We're initiating a new position in Broadcom (AVGO), buying 33 shares at roughly $865 each. Following Thursday's trade, AVGO will represent a 1% weighting in Jim Cramer's Charitable Trust, the portfolio we use for the Club. We would have also exited Advanced Micro Devices (AMD) if not restricted. We plan to sell AMD when we are cleared. As members know, the structure of data centers is changing. The age of artificial intelligence has brought on the need for larger and larger data sets to be analyzed faster than ever before. With Moore's Law coming to an end, CPUs (central processing units) are not capable of handling these needs on their own — and as a result, accelerated computing, using GPUs (graphics processing units), is fast becoming the new standard. (The semiconductor industry is a complex web of designers, manufacturers, and support companies. Here's our primer on how it all works and the important terminology.) After Nvidia's (NVDA) blowout quarter Wednesday evening, we wanted to share two important quotes from CEO Jensen Huang about the kind of computing power companies need in this day and age. "The world has something along the lines of about $1 trillion worth of data centers installed, in the cloud, in enterprise and otherwise. And that $1 trillion of data centers is in the process of transitioning into accelerated computing and generative AI. We're seeing two simultaneous platform shifts at the same time," Huang said. The CEO and co-founder later explained that when it comes to high demand for the company's products, "the world is transitioning from general-purpose computing to accelerated computing [and] the best way for companies to increase their throughput, improve their energy efficiency, improve their cost efficiency, is to divert their capital budget to accelerated computing and generative AI." However, we are at the point where a single server, even an accelerated one, isn't powerful enough to handle the workloads new AI developments, such as large language models (like Alphabet's BARD and ChatGPT from Microsoft -backed OpenAI), require. That's where the backend network comes into play. This issue is not solved by having several disconnected serves work independently. Rather a network is needed to connect multiple serves together and allow them to operate as one. Think about the human brain, it doesn't work if each lobe is disconnected from one another. They need to be able to communicate together and transfer information, working as a single unit for the human mind to operate as it should. A modern data center must be able to operate in a similar manner, which is why we're getting into Broadcom . Broadcom's AI Network solution, Jericho3-AI , was released earlier this year. It's the company's answer to increasingly complex backend AI network requirements. According to Broadcom, "Jericho3-AI revolutionizes AI networking with best-in-class capabilities such as perfect load balancing, congestion-free operation, ultra-high radix, and Zero-Impact Failover, all culminating in significantly shorter job completion times for any AI workload." In English, that basically means that Jericho3-AI "provides the highest performance while also enabling the lowest total cost of ownership." Another leg to the Broadcom story: It's in the process of acquiring the enterprise software company VMWare (VMW) in a deal that was originally valued at $61 billion . In the deal announcement, Broadcom touted VMware, saying that its "multi-cloud portfolio, spanning application modernization, cloud management, cloud infrastructure, networking, security, and anywhere workspaces, forms a flexible, consistent digital foundation on which the largest and most dynamic enterprises across industries build, run, manage, connect and protect their most important and complex workloads for the benefit of their customers." One big partnership that VMware has that we believe will be particularly lucrative is with Nvidia, the kingmaker of AI. The two companies announced Tuesday an expansion to their strategic partnership "to ready the hundreds of thousands of enterprises that run on VMware's cloud infrastructure for the era of generative AI." The new offering from that alliance, called the VMware Private AI Foundation, will enable enterprises to customize models and run gen AI applications like chatbots, assistants, search and summarization. Additionally, Nvidia announced AI-ready servers to support the VMware Private AI Foundation and help businesses customize and deploy generative AI applications using their proprietary data. On last Wednesday evening's earnings call, Nvidia's Huang called VMware "the operating system of the world's enterprise," adding that with the software company "we're going to bring generative AI to the world's enterprises all the way out to the edge." We think this an emerging opportunity for VMware, and therefore Broadcom, which has not been fully priced into these stocks despite their year-to-date runs. There is still, of course, one hurdle that Broadcom and VMware must get through to take this deal across the finish line. The holdup isn't from British regulators who have already approved the deal or the U.S. Federal Trade Commission, which so far hasn't publicly commented, letting the mandated premerger waiting period expire. The issue is in China, the last major regulator required to weigh in. Recent history suggests this could be an issue. A few weeks ago, Intel (INTC) dropped its $5.4 billion acquisition of Tower Semiconductor (TSEM) after Chinese regulators failed to approve the deal in a timely fashion. If the VMWare deal were to break down, we would look to take advantage of any potential volatility by buying VMware stock and/or more Broadcom shares, which we would expect to announce a massive share repurchase program to make up for the deal. We are initiating our new Broadcom position with a $1,000 price target. But, that could be tweaked next week after we get a fresh view of the company when it reports earnings after the closing bell on Aug. 31. AVGO YTD mountain Broadcom YTD performance It's not in our nature to buy stocks around their all-time highs. Broadcom is near last month's record levels. We also rarely buy immediately ahead of a quarter. That said, our plan is to offset this purchase by selling and exiting Advanced Micro Devices . Shares of AMD have also surged higher in this year's AI trade, though the role the company plays in the field is less clear. We cannot make the AMD sale Thursday because of our trading restrictions. But we are downgrading our AMD to a 3 rating to reflect our plan to sell the stock and lock in the rest of our big gains as soon as we are cleared. As we have discussed before, Advanced Micro Devices has become a real quandary. We've stuck with this name this year, betting we'd see a bottom in the PC cycle. That has, indeed, played out as expected with the inventory finally cleaned out. Unfortunately for AMD, the company makes CPUs and the new world will be dominated by Nvidia's new Grace Hopper Superchip, which combines its Hopper GPU with an ARM-based Grace CPU. To be fair, AMD should see a ramp of its MI300 integrated data center CPU and GPU later this year, and that could be a good opportunity for the company. However, we prefer to stick with the chip system that offers the best performance. That's Nvidia. And, since we do not want to lose our exposure to semiconductors that benefit from the billions of dollars being invested in this emerging technology, we like the idea of swapping out AMD (again, when we are not restricted) for a company like Broadcom that benefits from a different part of the AI ecosystem. AMD YTD mountain Advanced Micro Devices YTD performance Yes, Broadcom may be around all-time highs, but it's also trading at half the valuation of AMD: 20 times forward earnings for AVGO versus a 40 times valuation for AMD. Broadcom also sports a 2% annual dividend yield. The company is also unique to both AMD and NVDA in that it works with hyperscalers to develop their own custom in-house chips. Alphabet's TPUs (tensor processing units) , for example, are designed with the help of Broadcom. That should help us increase diversification within the semiconductor industry without giving up exposure to the shifting design of data centers in general. (Jim Cramer's Charitable Trust is long AVGO, AMD, NVDA, GOOGL, MSFT. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Eduardo MunozAlvarez | View Press | Getty Images
"Exit" - Google News
August 24, 2023 at 10:14PM
https://ift.tt/uft1r9S
We're buying a new chip stock and looking to exit another to expand our AI exposure - CNBC
"Exit" - Google News
https://ift.tt/MdbATt8
https://ift.tt/dkMvXBn
Bagikan Berita Ini
0 Response to "We're buying a new chip stock and looking to exit another to expand our AI exposure - CNBC"
Post a Comment