AMD can monetize AI infrastructure through both accelerators and a widening server-CPU opportunity, while customer deployments and software support reduce adoption friction.
AMD’s growth case remains exposed to constrained supply, uncertain GPU economics and conversion, and a valuation that may already discount execution.
Downstream companies like AMD could face more concentrated laser/DSP supply due to the potential LITE/MRVL deal.
The company's stock increased due to positive checks across the board, indicating better MI450 builds and CPU supply in the second half of 2026, with a likely CPU price increase in Q1.
The company committed to investing in META's equity offering and confirmed its first Helios racks are installed and being tested at META, which also shared a release on training production recommendation models using the company's GPUs.
Meta has deployed millions of the company's EPYC processors and is a lead customer for 6th Gen EPYC CPUs, with Instinct GPUs in production for training, ranking, and recommendation.
The company provided Day-0 support for Muse Spark Open on Instinct GPUs and Muse Glimmer on Ryzen AI PCs and Radeon GPUs, and expanded software collaboration with Meta for large-scale AI training, optimizing recommendation models and contributing to open-source AI frameworks.
Evercore's checks with AI engineers indicated that agentic fan-out is pushing server CPU:GPU ratios towards 1:1, leading to increased CPU ASPs.
AMD +3.4% as Evercore’s checks with 20+ AI engineers found agentic fan-out pushing server CPU:GPU ratios from one per 2-4 GPUs toward 1:1, with CPU ASPs up 10-15% q/q in 1H26 and seen up 15-20% in 2H26 (one expert models 20-25% through 2027 until CoWoS ease...
Clev is positive, stating that CPU demand remains strong with mid-1M mi4xx GPU shipments in CY27.
SemiAnalysis argues 2027 data-center growth is materially underestimated, forecasting ~50GW of new capacity in a single year with >24GW already in advanced construction across just 40 campuses and ~60GW under construction more broadly. The key leading indic...
The company's China-developed open-source TileRT demonstrated superior performance compared to NVDA in extreme high-end GLM 5.2/5.3 inference, as highlighted at OCP.
JPM sees networking as an increasingly critical gating factor in AI data center infrastructure, which favors AMD alongside other compute and foundry beneficiaries.
Microsoft's Scott Guthrie expects prolonged storage shortages and intense model price competition, but also indicates that Microsoft will continue to purchase chips from Nvidia, AMD, and Intel as its own Maia program expands, suggesting continued investment in AI infrastructure.
The company has significantly increased its CPU Total Addressable Market (TAM) estimate, now exceeding $200 billion by 2030, driven by the new workload of agent execution in AI data centers, which requires substantial CPU processing alongside GPUs.
Helios has transitioned from launch to initial customer deployment, with production shipments starting in September and installations at major hyperscalers like OpenAI, Meta, Microsoft, and Oracle, indicating a meaningful execution milestone.
The new Helion rack specifically targets agentic execution with CPU-heavy configurations for hyperscalers and integrated CPU/GPU configurations for enterprises, representing a distinct and deliberate second deployment architecture.
Verano, a Venice-family EPYC variant, is optimized for lower system power and features an LPDDR-based memory approach and updated CPU-to-GPU connectivity, demonstrating AMD's strategy to address diverse agent-related CPU workloads beyond just maximizing core count.
AMD is addressing a larger CPU opportunity while simultaneously advancing its GPU proposition towards a complete deployable system, indicating a comprehensive strategy for the AI data center market.
While Helios production shipments have begun, initial installations are not equivalent to fulfilling previously announced multi-gigawatt commitments, and the pace of conversion to large-scale deployment remains an open question.
The company's claim of a 30% tokens-per-dollar advantage for Helios lacked sufficient comparison detail to extrapolate this benefit across all workloads, making it difficult to fully assess the economic advantage.
Neither the TAM forecast nor a vendor benchmark definitively establishes AMD's eventual revenue share or margin contribution in the evolving market.
The company is preparing price increases on some CPUs due to advanced foundry and packaging costs.
INTC -3.2%; AMD -2.3%; QCOM -61bps got no help from Commercial Times reporting CPU price hikes spreading: INTC +15-20% on high-end Panther Lake, QCOM ~10% on Snapdragon X, AMD preparing its own.
Checks were positive, with some products shipping ahead of schedule.
The shift from GPU-centric clusters toward heterogeneous GPU/CPU/ASIC/wafer-scale systems and open rack-scale standards is positive for the company.
Citi expects OCP to highlight a shift from GPU-centric clusters toward heterogeneous GPU/CPU/ASIC/wafer-scale systems and open rack-scale standards, positive for AMD/MRVL/AVGO. Power should center on 800V DC adoption, favoring TXN/MPWR/AEIS/Delta, while con...
Evercore says ANET’s new Etherlink architecture extends Ethernet into scale-up, supporting 144 XPUs single-hop and 1,024 cross-rack, while next-gen optics lift capacity ~4x to 6.5 Pb/s/rack and cut DC footprint ~50%. Flexible designs spanning AMD/Arm/Broadc...
The other semis that were hitting 52wk highs this week: LITE and TSM charts also look similar dipping below previous resistance after all-time highs, but AMD still holding its breakout.
The company is actively exploring foundry and memory partnerships with Samsung to secure supply.
AMD has extended joint supply-chain planning with customers to as much as five years, coordinating advanced wafers, packaging, and memory.
The company plans to maintain its asset-light model while needing more advanced wafer capacity in the coming years.
Reuters reports Microsoft and Nvidia will unveil the Surface Laptop Ultra on Wednesday, using Nvidia’s RTX Spark chips to run AI agents and coding workloads locally. The launch extends Nvidia into a PC market dominated by Intel and AMD. Pricing is a key iss...
The company faces challenges in securing high-layer-count ABF substrate capacity, which is the tightest part of the market with equipment lead times exceeding 12 months, potentially leading to significantly higher substrate pricing in 2027.
Kinsus, a key supplier, views NVIDIA's CPU as its primary growth driver into 2026-27, indicating potential market share gains for NVIDIA against AMD and Intel.
