AMD Β· Advanced Micro Devices, Inc.Advanced Micro Devices, Inc.
π AI hyperscaler demand is causing widespread supply chain bottlenecks across semiconductors, materials, and infrastructure components, benefiting companies exposed to CPUs, semiconductor equipment, photoresists, grid components, optical modules, and photonics
AI hyperscaler demand is causing widespread supply chain bottlenecks across semiconductors, materials, and infrastructure components, benefiting companies exposed to CPUs, semiconductor equipment, photoresists, grid components, optical modules, and photonics.
Evidence & details
- +Hyperscalers may build their own datacenters and drop contracts, akin to Apple dropping Qualcomm.
- +GPUs depreciate over 4 years, affecting full-stack providers.
- +Credit tightening could force dilutive raises or high-interest debt.
- +LLM software breakthroughs could reduce GPU utilization, crashing demand.
- β‘Hyperscaler vertical integration
- β‘GPU depreciation curve
- β‘Credit market conditions
- β‘LLM efficiency breakthroughs
- βCustomer loss to hyperscaler self-build
- βGPU value erosion
- βDilution or expensive debt
- βOvercapacity from software improvements
Chart
Post Timeline Β· 1 posts
@LoganMacGregor9 Good question about risk, I enjoy talking about the downsides too. 1. Hyperscalers completely their own datacenter buildout with GPUs (TPUs with $GOOGL, Tranium with $AWS), energy, locations in 5+ years. This is the analogy of Qualcomm with Apple. Apple just uβ¦
Thesis: Lists four main risks for NBIS and other AI infrastructure companies: hyperscaler vertical integration, GPU depreciation, credit tightening/dilution, and overbuildout due to software efficiency breaktβ¦
AMD is undervalued compared to NVDA, with potential for significant market cap growth as it is seen as the next NVDA.
source βRotating into high-growth positions: AMD (forward revenue doubling), FLY (neocloud similar to RocketLab), WLAC (neocloud IPO with high margins), MU (memory demand from Stargate/OpenAI), FLNC/SEI/DFLI (energy and battery plays, high risk high reward). Trimming laggards for tax harvesting, will repurchase later.
source βAMD's 8.2% drop due to a tweet is a buy opportunity as macro fears are overblown.
source βAMD is bullish because as a fabless company it is less impacted by rare earth export controls, and the current selloff resembles April when a trade deal resolved the downturn.
source βAMD is a buy on a 1-year horizon due to a $100B contract with OpenAI starting to ramp. MU is affected by China disputes due to its large DRAM/NAND exposure and is advised to hold. Overall macro risk from Trump's unpredictability tempers conviction.
source βBullish on RDDT at $192, AMD at $210, VIRT at $33. VIRT seen as undervalued hedge with upcoming earnings.
source βRaising 1-year price target on NBIS to $450 and reiterating outperform on other stocks due to infinite energy consumption in the Neocloud space.
source βThe author presents a broad bullish thesis on multiple growth stocks, particularly in AI/data center buildout (neoclouds), energy, and cybersecurity, citing catalysts such as rate cuts, government reopening, and seasonal trends. Key themes include the undervaluation of certain tech stocks after corrections, the importance of TSM as a central supplier, and the potential for re-rating in sectors like nuclear and memory.
source βAuthor presents a comprehensive trading plan for rate cut week, recommending longs in neocloud, semiconductors, and select other stocks while selling overvalued quantum, space, and crypto assets. Key macro view: market fear is an ideal entry point ahead of expected rate cut.
source βAMD has potential to reach NVDA's current market cap if frontier models use AMD effectively, as suggested by Sam Altman and Elon Musk. However, OpenAI's revenue promises to AMD are uncertain, and position sizes are small relative to TSM.
source βAMD has extremely high upside potential based on NVDA's market cap and endorsements from Sam Altman and Elon Musk regarding utility for frontier models. HOOD is a generational company with high conviction long-term, though short-term overvalued.
source βAI capex is exponentially ramping up and flowing directly to neoclouds, connectivity, energy, semi/foundries, and memory stocks, with multiple recent catalysts confirming record growth with no slowdown in sight.
source βReduced bullishness on AMD due to weakening OpenAI dominance; also bearish on OpenAI-dependent stocks (CRWV, ORCL, MSFT).
source βOpenAI's competitive position is deteriorating (Gemini, Claude surpass it), leading to contagion in companies that built capacity for OpenAI's capex (ORCL, CRWV, AMD). However, for companies not materially impacted by OpenAI's specific promises, this creates a buying opportunity.
source βElon Musk's $134B lawsuit against OpenAI poses downside risk to AI stocks levered to OpenAI (AMD, CRWV, ORCL, MSFT, APLD, CORZ, RIOT) in worst-case scenario, but author expects middle ground with a large fine and MSFT stepping in, limiting contagion.
source βTrump administration proposal for AI chip export permits is unlikely to pass; current selloff is overreaction.
source βUS Commerce Department denies return to AI diffusion rule, countering negative reporting that caused sell-off in semis, supporting NVDA and AMD.
source βReaffirms long-term bullish thesis on a basket of semiconductor and related stocks, noting that short-term entry points vary but overall market validation continues.
source βCPU supply shortage driving price hikes and capex increases, benefiting CPU makers and foundry.
source βAI hyperscaler demand is causing widespread supply chain bottlenecks across semiconductors, materials, and infrastructure components, benefiting companies exposed to CPUs, semiconductor equipment, photoresists, grid components, optical modules, and photonics.
source β