- Leverage & Inverse ETPs
- Instruments
- Why Us
Your capital is at risk if you invest. You could lose all your investment. Please see the full risk warning here.
Originally starting out in 1961 as the semiconductor products division of Hewlett-Packard, Broadcom (ticker: AVGO) went through a number of iterations before attaining its current form. In the present day, the company designs and develops a wide range of semiconductor and infrastructure software products. The company has been accruing attention from investors over the past couple of years owing to the ongoing AI Boom.
Inside Broadcom’s AI Infrastructure EdgeIn the AI infrastructure arms race, Broadcom’s strength lies in its:
Deep relationships with hyperscale cloud providers, enabling it to design custom silicon (ASIC/“XPU”) tailored to specific AI workloads.
Growing presence in networking / interconnect chips (lately via its long-running “Tomahawk” series of products) that tie AI compute clusters together.
As of the company’s third quarter (Q3) of its fiscal year (FY) 2025, well over half of its revenue was from semiconductor solutions.
Websim is the retail division of Intermonte, the primary intermediary of the Italian stock exchange for institutional investors. Leverage Shares often features in its speculative analysis based on macros/fundamentals. However, the information is published in Italian. To provide better information for our non-Italian investors, we bring to you a quick translation of the analysis they present to Italian retail investors. To ensure rapid delivery, text in the charts will not be translated. The views expressed here are of Websim. Leverage Shares in no way endorses these views. If you are unsure about the suitability of an investment, please seek financial advice. View the original at
However, its infrastructure segment has been accruing a significant amount of investor attention as AI datacenters continue to be built out: large-scale AI systems are less about raw compute and more about how fast one can move data between compute blocks. Broadcom’s interconnect chips and solutions act as a sort of “glue” in the AI stack.
In June 2025, Broadcom began shipping Tomahawk 6, which doubles performance over its predecessor and integrates multiple chips in a single package. This helps reduce the number of switches needed and improves efficiency in large AI data centers. In the next month, it launched Tomahawk Ultra, a networking processor intended to accelerate AI-scale systems by interconnecting chips in rack-scale “scale-up” computing. By this launch, Broadcom effectively competes with Nvidia’s NVLink Switch, with the Ultra being positioned as being capable of connecting four times more chips using an enhanced Ethernet-based protocol.
Its relationship with the likes of Nvidia, however, isn’t necessarily competitive across the board. For instance, Broadcom completed its acquisition of cloud computing and virtualization technology company VMWare in November 2023 – giving it a foothold in enterprise software, virtualization, and hybrid cloud infrastructure. Lately, the company is expanding partnerships to combine VMware Cloud Foundation with AI/ML workloads – including integration with NVIDIA GPUs/DPUs – to allow enterprises to run AI/ML alongside traditional enterprise workloads. The ready offering of solutions that integrate NVIDIA GPUs / DPUs indicates that the company’s approach to the industry is nuanced; it seeks a collaborative—rather than purely competitive—relationship in parts of the stack.
There are a number of peers in the space that Broadcom is competing with (and sometimes collaborating with):
NVIDIA (NVDA): While the dominant player in AI accelerators/GPUs, Broadcom is increasingly positioned as an alternative or complement to Nvidia in AI infrastructure buildouts.
AMD (AMD): Over the past couple of years, AMD has made massive inroads into the AI server / datacenter accelerator segment. While there is some overlap – like with Nvidia – there is also potential for complementary alignment.
Marvell Technology (MRVL): Being an established player in the networking, data infrastructure, and custom silicon segments, Marvell might be considered more of a competitor rather than a potential collaborator.
Cisco (CSCO), Arista Networks (ANET): Established players in the networking/switching space, both might be considered as competitors especially for Broadcom’s data center interconnect hardware.
Towards the end of 2024, Broadcom estimated1 that the AI infrastructure opportunity (including custom silicon, networking and connectivity) could present a $60 billion to $90 billion revenue opportunity in 2027, more than four times the current size of the market. The company anticipates that the AI segment will dominate its future growth, with networking and software infrastructure providing durable support. If Broadcom can scale AI revenues aggressively while maintaining margins, the software and infrastructure side can help stabilize earnings in cyclical periods of semiconductor demand.
Thus, for investors, Broadcom presents a compelling asymmetric bet in the AI infrastructure space: on one hand, it has many of the required pieces – custom silicon, networking, deep customer relationships, software backbone, and so forth. On the other hand, the scale, complexity, and competition are immense.
Through the wax and wane of AI cycle and investor sentiment, professional investors might like to consider the +3X Long Broadcom ETP (AVG3) and the -3X Broadcom ETP (AVOS) during bullish and bearish trends in the stock.Leverage Shares has an extensive catalogue of products offering professional investors exposure to the semiconductor industry. The +3X Long Nvidia ETP (NVD3) is a frequent top-traded product in its segment at the London Stock Exchange and the -3X Short Nvidia ETP (NV3S) is an “inverse” alternative to NVD3. The +3X Long AMD ETP (AMD3) provides leveraged exposure to the performance of AMD’s stock while the -1X Short AMD ETP (AMDS) aims to deliver the same impact as holding a short position in AMD without the need to maintain a margin balance. Also available are the +3X Long Taiwan Semiconductor (TSMC) ETP (TSM3) and the -3X Short Taiwan Semiconductor ETP (TSMS). For a broader yet leveraged exposure, the +3X Long Semiconductor ETP (SMH3) and the -3X Short Semiconductor ETP (SMHS) are at hand.
Footnotes:
Your capital is at risk if you invest. You could lose all your investment. Please see the full risk warning here.
Share this:
INVESTOR TYPE:
LOCATION:
Never miss out on important announcements. Get premium content ahead of the crowd. Enjoy exclusive insights via the newsletter only.