AMD -34bps with Lisa Su in Taiwan hunting packaging and substrate capacity. Good color on that bottleneck from Terminal Junkie's Kinsus meeting in Slack here: high-layer-count ABF is the tightest part of the market, equipment lead times are >12 months,...
The company plans to substantially increase supply in 2027, including securing memory, which is a positive indicator for future growth.
Analysts are projecting significant growth in the CPU TAM, with Citi estimating ~$300 billion by 2030 and Mizuho estimating ~$209 billion.
The company gained market share in the Q3 laptop model-share tracker, increasing by 3.1 percentage points to 29.3% at Intel's expense.
AMD +3.6% kept going after Lisa Su told Reuters AMD plans to "substantially increase" supply in 2027, including securing memory, with Citi (CPU TAM to ~$300B by 2030) and Mizuho (~$209B) out with agentic-CPU math, and Susquehanna's Q3 laptop model-share tra...
The company is actively seeking packaging and substrate capacity, indicating a bottleneck in its supply chain.
High-layer-count ABF substrates are the tightest part of the market, with equipment lead times exceeding 12 months, and a base case of ~30% higher 2027 substrate pricing due to materials and capacity-driven hikes.
Kinsus, a key supplier, views Nvidia's CPU as its primary growth driver into 2026-2027, with potential share gains against AMD and Intel.
AMD -34bps with Lisa Su in Taiwan hunting packaging and substrate capacity. Good color on that bottleneck from Terminal Junkie's Kinsus meeting in Slack here: high-layer-count ABF is the tightest part of the market, equipment lead times are >12 months,...
The company's 2027 EPYC 9006 Venice capacity is sold out, with orders extending into 2028, indicating strong demand for its server CPUs.
AI agents are increasing orchestration and execution workloads, leading to tighter supply and higher demand for server CPUs.
The company's 2027 Venice EPYC capacity is reportedly sold out, with orders now being taken for 2028, indicating strong demand driven by AI agents pushing server CPU lead times higher.
The acquisition of World Labs is strategically more important than its near-term financial contribution, bringing Dr. Fei-Fei Li as Chief Scientist and giving AMD direct insight into frontier workloads to shape future compute architectures.
World Labs expands AMD beyond LLMs into spatial intelligence, simulation, robotics, and physical AI, while strengthening its open-source/open-weight positioning.
The deal narrows AMD’s strategic gap with NVDA over time, and StoneX reiterates a Buy rating with a $685 price target.
Stifel says AMD is effectively buying a seat at the model frontier, with World Labs adding spatial/world-model expertise for 3D simulation, robotics, and physical AI, and Dr. Fei-Fei Li joining as Chief Scientist.
The strategic value is tighter model-to-hardware/software co-design and a more credible open-model alternative to NVDA’s Cosmos stack, rather than near-term P&L contribution.
Stifel also sees World Labs’ simulation capabilities helping address the real-world training-data bottleneck in robotics, and maintains a Buy rating.
The company's stock significantly outperformed peers over the past month.
The company's stock was one of only three to be green over the past month among SOXX or SMH components, indicating relative strength.
The World Labs deal is not a near-term moat-closing event against NVDA.
BofA expects HPE’s Sept. 30 Networking Investor Day to update the TAM, Juniper integration and AMD Helios opportunity, with networking revenue capable of a 15– 20% CAGR through 2030. Helios could scale from a few hundred million of switch revenue to several...
The analyst notes that Akamai's cloud already utilizes AMD EPYC processors, with the current top tier being 5th-gen EPYC and the previous being 3rd-gen, indicating AMD's existing presence in Akamai's infrastructure.
AKAM’s cloud already runs on AMD EPYC. The current top tier is 5th-gen EPYC and the one before it is 3rd-gen. I found no Intel or Arm servers anywhere in their lineup. The dollars are modest for AMD. My rough math is about $1–1.4B of CPUs over 2027– 28.
The stock reversed sharply from morning lows, shrugging off negative shipment data, and is benefiting from increased pricing for its EPYC CPUs.
The AKAM deal with Anthropic, which uses AMD EPYC, indicates significant demand for CPU capacity across the board.
AKAM’s cloud already runs on AMD EPYC. The current top tier is 5th-gen EPYC and the one before it is 3rd-gen. I found no Intel or Arm servers anywhere in their lineup. The dollars are modest for AMD. My rough math is about $1–1.4B of CPUs over 2027–28.
AMD +2.4% / INTC +3.9% reversed hard off the morning lows (INTC closed at the highs), shrugging off the M-sci shipment datapoint on AMD we flagged this morning. KeyBanc’s bus-tour note helped INTC (14A “~1 year ahead of where 18A was”), and NBIS told custom...
New CPU shipments were significantly lower in the first half of September compared to the first half of August, indicating a negative year-over-year inflection.
The stock is a thematic winner in AI infrastructure, specifically within CPUs, and is trading near highs.
In semis, it’s more nuanced than just AI winners vs. losers because most names are perceived winners, but the outperformance since the Leopold lows has been centered on CPUs and on-prem/enterprise. The SOX is ~14% below its June highs, yet just a few names...
Conversations with various companies consistently indicated strong AI infrastructure demand despite ongoing debates about AI regulation.
Memory and storage demand continues to exceed supply.
Agentic CPU demand is expected to broaden from head nodes and traditional servers in 2027 to sandbox workloads in 2028, benefiting next-generation Vera, Venice, and Diamond Rapids CPUs.
CXL is gaining traction, with CXL 2.0 ramping at Meta and Microsoft, and CXL 3.0 being crucial for broader adoption.
BNP Paribas says conversations across AMD, ANET, AMZN, AVGO, NVDA and others consistently pointed to AI infrastructure demand remaining strong despite calls for tighter AI regulation.
The stock saw a significant rally as investors increasingly connected consumer agentic harnesses like MUSE with the next wave of compute demand.
The company joined DELL/HPE as one of the few large-cap AI semi stocks to hit all-time highs, crossing the $1 trillion market capitalization, which confirmed to investors that parts of the AI trade are still viable.
