NVIDIA shares (NVDA) is on track to experience a comfortable triple-digit percentage gain again in 2024, after an increase of almost 240% in 2023.

This year’s increase is largely due to a product that didn’t even ship until the last quarter of the year: Blackwell.

It is the largest GPU (graphics processing unit) ever built, created by connecting two chips through a high-bandwidth interface (HBI). Simply put, this translates to a lot of power with high efficiency, which is why it is in such demand by so-called hyperscalers – companies like Alphabet (GOOG) and Microsoft (MSFT) who are building huge data centers to power large language models (LLM).

This strong demand for Blackwell and the 170% rise in Nvidia shares is what led Yahoo Finance to name the chip its 2024 Product of the Year.

Learn more: Here’s why Walmart won the 2024 Yahoo Finance Company of the Year award

“Technically, Blackwell is a beast,” Matt Kimball, an analyst at Moor Insights & Strategy, told Yahoo Finance in an email. “Blackwell represents an exponential leap forward as it is a dual GPU on a single chip with faster connectivity, a larger pool of high bandwidth memory (HBM3E) and the introduction of the decompression engine of NVIDIA to make data processing much faster (up to 6 times compared to last year). to Hopper).

First announced in March, Blackwell is in the right place at the right time. Data centers that power Generative AI or LLMs are currently busy entering data and training these models. Blackwell and his predecessor, Hopper, are well suited to the job.

Blackwell represents a huge step forward in power. It contains 208 billion transistors, more than two and a half times Hopper’s number.

Its GB200 NVL72 server, which combines 72 Blackwell GPUs with 36 Grace processors, delivers up to 30x better performance than the same number of Hopper GPUs for LLM inference workloads. It also consumes up to 25 times less energy.

“For us to leap forward by an order of magnitude is pretty unheard of,” said Dion Harris, director of accelerated data center, HPC (high performance computing) and AI at Nvidia, during a telephone interview. “We were limited by physics, but we recognized that innovating with the HBI would allow us to scale communication and computation to the chip level.”

Spending by U.S. companies on generative AI has increased sixfold in one year, from $2.3 billion in 2023 to $13.8 billion in 2024. according to Menlo Ventures. And the trend is only growing, as major companies, from banking and retail to technology and hospitality, race to introduce chatbots and advanced assistants to their customers.