Broadcom vs. Nvidia: 1 Critical Metric Shows Which Artificial Intelligence (AI) Chipmaker Is the Better Buy After Earnings
- NVDA
- AVGO
- AMD
Broadcom (NASDAQ: AVGO) and Nvidia (NASDAQ: NVDA) are two of the biggest names in the chipmaking business. While AMD is sometimes mentioned in the same conversations as those two giant chipmakers, it doesn’t hold a candle to them in the AI accelerator space. Both Broadcom and Nvidia have recently provided quarterly updates to investors, and both of them are doing incredibly well.
But which one is the better buy now?
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a “Double Down” signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same “Total Conviction” signal is flashing for a company 1/100th the size of Nvidia. Continue »
Nvidia has a broad business, while Broadcom is specialized
Nvidia’s chief products are graphics processing units (GPUs), which are broad-purpose accelerated computing units. These flexible chips can be tasked with nearly any type of workload that’s suitable for a parallel processor and deliver incredible results. However, many of a GPU’s capabilities can go to waste if that computing unit only deals with one type of workload in its service life.
Broadcom addresses that problem by designing ASICs — application-specific integrated circuits. These chips are purpose-built for narrow workloads, so anything extraneous to those workloads can be dropped from their design, which saves money in several ways.
ASICs are nothing new; they’ve been used in industry for a long time. Recently, Broadcom has been catering to the hyperscaler and frontier labs markets, designing ASICs to handle the specific workloads generated by AI. Broadcom has partnered with hyperscalers such as Alphabet and Meta Platforms, and large language developers like Anthropic and OpenAI. Orders for those chips are starting to ramp up, which is boosting its results.
Both GPUs and ASICs have their merits, and no AI hyperscaler will ever be able to fully switch away from broad-purpose GPUs from a company like Nvidia. However, as hyperscalers look to optimize their processes, they may find that data centers with a higher share of custom AI chips from Broadcom could make the most sense. We’ll see how it pans out, but currently, it looks like Nvidia is at higher risk of losing some of its massive market share in data center chips to custom designs than Broadcom is of losing share to Nvidia.
This tips the investment thesis scales in favor of Broadcom, as it’s easier to be the hunter than the hunted, but Nvidia is still doing quite well in its own right.