The stock's chart looked strong, breaking out to a two-month high.
There are rumors of a 10% chip price increase due to higher TSMC wafer costs.
Muse responded that its VM is an AMD EPYC 9D25 with 2 vCPUs, indicating its use in advanced AI systems.
The stock continues to benefit from increased CPU usage, particularly with Muse.
AMD has secured sufficient leading-edge wafer, HBM, substrate, and packaging supply to meet or exceed current targets, with Helios entering volume production and expected to contribute significantly to AI revenue.
Data center revenue is projected to more than double in CY27, and server CPU revenue is expected to grow over 70% year-over-year as Venice ramps, with demand continuing to outstrip available supply.
Gross margins are expected to remain around 55% in CY27, but significant operating leverage from AI growth is anticipated.
Helios rack production is starting, with broader volume ramping into December/2027, and Venice server CPUs are expected to drive over 70% year-over-year CPU growth as AMD targets over 50% server share long-term.
AMD has secured sufficient wafer, HBM, and packaging supply to support guidance, with demand still exceeding availability, and the data center mix is driving higher gross profit dollars despite margins hovering around 55%.
The company's stock was helped by the agentic CPU bid and positive pre-quiet period calls with sell-side analysts, with Piper noting that Helios rack production is starting now and broader volume will ramp into December 2027.
Venice server CPUs are expected to drive over 70% year-over-year CPU growth, as the company targets over 50% server share in the longer term.
The company is expected to be a major incremental driver of CoWoS consumption in 2027, with a projected 165% y/y increase to ~345K wafers for MI450/455 and Venice products.
AI Semis/AMD/INTC: MS sees 2.5D packaging capacity +~50% in 2028 vs CSP capex +12%, reinforcing compute as the priority spend
Piper initiated coverage with an Overweight rating and a $600 price target, anticipating that agentic AI will boost CPU demand and that MI450 will drive GPU shipments to 1.5M in 2027.
Piper believes agentic AI will expand server CPU demand, alongside continued AMD share gains against Intel, adding another dimension to the AI compute story beyond GPUs.
Piper models approximately 1.5M AMD GPU shipments in 2027, roughly doubling year-over-year as the MI450 ramps up.
HPE -6.2% as IR chief Shannon Cross at Goldman said the FY26 networking-for- AI order target has gone from $1.5bn to $2.5-3bn, with the gigawatt Oracle win ‘a competitive bid’ won on the liquid-cooled Tomahawk 6 switch and talks under way with others. New:...
There are positive indications regarding CPU demand and pricing upside.
There is potential for upside in GPU performance.
AMD/INTC: Hearing Clev positive on CPU Demand and pricing upside; AMD GPU upside
The company benefited from a positive read-through to CPUs due to increased computer usage, suggesting higher demand for its processors.
The CFO and CEO expressed confidence in the MI450/Helios ramp, despite execution concerns, indicating strong internal belief in the product's success.
Demand and volume for 2027 are now exceeding AMD's original plans, with all three anchor customers (Meta, OpenAI, Anthropic) forecasting higher demand than expected.
Supply, rather than demand, is the primary constraint for the company, suggesting robust underlying demand for its products.
The company plans to pass wafer-cost increases through to customers, indicating an ability to maintain margins despite rising input costs.
The data center business is projected to double next year, and server CPU growth is expected to be over 80% in the second half and over 70% in 2027, highlighting significant growth in key segments.
Optical (LITE +11% / COHR +7%) — Getting the bid off AMD at Citi TMT: the company plans to offer both copper-based and near-package optical connectivity for larger scale-up domains on the 2H27 accelerator, with copper and optical likely coexisting for multi...
Gross margin in Q4 and through 2027 is expected to be slightly below Q3's 56% guide due to the MI450 ramp, although this is offset by server CPU and embedded mix.
Competition from the company's TPUs is a concern for bears.
AMD's Helios architecture demonstrates a much more mature open-Ethernet scale-up stack, which is positive for the company.
William Blair’s main Hot Chips takeaway is that differentiation is moving beyond raw FLOPS toward full-system architecture across compute, memory, networking, software and reliability. Nvidia still has the clear lead, but AMD’s Helios architecture shows a m...
The company offers the best combination of direct earnings leverage, datacenter positioning, and potential share gains in the server CPU market, which is expected to reach ~$201B by CY30, including ~$85B of agentic CPUs and ~$83B of AI head-end CPUs.
Agentic AI is creating a distinct CPU demand pool as agents increasingly handle orchestration, retrieval, databases, security, and tool execution around each model call.
RJ's upgrade suggests the server CPU market will reach approximately $201 billion by CY30, with significant portions from agentic and AI head-end CPUs.
The stock has fallen 21.9% from its high, indicating a substantial downturn.
JPM says recent concerns around Google custom silicon competition are misplaced: AVGO’s 5-year Google agreement locks in the next four TPU generations, while TPU v9 is now ramping and the next 2nm XPU ASIC remains on track for CY28 with no CoWoS cuts or del...
AMD is identified as a key ecosystem beneficiary in evolving inference architectures, particularly through its collaboration with Cerebras for disaggregated inference.
The partnership with Cerebras, where AMD Helios handles prefill for disaggregated inference, aims for significantly better tokens-per-watt efficiency compared to WSE-only systems.
CBRS/AMD/ANET: BNP Says Cerebras CS-4 Targets 30x GPU Inference Speed, with AMD and ANET Key Ecosystem Beneficiaries
The stock saw a positive move, potentially due to the company working with Google on a TPU project.
BNP expects CXL memory-expansion products from MRVL/ALAB to ramp in 1H27, with true memory pooling ramping later in 2027 as PCIe 6 CPUs such as AMD Venice and Intel Diamond Rapids reach volume.
AVGO: Semianalysis says AMD might be working with GOOG on TPU project and AVGO is seeing lower supply gated TPU demand (out Friday, but just flagging for the reason for the AMD +3% / AVGO -5%) move
Disaggregation. The AMD Helios partnership places GPUs on prefill and CBRS on decode, preserving CBRS’s speed while raising throughput by roughly 5x. The
AMD quietly acquired Taalas, a startup that permanently 'prints' AI models into chips, which is clever engineering.
