Accumulate shares of Meta Platforms (META) as it currently trades as the cheapest Magnificent 7 stock, sitting roughly 20% cheaper than NVIDIA (NVDA).

Take advantage of NVIDIA’s favorable valuation near the $200 price level, as its fabless business model allows for rapid scaling during the ongoing AI boom.

Target Micron Technology (MU) around the $820 level as a top hardware pick, supported by an ultra-low P/E ratio of 4.5 to 5 and a shift away from pure cyclicality.

Invest in Nebius (NBIS) while it trades under the $200 threshold, capitalizing on its temporary valuation dip and multi-billion-dollar enterprise contracts.

Avoid overvalued AI hardware plays like Advanced Micro Devices (AMD) and Broadcom (AVGO) until their valuation multiples drop to align more closely with industry peers.

NVIDIA (NVDA)

• Considered to be very cheap at the $200 price level, ranking as the second cheapest among the Magnificent 7 stocks based on a modification of the PEG ratio (enterprise value over gross profit over revenue growth).
• Operates in a fabless business model, meaning it designs chips and operates in the world of bits rather than physical manufacturing, allowing for much faster scaling and growth compared to physical foundries.
• Highlighted as one of the top hardware picks alongside Micron due to explosive AI demand and hardware needs.

Takeaways

• Consider accumulating shares while the valuation remains favorable relative to other mega-cap tech stocks.

Amazon (AMZN)

• Recently saw a 15% stock price increase following an earnings call where CEO Andy Jassy highlighted crazy demand for AWS extending into 2028.
• Ranks as the third cheapest among the Magnificent 7 on the speaker's comparative valuation metric with a score of 0.54, placing it in the middle tier of being neither overly cheap nor expensive.
• Successfully announced an increase in capital expenditures (CapEx) while still seeing its stock price rise, indicating that the market is comfortable with its heavy AI investments.

Takeaways

• Strong long-term demand trends suggest Amazon remains a solid infrastructure player in the AI boom.

Meta Platforms (META)

• Identified as the cheapest Magnificent 7 stock, trading roughly 20% cheaper than NVIDIA following recent market drops.
• Moving toward becoming a hyperscaler, with CEO Mark Zuckerberg indicating openness to potentially selling excess compute capacity.

Takeaways

• Represents a high-value opportunity within the mega-cap tech space for investors looking for relative value in AI infrastructure.

Advanced Micro Devices (AMD)

• Viewed as significantly too expensive, noted as being more than twice as expensive as NVIDIA on a relative valuation basis.
• Suffered from a rapid run-up in stock price that outpaced its fundamental valuation appeal.

Takeaways

• The speaker suggests avoiding or eliminating AMD from consideration until its valuation drops to align more closely with industry peers.

Super Micro Computer (SMCI)

• Extremely cheap in the hardware space with a valuation score of 0.22.
• Plagued by drama and negative sentiment surrounding alleged China ties and regulatory concerns, which artificially depresses its market price.

Takeaways

• While statistically cheap, investors must weigh the heavy headline risk and dramatic market sentiment associated with the company.

Micron Technology (MU)

• Highlighted as extraordinarily cheap, trading at a valuation score of 0.12 and a low price-to-earnings (PE) ratio of roughly 4.5 to 5.
• Historically viewed by Wall Street as a cyclical business, but this narrative may be shifting as Micron secures 17 five-year long-term contracts featuring upfront payments and cancellation fees.
• Memory is argued to be evolving into a cornerstone of AI alongside GPUs, potentially positioning Micron to eventually reach a multi-trillion-dollar valuation.
• Traded down recently to around $820 (with previous buyer interest noted around $600 to $700), though the speaker fears it may quickly bounce back up to $1,000.

Takeaways

• Considered one of the top hardware picks due to its ultra-low valuation and changing business dynamics away from pure cyclicality.

Broadcom (AVGO)

• Praised for strong management and its position as the king of custom ASIC chips, but evaluated with a valuation score of 0.66.
• Dismissed as too expensive relative to its growth prospects, putting it in a similar category to AMD.

Takeaways

• Easily eliminated from a value-focused AI portfolio due to stretched valuation multiples.

Taiwan Semiconductor Manufacturing Company (TSMC)

• Acts as the world's largest chip foundry, meaning it benefits regardless of which specific company wins the AI race.
• Considered too slow-growing for the AI era, projected at 33% revenue growth over the next 12 months, which falls short of the speaker's preference for stocks capable of doubling revenues in a year.
• Bound by physical constraints ("moving at the pace of atoms" in the brick-and-mortar world), making it harder to scale as rapidly as fabless designers like NVIDIA.

Takeaways

• Excluded from top picks because physical manufacturing bottlenecks limit its hyper-growth potential compared to software or fabless hardware peers.

Nebius (NBIS)

• Identified as the speaker's favorite neo-cloud pick, specializing in pure-play GPU data centers and cloud infrastructure.
• Secured major multi-billion-dollar backlogs and contracts with major players like Microsoft and Facebook.
• Valuation dipped temporarily to a score of 0.16 during a recent market drop, with the stock trading under $200.
• Risk factors mentioned include the potential burden of debt required to fund heavy capital expenditures.

Takeaways

• Viewed as a strong investment under $200, with future revenue growth heavily supported by secured enterprise contracts.

CoreWeave (CRWV)

• Operates as a neo-cloud GPU provider similar to Nebius, leaning more toward bare-metal offerings alongside some software solutions.
• Noted as slightly more expensive than Nebius and carrying higher debt to fund necessary infrastructure investments.

Takeaways

• Requires market comfort with heavy CapEx and debt loads, though it benefits directly from surging global demand for compute power.