START TYPING KEYWORDS TO SEARCH OUR WEBSITE

Is Nvidia about to change forever?

Will it be the same but better, or totally new?

Posted: By Jon Peddie 01.04.21

Nvidia was founded on April 5, 1993, by Jensen Huang, who used to be the director of CoreWare at LSI Logic, Chris Malachowsky, an electrical engineer who worked at Sun Microsystems, and Curtis Priem, previously a senior staff engineer and graphics chip designer at Sun Microsystems. They wanted to create a virtualized environment for all things IO (including Graphics) and audio, etc., with a strong emphasis on gaming. Their first product, the NV1, was one of the earliest examples of accelerated computing—one where the peripherals are the accelerators. Their idea was that a virtualized programming interface allows hardware to change while software continues to work – i.e., programming architecture. They did that, and the company’s GPU strategy today goes all the way back to that very beginning concept.

But over the past 27 years, the company hit a few speedbumps. Not the least was the market’s acceptance of Nvidia’s first product, the NV1. The NV1 processed audio, graphics using a revolutionary quadratic approach, video, peripherals (joystick), and 2D. Still, Nvidia was too far ahead of the curve, as it turned out, and the industry chose to stay with polygonal graphics. There were criticisms with Nvidia’s approach that 2D graphics performance couldn’t compete with other options at that time, especially in DOS. And its audio quality was not as good as hoped for. Being an expensive AIB, it failed to win the market.

Also, the company had a demanding customer, Sega. Nvidia developed the NV2 with Sega for the forthcoming Dreamcast console. But Sega then turned their back on the partnership and dropped Nvidia. The loss of the forward business and the investment Nvidia made in designing the NV2, which never came to market, almost killed the company. In one of the industry’s most dramatic and daring turnarounds, Nvidia dropped its spherical approach, repositioned the company to a polygon design, and two years later introduced the Riva 128—one of the most successful graphics chips ever.

It was what today we call a pivot.

The company did it again when it pivoted out of the smartphone business, but not before it had developed (with the help of a few acquisitions) an Arm-based SoC. That SoC has gone on to power Nintendo Switch consoles and autonomous vehicles.

The company was also quick to pivot its gaming GPUs into GP compute accelerators, which led to AI and ML applications running on them. Nvidia found itself in the data center’s heady world, with new demands, rules, and customers.

Although Nvidia continued to defy gravity and increased its gaming business while the overall PC market declined, the company knew it needed deeper support for these new markets. So, in March 2019, Nvidia announced a deal to buy Mellanox Technologies for $6.9 billion. It was the largest acquisition the company ever made and gave them the premiere data center networking company. Nvidia now had two of the four essential hardware elements of HPC, data center, AI, and VM. It still didn’t have a storage or a CPU offering—things needed to be a full-fledged data center company.

A year and a half later, in September 2020, the company announced it would buy Arm Holdings from SoftBank Group for $40 billion. Arm had won sockets in supercomputers, servers, and other data center devices. Nvidia now had three of the four essential data center components.

Along the way, in 2018, the company shocked the world and introduced a real-time ray-tracing capable GPU. Nvidia had been a long-term player in the ray-tracing market, highlighted by its acquisition of the leading ray tracing software company, Mental Images, in 2007. No hardware or software company knew more about ray tracing than Nvidia. And one of the most critical uses of ray tracing is supercomputer users who do massive visualizations of airplanes, rockets, viruses, and hydrogen bombs.

But, isn’t Nvidia just a gaming company?

Indeed. Is it?

Nvidia began pivoting itself from a component supplier to a system and platform supplier in the early 2000s. The company had been a GPU-based accelerator board supplier for supercomputers for over a decade and is found in the top 10 of the world’s top 500 supercomputers. Demonstrating how it had become a full-fledged system company, a high-performance systems company, Nvidia announced In October 2020, it would build the most powerful computer in the United Kingdom in Cambridge. Named Cambridge-1.

But, isn’t Nvidia just a gaming company?

So, who and what is, or will be, Nvidia?

Nvidia WAS a gaming company. Now Nvidia is a computer company with strong interests in gaming, but its future is computing. CPU computing, GPU computing, and network computing. Dell dropped the name Computer in 2003. Maybe Nvidia should pick it up and become Nvidia Computer, or Nvidia computing.

Whatever the company calls itself, it is not the old Nvidia we once knew. That’s not to suggest Nvidia people will lose their lust for gaming, they won’t, but the payoff, the ROI, isn’t what it was and won’t be as crucial to the company as it was either.

The percentage of gaming in terms of its overall sales has been declining for a while now, as depicted in the following chart.

Nvidia’s gaming revenue percentage over time

 

Nvidia will remain a leader in gaming. But its future will be in computing—computing from simple little beeping IoT devices to supercomputers. The financial statement will shift as Arm sales (est ~$1 billion, maybe a little less today) start to click in. Revenue-wise, that’s not much compared to the ~$6 billion the company pulls in from gaming. But Nvidia didn’t buy Arm for revenue. It bought it for sockets, customers, and associative opportunities.

What do we think?

As Intel enters Nvidia’s cash cow GPU market, Nvidia enters Intel’s cash-cow CPU market. And Arm outsells Intel in unit CPUs by 20 to one or more. Intel can leverage its CPU sockets and offer a GPU. Now Nvidia can do the same thing. And Arm demonstrated how effective that can be with their puny Mali GPU. Just imagine what Arm customers will do with an Nvidia GPU.

So, gaming got them here, and gaming will forever be a part of Nvidia, but Nvidia is no longer just a gaming company.