UBS says AMD’s Taala acquisition adds another differentiated inference architecture by hard-coding model weights into silicon, potentially reducing external memory fetches and improving performance-per-watt versus conventional GPU or SRAM-based approaches.
The analyst believes AMD wants Taalas for niche uses, not for the big AI datacenter story being marketed.
The company acquired Taalas, a chip startup that hardwires AI models into its silicon, which promises less-expensive chips that can produce output for specific models thousands of times faster than traditional GPUs.
The stock experienced a decent bounce back after a decline on the previous day.
The last few days have dinged some AI favorite names which now look range bound near-term: WDC, SNDK, ALAB, AMD among others. Despite the LT bull cases remaining intact, this earnings season has seen a lot more dispersion and smaller beats than the early Ag...
The company is developing two independent AI growth engines: Instinct accelerators for the GPU market and EPYC processors for server share gains.
ANET has relatively limited scale-up exposure when NVIDIA supplies the full NVLink stack, but its addressable market expands with AMD MI accelerators, Google TPUs, custom hyperscaler silicon, inference accelerators and open Ethernet-based architectures. Bul...
Near-term results failed to exceed elevated expectations, and the company faces margin dilution from its AI business alongside rising operating expenses.
We’ll cover SPCX AMD ANET ALAB BKNG and SHOP below then onto the usual.
The analyst anticipates the CEO will express optimism regarding future product cycles in data center and GPU, and expects updates on the Helios and GPU ramp in Q4 and CY27.
Post-close ANET/ALAB delivering the goods while fell AMD short of bogeys. These comments from Elon have NVDA +1% in the post:
The company's Q2 and Q3 guidance fell short of elevated buy-side expectations, despite a beat-and-raise, indicating a miss relative to the high expectations for a widely held stock.
Benchmark reiterated a Buy rating and a $685 price target, expecting an essentially in-line 2Q but highlighting the September Helios ramp, gross margin, and evidence of disclosed AI deployments converting into revenue as important catalysts.
Benchmark models 3Q AI revenue of approximately $3.45B, up 27% quarter-over-quarter, driven by MI350 execution and initial MI455X/Helios shipments.
Improving HBM4 and packaging visibility, resilient EPYC demand, and the stock's 19% pullback create a more favorable risk-reward profile.
NVDA/AMD/AVGO: Jefferies Fubon Raises 2027 CoWoS Outlook as AI Packaging Demand Broadens
Jefferies Fubon expects TSMC CoWoS capacity to reach roughly 180kwpm by end- 2027 and 220kwpm by end-2028, with NVDA remaining the largest customer, AMD allocation rising toward 300k wafers in 2027 and AVGO/MediaTek also gaining capacity. Jefferies Fubon fo...
Mizuho reiterated Outperform and raised its price target, citing up to 14GW of MI455 Helios commitments from major tech companies, supporting a higher 2030 TAM estimate.
AMD and CBRS are developing a disaggregated inference platform with roughly 5x the throughput of WSE-3, with initial CBRS Cloud deployment expected in 2H26 and on-premise availability later.
AMD/CBRS: Mizuho Raises Targets as Helios Commitments and Fast-Inference Partnership Expand the AI Opportunity
The stock declined after outperforming leading into AI day, with some investors selling the news post-event, especially given INTC's poor price action after its earnings.
AMD -5% - stock had outperformed nicely heading into AI day but with that in rear-view and INTC’s ugly price action post EPS, some sell the news post the event.
Clev is positive on the company, noting robust CPU demand, positive GPU feedback, and supply upside.
The company is expected to see strong server CPU growth and is gaining market share as competitors face delays.
AI GPU revenue estimates have been raised due to a bullish supply-chain outlook and increased CoWoS allocations.
KeyBanc reiterated an Overweight rating and raised its price target, citing benefits from server CPU capacity supporting significant unit growth.
The company's server CPU capacity is expected to support substantial unit growth in the current and next year.
The MI455/Helios product remains on track for the second half of the year, indicating continued product development and market presence.
AI GPU revenue is projected to grow significantly in the coming years.
KeyBanc reiterates Overweight, raises estimates and lifts its price target to $155 from $110, citing outsized server CPU demand from agentic AI, with Intel capacity supporting 25%–30% unit growth this year and more than 50% next year. The firm says 18A yiel...
AMD/MU/MRVL: KeyBanc Raises Targets as AI Compute, Memory Pricing and Custom Silicon Visibility Improve
The company is positioned as a key technology and memory enabler in the AI data center market.
TD Cowen is neutral on ALAB, AMD, ANET, CBRS, CEVA, COHR, INTC and MTSI, cautious on QCOM, and sees PC/smartphone demand as weak due to persistent memory pricing pressure.
The stock experienced a decline.
Citi favors AMD due to strong AI infrastructure demand, which remains supply-constrained.
Goldman Sachs reiterates a Buy rating and raised its price target to $640 from $450, with 2027 EPS estimates moving to $14.50, approximately 13% above Street expectations.
Goldman Sachs expects a strong quarter and guide, driven by server CPU demand tied to agentic AI/infrastructure spending.
Commentary on MI450 timing, Meta/OpenAI deployments, and new customer engagements could potentially push 2027 datacenter GPU estimates higher.
The July 23 Advancing AI event is expected to be a positive catalyst.
The delay of NVDA's Kyber NVL144 is expected to be beneficial for AMD's competitive positioning with its MI series.
AVGO leading the way higher after renewing with AAPL through 2031, the usual suspects on a semi up day (HDDs/Memory/semicap) outperforming early, and AMD catching a bid on the NVDA Kyber delay news getting passed around.
Citi expects another strong semiconductor earnings season, arguing AI infrastructure demand remains supply-constrained despite recent concerns around excess compute capacity. The firm continues to favor AMD, TXN, and AMAT, sees DRAM shortages as the biggest...
