The acquisition of Arm by Nvidia is now official. This is a real turning point for players in the new technologies and semiconductor market, but also the second major acquisition for Nvidia in 2020, after that of Mellanox in April.
These two transactions are complementary, as they are both fundamental to Nvidia’s plan to acquire and maintain a leading role in AI workloads, in data centers and beyond. As we noted, GPUs are a boon to machine learning workloads. Nvidia also took note of it and responded to it very early and successfully. Result: the company has built an additional and growing market. Machine learning is now on everyone’s lips, as is cloud computing.
Machine learning workloads are indeed a great addition to the cloud. For starters, the training phase for machine learning algorithms is quite demanding in terms of computation. For many organizations, it doesn’t make sense to purchase the kind of infrastructure needed for these workloads, and that’s where cloud computing comes in.
A multi-pronged strategy
Besides on-demand usage and elasticity, there are other reasons why sending machine learning workloads from machines to the cloud makes sense in many cases. AI workloads are best performed by specialized hardware, which is why Nvidia has expanded its presence in data centers.
The acquisition of Mellanox was already an important piece of this puzzle, as its technology enables better networking in data centers for Nvidia’s chips. This is a huge advantage. Especially since the company has contributed to the solid profits reaped by Nvidia in the second quarter of the current fiscal year.
But it’s the big picture of those profits that’s really important here: Nvidia beat second-quarter estimates with record data center sales. Affected revenues reached $ 1.75 billion, an increase of 167% over the previous year. Showing once again that data centers are a growth engine for Nvidia.
Go beyond data centers
The acquisition of Arm also plays a role in this register. The artificial intelligence data center workload is increasing, and competition for a piece of it is intensifying from Intel’s two emerging start-ups. Faced with this competition, Nvidia seeks to obtain a double result: better performance and better economy.
This was one of the key themes of the recent unveiling of Nvidia’s AI Ampere chip. But also a key theme in the existing collaboration with Arm. Recently, Nvidia added support for Arm processors. While the performance of these processors is not yet up to Intel’s, their low power consumption makes them an attractive option for data centers.
Data centers are the latest expansion of Arm’s central units. Traditionally, Arm’s strength lies beyond the data center. The fact that they are used by Qualcomm – its Snapdragon models are found in a large number of cellphones – is proof of this. Nvidia has been adamant that it intends to expand beyond the data center as well.
The fact that Nvidia has been working with Arm for a while now probably means that we can expect the software side of things, in terms of Arm processor support, to evolve smoothly as well. With the acquisition of Arm, Nvidia continues to execute its plan, while posing an ever greater challenge for those working on new architectures. Let it be said: Nvidia’s competitors will now have to beat the founder, not only on the performance front, but also on the economic and ecosystem aspects, which both seem to have been improved.
Source : ZDNet.com