for 63% and puts making up 36%.
NVIDIA CEO Jensen Huang plans to visit China in late January to explore the artificial intelligence chip market. While it is unclear whether he will meet with senior Chinese officials, he is expected to attend company gatherings and visit Beijing before the Lunar New Year holiday. This trip to China is part of Jensen Huang's routine schedule and is a critical moment for NVIDIA, as the United States has relaxed regulations on the export of H200 chips to China, while the Chinese government is deciding how many chips to allow for import
On Tuesday, Cathie Wood-led Ark Invest made significant trades involving Broadcom Inc., WeRide Inc., BYD Co. Ltd., and Advanced Micro Devices Inc. These trades were executed across various Ark ETFs.