The firm’s preferred tactical ideas are AMAT, AMD, and ON, citing upside from server CPU demand, AI infrastructure spending, and analog pricing, while remaining cautious on KLAC, ARM, and Qorvo due to tougher near-term setups. Goldman also expects continued...
The stock was a significant beneficiary of the Kyber delay news, as its MI500 product now competes against a dual-die Rubin Ultra in a legacy-style rack, rather than the quad-die Kyber.
The potential push of Kyber to 2028 means AMD's Mi500 "MegaPod" could arrive earlier, allowing AMD to offer a larger coherent scale-up domain.
The near-term narrative for AMD appears strong heading into its earnings print and AI day.
NVDA +37bps underperformed all day while AMD +6.6% seemingly the big winner from the news as MI500 now lands against a dual -die Rubin Ultra in a legacy-style rack instead of the quad-die Kyber.
Key risks include slower agentic AI adoption, weaker AMD GPU deployment at Meta, x86 share erosion, and limited operating leverage.
Wells Fargo raised its price target, arguing the market continues to underestimate the durability of AMD's server CPU and AI growth.
The firm sees accelerating EPYC demand, improving pricing, and growing adoption of high-core-count server CPUs driven by agentic AI and enterprise modernization.
Wells Fargo is raising CY27/CY28 EPS estimates, with increasing confidence in AMD's long-term path to $20+ EPS.
The firm highlighted a new order for automated wafer-level burn-in systems tied to datacenter transceiver deployments as evidence that industry participants are actively preparing for higher production volumes, while recent checks suggest AMD is acceleratin...
Bernstein argues that agentic AI is driving a meaningful upward revision to server CPU demand, benefiting both AMD and Intel as CPUs become increasingly important for orchestration, memory management, and agent workflows.
AMD remains the preferred beneficiary due to its continued product leadership and accelerating share gains, with Bernstein modeling strong EPS for CY27 and CY28.
AMD is working with Samsung on some chip projects due to capacity constraints at TSMC, and is in talks to build some future CPUs with Samsung from 2028.
(MU/SNDK +2-3%; INTC/AMD/ARM +2-3%).
The shift from chatbot AI to agentic AI fundamentally changes data-center architecture, making CPUs increasingly critical for orchestration, memory management, networking, and agent coordination.
The 2030 server CPU TAM forecast was raised significantly, with the prior forecast now representing a bear case, suggesting CPU to GPU workloads could move toward a 1:1 ratio.
We laid out some ARM short to hedge against our other CPU longs. We usually stay away from this stock both long and short unless it’s a tactical trade if the chart gives us an opportunity. No strong bias either way — we like the CPU tailwinds & the Vera/Tra...
The company's stock showed positive movement despite other CPU names being softer.
The acquisition of MEXT is viewed as a potentially important technology for AI memory optimization, which could help reduce AI infrastructure costs by making flash storage behave more like DRAM for certain workloads.
In Techlandia, strength in semis continues although more bifurcated this morning with HDDs/Memory leading us higher (WDC/STX +8-10%) and SNDK/MU +2-3% while CPU names softer (ARM -2%, INTC -1% although AMD +50bps)
The company's recent meetings with sell-side analysts suggested positive developments, contradicting a negative report from M-sci.
Third-party sources expect a slight downside to data center revenue in Q2 due to slowing CPU deployments.
M-sci reported that the company's Q2 data center revenues are tracking below street expectations, primarily due to softer EPYC CPU deployments, although this was partially offset by better GPU performance.
Wolfe is constructive on AMD, believing the Street continues to underestimate both its GPU and CPU opportunities.
Management indicated the MI450 remains on track, with revenue expected to increase materially in Q4 and Q1 as deployments at Meta and OpenAI ramp, and AMD is comfortable with its wafer supply through CY27.
The firm highlighted a larger CPU opportunity driven by agentic AI, where high-core-count CPUs increasingly handle orchestration and reasoning workloads, supporting a server CPU TAM above prior expectations.
Wolfe believes AMD's earnings power remains well above consensus, driven by incremental MI450 deployments, expanding CPU demand, and continued datacenter share gains.
Northland estimates ALAB’s scale-up and scale-out switch revenue will reach roughly $350M this year, up 180% y/y, driven by accelerating adoption of its Scorpio product family and growing deployments with AMZN, AMD, and other hyperscale customers.
Wolfe and JPM had positive takeaways following pre-quiet period meetings with management, indicating favorable business developments.
Irrational Analysis was positive on the company's CPU opportunity, stating it could add significant gross profit dollars if it secures incremental N3P wafers from TSM.
AMD +7%: Wolfe/JPM had some positive takeaways following pre-quiet periods with management (we covered them in our morning wrap). Irrational analysis was also positive on AMD’s CPU opportunity said they could add significant gross profit dollars if they get...
Citi upgraded the stock to Buy and raised its price target, believing investors are underestimating the CPU opportunity driven by agentic AI and AMD's strong position in custom AI accelerators.
The server CPU Total Addressable Market is projected to reach approximately $137 billion by 2030, an increase from previous estimates, as agentic workloads increasingly shift orchestration, reasoning, memory management, and tool-use functions towards CPUs.
Meta could become a significantly larger customer than currently anticipated, with custom MI450 deployments potentially generating approximately $15 billion in revenue.
Citi raised its calendar year 2026 and 2027 EPS estimates by 12% and 13% respectively.
BofA remains constructive on AMD due to the expansion of the server CPU TAM driven by agentic AI.
CPUs/AMD/ARM: BofA Raises Server CPU TAM to $170B, Says Agentic AI Fundamentally Expands the Opportunity
Against this backdrop, BofA raised targets across CPU-exposed names, double-upgraded INTC to Buy, remains constructive on AMD and ARM, and continues to view
The company's stock price increased following BofA's upgrade of INTC, which also talked up CPUs.
China's plan to spend $295 billion on AI infrastructure over five years aims to rely on local suppliers like Huawei for at least 80% of technology, effectively squeezing out Advanced Micro Devices.
