By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panelists agreed that AMD's strong performance is tied to the 'agentic AI-driven server CPU renaissance,' but they expressed concerns about valuation, competition, and supply constraints, particularly high-bandwidth memory (HBM3e).
Risk: Supply constraints of high-bandwidth memory (HBM3e) could compress AMD's EPYC margins and allow Nvidia to capture pricing power.
Opportunity: The potential growth in agentic AI workloads driving demand for AMD's EPYC architecture.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Columbia Threadneedle Investments, an investment management company, released its "Columbia Seligman Global Technology Fund" second quarter 2026 investor letter. A copy of the letter can be downloaded here. During the quarter, the Fund's Institutional Class shares returned 50.34%, outperforming the MSCI World Information Technology Index's 33.65% gain. Stock selection in semiconductors, technology hardware and software, together with an off-benchmark electrical equipment allocation, supported relative performance, while exposure to financials, consumer discretionary and healthcare detracted. Technology stocks rallied as concerns over the Iran conflict eased and AI infrastructure spending boosted demand for semiconductors, memory, networking, servers and power solutions. The Fund expects AI and data-centre investment to remain strong, supported by broadening earnings growth and improving software bookings, cloud consumption and customer spending. However, geopolitical uncertainty, higher interest rates and heavy AI investment could pressure valuations and free cash flow. The strategy holds 50–75 technology companies across market capitalisations and uses bottom-up GARP research to identify misunderstood and undervalued businesses in the technology industry. In addition, please check the Fund's top five holdings to know its best picks in 2026.
In its second-quarter 2026 investor letter, Columbia Seligman Global Technology Fund highlighted Advanced Micro Devices, Inc. (NASDAQ:AMD). Advanced Micro Devices, Inc. (NASDAQ:AMD) is a leading semiconductor company that designs and manufactures of AI accelerators, microprocessors, and graphics processing units. On July 27, 2026, Advanced Micro Devices, Inc. (NASDAQ:AMD) closed at $494.95 per share. One-month return of Advanced Micro Devices, Inc. (NASDAQ:AMD) was -14.80% and its shares gained 178.94% over the past 52 weeks. Advanced Micro Devices, Inc. (NASDAQ:AMD) has a market capitalization of about $807.07 billion.
Columbia Seligman Global Technology Fund stated the following regarding Advanced Micro Devices, Inc. (NASDAQ:AMD) in its Q2 2026 investor letter:
"The fund did not own some of the larger central processing unit (CPU) companies such as
Advanced Micro Devices, Inc.(NASDAQ:AMD) and the server CPU opportunity is having a renaissance driven by agentic AI. This lack of exposure ultimately detracted from performance, as the companies' share prices moved sharply higher during the quarter."
Advanced Micro Devices, Inc. (NASDAQ:AMD) ranks 20 on our list of 40 Most Popular Stocks Among Hedge Funds Heading Into 2026. According to our database, 134 hedge fund portfolios held Advanced Micro Devices, Inc. (NASDAQ:AMD) at the end of the first quarter, up from 132 in the previous quarter. While we acknowledge the potential of Advanced Micro Devices, Inc. (NASDAQ:AMD) as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you're looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock.
Four leading AI models discuss this article
"The article contains an apparent factual error on fund ownership that undermines its narrative, leaving AMD's valuation looking full at current levels despite the AI tailwind."
The article's claim that the Columbia Seligman Global Technology Fund 'did not own' AMD yet was hurt by its rally is internally inconsistent and likely a reporting error; the fund's 50.34% return significantly outperformed the MSCI World IT Index's 33.65%, so missing a strong performer would have helped, not detracted. AMD's 179% 52-week gain to $495 (market cap $807B) reflects the server CPU renaissance from agentic AI, but at ~55x forward P/E on 2026 consensus EPS, the stock prices in multiple years of 25-30% growth. Geopolitical risks, potential AI capex digestion, and higher rates cited by the fund itself remain material headwinds.
If the fund truly had zero exposure, the outperformance implies even stronger stock selection elsewhere; AMD's surge could prove unsustainable if agentic AI hype cools or if Nvidia's CUDA moat keeps limiting AMD's data-center share below 15-20%.
"The transition to agentic AI shifts the bottleneck from raw GPU training throughput to CPU-heavy orchestration, favoring AMD’s EPYC server architecture, but the current valuation leaves zero margin for execution error."
The 'renaissance' in server CPUs driven by agentic AI is a critical narrative shift. While GPUs handle training, agentic AI—which requires autonomous, multi-step reasoning—is proving CPU-intensive for inference and orchestration. AMD's EPYC architecture is uniquely positioned to capture this shift, yet a $807 billion market cap implies significant growth is already priced in. A 178% 52-week run-up suggests the stock is vulnerable to a 'sell the news' correction if enterprise software bookings don't translate into immediate hardware demand. Investors are betting on a massive refresh cycle, but the high interest rate environment mentioned in the letter remains a structural headwind for capital-intensive data center expansion.
