NVDA Stock Hit All-Time High, Nvidia Unveils Ampere GPU Architecture for AI Boost

UTC by Darya Rudz · 3 min read
NVDA Stock Hit All-Time High, Nvidia Unveils Ampere GPU Architecture for AI Boost
Photo: NVIDIA

Nvidia (NVDA) stock is rising partly thanks to the company’s new GPU architecture named Ampere. On Thursday, Nvidia introduced the architecture and a new line of chips based on Ampere target AI acceleration. 

Nvidia Corporation (NASDAQ: NVDA) stock has hit its all-time high this week. On May 11, it closed at $322.62 but later slightly plummeted with the overall market.

Yesterday, Nvidia shares ended the day 3.22% up at $321.22. After hours NVDA stock rose by 0,12% to $321.60. In the pre-market today, Nvidia shares are trading at $324.25, or 0.94% higher from the previous close. Such good performance can not but fuel analysts’ confidence in the stock. For example, on Wednesday, Wedbush analyst Matt Bryson raised his price target on the NVDA stock from $311 to $340. As Bryson has explained, he sees “a room for upside” for the stock, especially after May 21 when Nvidia delivers its quarterly earnings report.

“Mad Money” host Jim Cramer believes Nvidia stock is now worth buying as well. He said:

“In a market that’s suddenly rolling over after an epic rebound from the lows, we need to circle the wagons around high-quality stocks that we can confidently buy into this newfound weakness, because they’ll be able to make a comeback even if the economy doesn’t.”

According to Cramer, if Nvidia can break its resistance level of about $322, the stock price can have more upside.

NVDA Stock Up Boosted by New Ampere GPU Architecture from Nvidia

The all-time high of Nvidia stock has been seen this week are partly due to the company’s new GPU architecture named Ampere. On Thursday, Nvidia introduced the architecture and a new line of chips based on Ampere target AI acceleration.

The first GPU to rely on Nvidia’s next-gen Ampere architecture is called A100. As Nvidia has said, A100 is already in full production. It packs a massive 54 billion transistors, compared with the 21 billion transistors used by the prior Tesla V100 server that relies on the older Volta architecture.

A100 GPU can be split into seven different instances on the same chip – MIG. Alternately, it can be combined with many different A100 GPUs via NVLink and create a single massive virtual GPU. To make the A100. Nvidia is using a more advanced 7nm TSMC process. Besides, A100 comes with 40GB of HBM2 graphics memory and has access to 1.6TB/s of memory bandwidth.

In a recorded keynote, Nvidia Chief Executive Jensen Huang said:

“Ampere is a miracle. Ampere is the largest, most complex processor the world has ever made.”

The price for the A100 begins at $200,000. The first A100 models have been earlier delivered to Argonne National Laboratory to study the coronavirus.

Rick Stevens, associate laboratory director for Computing, Environment and Life Sciences at Argonne, commented:

“The compute power of the new DGX A100 systems coming to Argonne will help researchers explore treatments and vaccines and study the spread of the virus, enabling scientists to do years’ worth of AI-accelerated work in months or days.”

Earlier, Argonne National Laboratory decided to use artificial intelligence to study COVID-19 and search for a potential drug to fight the virus.

Artificial Intelligence, Business News, Market News, News, Stocks
Related Articles