More gigawatt-scale customers for MI450-500 are in the works, and a Microsoft meeting felt confirmatory. AMD is retaining parts of GOOGL's head-node CPU, and the CPU opportunity is seen as a percentage of GW-scale footprint, with CPU-only rack setups becoming a significant percentage of the redesigned footprint (implied 10-20%).
ISI issued bullish notes after meetings with management, highlighting robust interest in higher-core-count Turin CPUs for agentic workloads and expected pricing improvements with increased core counts.
Management discussed upside from expanding CPU attach opportunities alongside its AI accelerator portfolio, including wins where customers selected AMD CPUs to pair with Rubin GPUs.
DB believes the announcements create incremental competitive pressure for AMD and INTC in datacenter CPUs and potentially QCOM in AI PCs.
TD Cowen raised its price target to $600 from $500, citing growing conviction that agentic AI is creating a meaningful incremental CPU opportunity.
The company's recently increased $120B CPU TAM could still prove conservative as enterprises deploy AI workloads requiring high-performance, low-latency CPUs, benefiting the EPYC franchise alongside Instinct accelerators.
Management reiterated that MI450 and Helios remain on track for a 3Q26 launch, with volumes ramping through 4Q26 and 2027, indicating strong product pipeline execution.
Cowen trimmed near-term accelerator estimates but raised CPU assumptions meaningfully, resulting in higher FY27 revenue and EPS forecasts.
Clev is positive on the company, raising numbers and stating that CY26 CPU growth is now +20% year-over-year, as the company continues to raise prices.
AMD/INTC: Hearing Clev positive raising numbers saying CY26 CPU now +20% y/y as both continue to raise price. Also saying ARM feedback stepping up due to x86 shortage & strong Verea feedback
The company's stock price decreased, indicating a negative market reaction.
AMD/ARM/NVDA/INTC: Wolfe Sees AI CPU TAM Driving ~30% CPU Market Growth Through CY28, AMD Biggest Beneficiary
The company's stock reached new highs, following the positive momentum of ARM.
Wolfe Research identified the company as the biggest relative beneficiary of Agentic growth, projecting server CPU revenue to reach $44 billion by calendar year 2028 and estimating incremental AI CPU revenue could add approximately $7 to EPS versus calendar year 2025.
AMD +5% following along to new highs although INTC -70bps lagged. Wolfe had the latest sell-side note talking up Agentic growth, saying AMD is the biggest relative beneficiary, modeling server CPU revenue reaching $44bn by CY28 and estimating incremental AI...
The company's stock reached all-time highs, indicating strong performance in the CPU trade.
CPU demand signals remain strong, and the company is actively building out alternative packaging supply paths.
AMD +4% as Lisa Su was in Taiwan sounding good around CPU demand. They also announced >$10B in Taiwan ecosystem investments ahead of Computex, as unlike NVDA< AMD is continuing to build a 2nd/3rd packaging supply path.
Demand for AI infrastructure is growing much faster than market expectations, with the CPU market expected to grow over 35% each year for the next five years due to the ramp-up of inferencing and agentic AI.
The company is expanding its supply capacity and plans to invest over $10 billion in Taiwan's ecosystem for advanced packaging, substrates, and AI rack-scale system capacity from 2026 through 2029.
The AI infrastructure market is large enough to support multiple architectures and suppliers, indicating that AMD can challenge Nvidia's dominance.
ASICs are considered reasonable and important, with the goal of using the right chip for the right job.
The CEO's comments in Taiwan indicated strong CPU demand, and the company announced over $10 billion in Taiwan ecosystem investments, distinguishing itself from NVDA by building a second and third packaging supply path.
AMD +4% as Lisa Su was in Taiwan sounding good around CPU demand. They also announced >$10B in Taiwan ecosystem investments ahead of Computex, as unlike NVDA< AMD is continuing to build a 2nd/3rd packaging supply path.
In terms of read-throughs, positive for ARM although we don’t know the actual #s from licensing deal with Vera. The $200B TAM is > than the $125B Lisa Su threw out on AMD’s EPS call, but AMD -2% down on concerns of share loss (INTC -3% down
The stock saw a significant increase, indicating strong market performance.
The company is preferred by some analysts due to expectations of continued share gains in the x86 enterprise market.
Inventec is expected to ramp the MI450 for Microsoft, with mass production starting in Q4 and more meaningful revenue in 2027.
TPU volume checks are mixed but some indicate significant year-over-year growth for 2026.
AMD +14% and INTC +7% as the CPU trade remains in full activation after INTC’s Lip Bu Tan was out talking positive on the CPU:GPU attach rate at JPM. Key quote:
ARM architecture is going to be more common in agentic AI BUT it doesn't 1) participate in pricing upside 2) own CPU is multiple years away and creates a conflict of interest with licensors 3) significant mobile phone exposure frontloaded 4) even if ARM hit...
* **$AMD**: Inventec is expected to ramp the MI450 for MSFT with mass production starting in Q4, with more meaningful revenue in C27. * TPU volume checks are mixed, with some indicating 3M units for C26, while others suggest 3.6-3.8M units for C26 (+150% Y/...
The analyst remains wary of the company's dependency on OpenAI and desires to see more deals in that area.
Concerns exist regarding the company's execution with GPUs.
The stock declined as some fast money investors opted not to hold long positions ahead of NVDA's earnings report.
AMD -1.6% as heard some very fast money types not wanting to be long into the NVDA earnings.
JPM materially raised its CoWoS estimates for 2027, modeling approximately 180K wafers/year, driven by stronger MI450/Venice ramps and increasing CPU/GPU packaging complexity.
TSMC/Packaging/AMD: JPM Raises CoWoS Outlook; AMD and AVGO Estimates Moving Higher
AMD -75bps outperforming as JPM raised CoWoS outlook, not modeling ~180k wafers/year driven by stronger Mi450/venice ramps and increasing CPU/GPU packaging.
AMD +1% / INTC -3.6%: Some noise around CPUs today, with some pointing to Tencent’s comments. Funda.ai had some good color, as always:
The stock saw a positive movement despite some noise around CPUs.