The thesis relies on the assumption that agentic AI will remain CPU-bound, ignoring the possibility that specialized NPUs or future GPU architectures could consolidate these workloads, rendering the 'CPU renaissance' a temporary bridge rather than a long-term growth engine.
"The article conflates a fund's FOMO miss with a validated investment thesis—AMD's agentic AI CPU opportunity is real but unquantified, and at 178% YTD gains, most upside may already be priced in."
This article is mostly a fund performance recap masquerading as AMD analysis. The real signal: Columbia Seligman *missed* AMD's 179% YTD run by not owning it, then rationalized the miss by claiming 'other AI stocks offer greater upside.' That's post-hoc justification, not analysis. AMD's server CPU renaissance thesis (agentic AI driving CPU demand) is credible—but the article provides zero specifics: no TAM expansion numbers, no competitive win rates vs. Intel/Nvidia, no margin trajectory. At $807B market cap with 50.34% fund returns, we're in momentum territory. The real question: does agentic AI actually require more CPU compute, or is this narrative retrofitting past gains?
AMD's 178% gain in 52 weeks already prices in most of the agentic AI upside; the fund's underperformance vs. the index (50.34% vs. 33.65%) suggests they caught the tail, not the thesis. If agentic AI adoption slows or proves GPU-bound instead of CPU-bound, AMD faces a sharp multiple compression from current valuations.
"Valuations for AMD may be ahead of durable earnings because AI demand is currently GPU-led and AMD's data-center CPU share remains modest."
The article frames AMD as a primary beneficiary of an 'agentic AI-driven server CPU renaissance,' backed by fund chatter and a sharp quarterly rally. The strongest caveat is that AI demand in servers is far from AMD-specific: AI workloads skew GPU-heavy (NVIDIA) and data-center CPU share for AMD remains modest versus Intel, limiting the incremental earnings leg. The rally could be front-running a multi-year capex cycle that depends on macro stability (rates, enterprise budget) and software monetization, not just hardware procurement. Also, valuations may already assume durable AI upside, leaving little cushion if tariffs, onshoring risks, or supply-costs intensify and cash flow falters.
The strongest counter is that AI server demand is largely GPU-driven and may not translate into proportional CPU upside for AMD; if Nvidia remains the AI accelerator winner, AMD's long-run growth and margins could disappoint.
"Outperformance despite zero AMD holding suggests stock selection skill, not rationalization, but socket-share data remains missing."
Claude's post-hoc justification claim overlooks that the fund's 50.34% return beat the MSCI World IT Index by 16.7 points despite 'missing' AMD. This implies superior picks elsewhere, not underperformance. The real gap: no panelist has quantified AMD's EPYC share in AI server sockets versus Intel's Xeon, critical if agentic workloads stay under 20% of total inference demand.
"The 'CPU renaissance' is constrained by HBM3e supply bottlenecks, not just socket share gains against Intel."
Grok, your focus on EPYC share misses the real bottleneck: memory bandwidth. Agentic AI requires massive inference throughput where AMD's Infinity Fabric architecture often outperforms Intel's Xeon, but high-bandwidth memory (HBM3e) supply constraints are the actual limiting factor for revenue, not just socket share. Claude is right to call out the lack of TAM numbers; without quantifying the HBM-to-socket ratio for agentic workloads, the 'CPU renaissance' remains speculative narrative rather than a valuation-backed thesis.
"AMD's CPU upside is real, but HBM supply constraints could lock margins below consensus while Nvidia extracts rent as the memory gatekeeper."
Gemini's HBM3e supply constraint angle is sharper than socket-share debates, but it inverts the risk. If HBM supply is the bottleneck, AMD's EPYC margins compress while Nvidia (which controls HBM allocation via GPU bundles) captures pricing power. This means AMD's 55x forward P/E assumes supply relief that may not materialize. Nobody's priced in the scenario where AMD gains sockets but loses margin to memory scarcity.
"Quantifying TAM alone won't justify AMD's high valuation; timing risk and the migration curve of AI workloads to CPUs vs GPUs matter as much, if not more."
Claude's critique of missing TAM is fair, but it shifts the conversation toward momentum without addressing timing risk. Even with a CPU-centric AI synergy, hyperscale buyers still want a balanced rack—GPUs, firmware, and software monetization—so rollout pace matters. If AI adoption slows or GPU-led architectures regain pricing power, AMD's elevated multiple could re-rate quickly. Quantifying TAM is only half the battle; you must map the migration curve, pricing, and capex cadence.
The panelists agreed that AMD's strong performance is tied to the 'agentic AI-driven server CPU renaissance,' but they expressed concerns about valuation, competition, and supply constraints, particularly high-bandwidth memory (HBM3e).
The potential growth in agentic AI workloads driving demand for AMD's EPYC architecture.
Supply constraints of high-bandwidth memory (HBM3e) could compress AMD's EPYC margins and allow Nvidia to capture pricing power.