“Moore’s Law Is Dead,” Says Nvidia CEO Jensen to Justify Rising Game Card Prices

By Wallace Witkowski

Silicon wafers used to make chips ‘not a little more expensive, it’s a ton more expensive,’ says Huang

Nvidia Corp. Chief Executive Jensen Huang said Wednesday he thinks it will be “a pretty great fourth quarter for Ada,” the company’s next-generation chip architecture unveiled this week, even as critics balk at a price hike during a slowdown in consumer demand. .

Nvidia (NVDA) expects strong demand for gaming chips using its next-generation “Ada Lovelace” chip architecture, named after a 19th-century English mathematician widely regarded as the world’s first female computer programmer for her work on Charles Babbage’s theoretical analytical engine.

A handful of sales will hit the current quarter as Nvidia’s $1,599 flagship RTX 4090 goes on sale October 12, with other cards like the $899 mid-tier 4080 to follow, and the “big majority” of the launch taking place in January. fourth fiscal quarter, Huang said.

Complaints have been circulating online about the unexpected price hike. For the respective chip class, the 4090 is priced 7% above the 2020 launch price of the 3090 it is supposed to replace. (As for the 3090, an upgraded version of the original was $1,100 at Best Buy in a reported $900 price drop.) Even more striking, the 4080 is priced 29% above the launch price. 2020 from 3080.

Lovelace is the successor to Ampere, which was unveiled in May 2020, approximately two months into the COVID-19 pandemic, amid high demand for game cards. Ampere-based game cards have been introduced in September 2020

Huang certainly paid for that optimism in the form of two quarters of “really tough medicine” after the chipmaker cut its outlook not once or twice, but three times and said $400 million in sales are now up in the air due to a U.S. ban on selling data center products to China and a $1.22 billion fee to clear Ampere-based inventory ahead of Lovelace’s launch.

Read: Nvidia’s ‘China syndrome’: is the stock melting?

“We’re very, very specifically selling in-market much less than what’s selling out-of-market, a significantly lower amount than what’s selling out-of-market,” Huang said. “And hopefully by Q4, in Q4, the channel would have normalized and made room for a great launch for Ada.”

To critics, Huang said he felt the higher price was justified, especially since Lovelace’s state-of-the-art architecture was needed to support Nvidia’s expansion into the so-called metaverse.

“A 12 inch [silicon] wafer is much more expensive today than it was yesterday, and it’s not a bit more expensive, it’s a ton more expensive,” Huang said.

“Moore’s Law is dead,” Huang said, referring to the standard that the number of transistors on a chip doubles every two years. “And the ability of Moore’s Law to deliver twice the performance at the same cost, or at the same performance, half the cost, every year and a half, is gone. It’s completely gone, and so the idea that a flea will shrink over time is unfortunately a thing of the past.”

“Computing is not a chip issue, it’s a software and chip issue,” Huang said.

Nvidia continues to develop its software

This is why, over the years, Nvidia has developed such an entrenched software ecosystem for its chips, that it has prompted some analysts to start viewing Nvidia as a rapidly emerging software company.

This time around, Huang unveiled a big expansion of the company’s so-called metaverse platform with Nvidia’s Omniverse Cloud, the company’s first Software-as-a-Service and Infrastructure-as-a-Service product. company, to design, publish, operate and experiment with metaverse applications.

Another push towards SaaS is Nvidia’s NeMo and BioNeMo cloud AI services. LLMs are machine learning algorithms that use huge sets of textual data to recognize, predict, and generate human language. While NeMo is the general model service, BioNemo specializes in the application of LLMs to biological and chemical research.

Seeing that Nvidia is essentially offering an RTX 3080 gaming chip as a service with its GeForce NOW Priority service which dropped in November, charging subscribers $99.99 for six months of RTX 3080 gaming chip performance, MarketWatch asked to Huang if he ever planned on using purchased physical GPU hardware replaced by cloud-based subscription services.

Read: Nvidia sales forecast about $1 billion lower than expected, shares tumble

“I don’t think so,” Huang said. “There are customers who want to own, and there are customers who like to rent.”

“Some people would prefer to outsource the factory,” Huang said. “And remember, artificial intelligence is going to be a factory, it’s going to be the most important factory of the future.”

“A factory brings raw materials in and something comes out,” Huang said. “In the future, factories are going to receive data, and what will come out of it will be intelligence, models.”

When it comes to factories, Nvidia needs to have options to serve all customers at scale. “Startups prefer to have stuff in opex,” Huang said. “Big, established companies prefer to have things in capex.”

Over the years, Nvidia has shown it’s not resistant to transformation, going from being a gaming chip company to becoming America’s largest chipmaker by market capitalization after data center designers discovered that Nvidia’s graphics processing units, or GPUs, weren’t working. just to make video games prettier, their parallel processors were very useful in machine learning.

Several other technology hardware companies, such as Cisco Systems Inc. (CSCO) and International Business Machines Corp. (IBM), have, over the years and with varying degrees of resistance and enthusiasm, transformed themselves almost out of necessity into software and services companies, as more and more companies migrate their data to the cloud rather than keeping them on site in a proprietary server.

Read: The end of single-chip wonders: Why Nvidia, Intel and AMD valuations have seen a massive upheaval

Of the 43 analysts covering Nvidia, 31 have buy ratings, 11 have hold ratings, and one has a sell rating. Of these, 13 lowered their price targets resulting in an average target price of $202, down from the previous $202.51.

The shares closed Wednesday up 0.7% at $132.61, against a 1.7% decline in the S&P 500 index.

During the year, Nvidia shares fell 55%, compared to a 36% drop in the PHLX Semiconductor index, a 20% drop in the S&P 500 index and a 28% drop for the index. Nasdaq composite, with a strong technological component.

As for the Ampere race, Nvidia’s stock price is down 4.7% since September 1, 2020, when Nvidia unveiled its RTX 3000 series Ampere-based gaming chips, down from a gain of 9 .3% by the S&P 500 over this period.

-Wallace Witkowski


(END) Dow Jones Newswire

9/21/22 1817 ET

Copyright (c) 2022 Dow Jones & Company, Inc.

About Leslie Schwartz

Check Also

PROFILE: Philip Ewell ’01, the cellist shaping the racial calculus of music theory

2022 Wilbur Cross recipient and music theorist Dr. Philip Ewell sat down with the News …