AMD +1% / INTC -3.6%: Some noise around CPUs today, with some pointing to Tencent’s comments. Funda.ai had some good color, as always:
AMD is taking server CPU market share from Intel, driven by hyperscaler AI demand.
Agentic AI increases CPU intensity, favoring AMD's high-core-count and multithreaded positioning.
AMD delivered strong Q1 results and materially raised its long-term x86 CPU TAM outlook.
CPUs: UBS Says Server Demand Remains Strong While AMD/ARM Continue Taking Share from INTC
The stock has experienced a sharp valuation rerating, leading to a downgrade in rating.
Agentic AI is significantly increasing CPU demand and CPU-per-GPU ratios, supporting a potential substantial CPU TAM by 2030 as orchestration and inference workloads scale.
The company is particularly well-positioned through its Turin/Genoa products and exposure to AI orchestration.
The firm highlights strong server CPU demand trends from AMD and Intel, with hyperscalers accelerating AI capex and internal ARM CPU deployments further validating the broader CPU opportunity.
AMD/INTC: Mizuho Raises PTs as Agentic AI Drives CPU Demand Inflection
The company is in advanced discussions with Samsung Electronics' Foundry Division for 2-nanometer (nm) chip orders, indicating efforts to address wafer supply constraints.
The firm is warming to the company as AI and CPU demand broaden beyond core accelerators.
The firm remains constructive on NVDA, AVGO, and semicap, while recently warming on AMD as AI and CPU demand broaden beyond core accelerators.
Despite the upside potential, BofA reiterates Underperform, arguing much of the opportunity is already reflected and preferring AMD/ARM as better-positioned CPU beneficiaries.
Fubon reports that the CPU shortage is worsening into 2026-2027, with AMD lead times now exceeding 20 weeks.
The company's CEO provided a very bullish outlook on CPUs, doubling the server CPU Total Addressable Market (TAM) forecast.
Server CPU revenues are expected to grow significantly in the current quarter and continue growing into future years.
Engagements with major customers like OpenAI and Meta for GPUs are progressing well, with customer forecasts exceeding initial plans.
Investors showed little pushback, with bulls anticipating higher EPS and arguing for a higher valuation multiple.
ARM +7% numbers mixed and some head scratching in conversations about why stock up 7% after being up 13% today. Licensing beat offsetting royalty with the guide largely in-line. Expects had been very high here following INTC/AMD and bulls want to hear simil...
AMD is expected to be a significant beneficiary of the agentic AI trend, which is projected to drive a substantial expansion in the server CPU TAM, with a potential 11% EPS impact.
UBS continues to see ARM as the biggest beneficiary (+24% potential EPS impact), followed by AMD (11%) and then INTC (7%), with near-term demand broadly supportive but longer-term differentiation emerging as workloads mature.
The stock is being aggressively bought on any minor pullback, indicating strong investor confidence.
The company's earnings were considered good enough, with better revenues.
The CEO provided a bullish outlook, expressing confidence in exceeding the $20 EPS target and achieving tens of billions in GPU revenue next year.
The server CPU TAM growth guidance was significantly raised, with the company's server CPU revenues showing strong year-over-year growth.
The company has a world-class supply chain and is making significant investments to expand capacity, with a clear path to exceed long-term financial targets.
The stock is not considered expensive given the potential for exceeding the $20 EPS target, which is now seen as a floor.
Charts are being bought aggressively on any mini-pull back (AVGO, LITE, AMD), plenty of names breaking out to new highs, beats are following through to the upside (NOK, STX, QCOM, WDC, TXN, VRT, INTC), and misses are being bought up (CLS, TER, GLW, ON, AMZN...
Memory names continue to lead the way higher with SNDK/MU up double digits as INTC +13% gets close to meme territory following news AAPL is looking to potentially use them as a foundry partne. Stocks are no longer even waiting for T+1 to go red to green - j...
Gross margins slightly missed internal targets.
The stock was down 6% following news about OpenAI missing internal targets and spending scrutiny, indicating market concern about its exposure to the AI complex.
The firm says advanced node capacity is moving into structural shortage by 2027 (15– 20% gap), with demand far exceeding supply as AI customers (NVDA, AVGO, AMD) compete for allocation. The analyst notes pricing power is set to inflect higher, with N2/A16 w...
Northland downgraded AMD due to increasing competitive pressure from Intel and Nvidia, which is likely to constrain AMD’s share gains and margin expansion.
Northland believes CY27 expectations for AMD appear too high and carry a risk of a reset, as AI infrastructure demand growth slows and hyperscaler capex faces constraints.
Northland noted that elevated R&D spend and supply advantages favoring Nvidia limit near-term upside for AMD despite solid execution.
Citi closed its positive catalyst watch on AMD due to the stock's recent rally.
Northland believes that while the upcoming quarter should be good, expectations for calendar year 2027 are too high.
Citi closed out its positive catalyst watch on the name, indicating that the recent rally has run its course.
AMD -3% as Northland downgraded to Market Perform saying increasing competitive pressure from INTC and NVDA, alongside TSMC allocation contrained dynamics, is likely to constrain AMD’s share gains and margin expansion. They think q should be good next week...
The firm upgraded the stock to Buy and raised its price target to $375, citing Intel's strength as a signal of a major CPU demand inflection, positioning AMD for significant upside as CPUs re-emerge as a bottleneck in AI infrastructure.
The analyst notes that AMD is well-positioned to gain share and exceed estimates, with 2026 revenue and EBITDA projected at $2B and $1.5B above consensus, respectively.
Tight CPU supply dynamics and accelerating enterprise/server demand support durable growth, while GPU contributions are becoming incremental to the story.
The firm expects sustained demand for advanced networking solutions as hyperscalers scale AI workloads.
AMD remains a strong x86 competitor, and ARM-based options from NVIDIA, AWS and Google create a medium-term share risk.
Net, Barclays sees multiple headwinds—including competitive risk (NVDA/AMD), high power consumption, and potential Samsung insourcing—keeping risk/reward skewed to the downside.
The stock continues to benefit from the CPU trade and positive checks regarding mi4xx, along with the Anthropic news, creating strong dual narratives that are difficult to counter.
The company is expected to benefit from improving CPU demand driven by agentic AI, although the stock's upside is considered more balanced compared to memory plays.
Morgan Stanley says INTC and AMD should benefit from improving CPU demand driven by agentic AI, but sees memory as the cleaner way to play the trend, favoring MU and SNDK on valuation and upside. The firm raises INTC PT to $56 and estimates on stronger serv...
MS adds AMD is better positioned fundamentally in servers but stock upside is more
Bernstein raised server assumptions, projecting approximately 50% year-over-year growth, and incorporated a new META AI deal, suggesting upside to 2027 estimates not yet fully reflected in consensus.
Bernstein is incrementally more positive on the company due to server and AI momentum, seeing clearer upside potential as AI and server demand ramps.
Feedback on CPU and GPU performance continues to improve, indicating positive market reception.
INTC/AMD: Bernstein Sees Server Strength Offsetting PC Weakness; More Constructive on AMD
For AMD, the firm raises server assumptions (~50% YoY growth) and incorporates a new META AI deal, seeing upside to ’27 estimates not yet fully reflected in consensus, while noting continued PC softness near term.
The stock led the market higher due to chatter that Semianalysis reported Anthropic plans to use AMD's Mi450, as Anthropic continues to be compute-constrained.
Bernstein issued a positive note, highlighting CPU tailwinds for the company.
EdgeW also provided a positive assessment, stating that CPU and GPU feedback continues to improve month-over-month.
The company is experiencing continued PC softness in the near term.
Clev is positive on the company, expecting upside to mi455 and strong server CPU demand and pricing.
Positive channel checks from Cleveland Research highlight potential upside to AI chip shipments and server CPU pricing.
difference between both: Sentiment here (and among minders) continues to pick given GPU pricing + demand/supply dynamics, especially more noise around DCs running into roadblocks in some places (today it was Maine announcing DC moratorium) Neoclouds ZIHEZ:...
TD Cowen sees medium-term upside from CPU share gains at cloud customers due to agentic AI reawakening CPU demand, as autonomous agents elevate the CPU to a central decision-making hub for multi-step workflows that GPUs don't service well.
TD Cowen flagged that approximately 90% of server CPUs are on TSMC's 3nm/5nm nodes, representing a tight capacity constraint.
The analyst sees solid AI pipeline visibility, including next-generation GPUs, and longer-term demand remains intact despite modest near-term trims.
TSM/NVDA/AVGO/AMD: JPM Raises CoWoS/SoIC Outlook; Near-Term AMD&NVDA Trim, ’27 Upside Intact TSM/NVDA/AVGO/AMD JPM CoWoS/SoIC
JPM adds strong upward revisions to AVGO CoWoS (TPU v7/v8 demand), with volumes trending higher. The analyst also sees solid AI pipeline visibility (incl. next-gen “Feynman” GPUs on A16/3D SoIC), while noting modest near-term trims for AMD on MI450 timing (...
There are modest near-term trims due to MI450 timing.
Goldman Sachs sees tactical long opportunities due to server CPU strength.
Goldman sees a constructive 1Q setup post-pullback, with most fundamental upside in compute (server CPUs/ASICs), HDD/NAND (tight supply), and semi cap (DRAM/foundry pull-ins). Tactical longs: TER on tester demand and GPU testing upside, AMAT on capacity pul...
The stock continued to perform better, supported by CPU tailwinds.
Semis: GS 1Q Preview; Broad Upside — Long AMAT, TER, AMD; Cautious KLAC, ON, ARM AMAT TER AMD KLAC ON ARM
Wells Fargo models Q1 server CPU revenue at $2.92B, significantly above typical seasonal trends, driven by Turin CPUs with higher blended ASPs.
The firm expects Q2 guidance around $10.3B and raised its CY26/CY27 DC GPU revenue forecasts, with potential for over $40B upside if additional GW-scale MI455X wins materialize.
AMD's CY27 pipeline for MI455X is estimated at 3.0–3.5+ GW, with line-of-sight to over $50B in annual revenue.
Wells Fargo added AMD to its 2Q Tactical Ideas List, expecting a positive setup into 1Q results driven by strength in EPYC server CPUs and new GW-scale GPU announcements.
The firm sees increasing evidence of upside to server CPU revenue growth, potentially exceeding 40% year-over-year, driven by cloud demand and market share gains.
Data Center GPU revenue could see an upside path to over $40 billion with additional deployment wins beyond OpenAI and Meta.
Wells Fargo sees a compelling risk-reward ahead of the July AI Accelerate event, with the stock trading at a discount to its 3-year median NTM P/E.
Early price action shows continuing of yesterday: Memory names weaker while large cap tech is outperforming (AVGO /GOOGL/NVDA/MTEA/AMZN/AMD +1.5% while memory/HDDs flat to down), the perception being what is bad for memory is good for hyperscaler ROI. There...
AVGO/GOOGL/NVDA/MTEA/AMZN/AMD +1.5% /HDDs
The stock held in relatively well despite the broader market weakness, especially given its underperformance compared to leading names since October.
Nothing spared, although NVDA -1.5%, AVGO -2.4%, and AMD -3% held in relatively well given all 3 large caps have underperformed the leading names by a ton since Oct…
Nikkei is also out saying shortages driving CPU prices higher for INTC (+4%) /AMD (+2.4%). This comes on the back of what has been positive checks/commentary from mgmt teams over the last couple months given the huge ramp in Agnetic AI to start the year.
Worsening supply constraints in CPUs are adding pain for PC and server makers, with a meaningful gap appearing between needed and obtainable volumes.
The supply crunch is increasing wait times and pushing up prices for CPUs for both servers and PCs, with some forecasting the situation to worsen.
Quote prices for CPUs have already increased multiple times this year, leading to an average price hike, and the company has told clients it will increase prices for all series of CPUs.
Edgewater is positive on the company's CPU and initial MI4xx feedback.
The stock tends to be bought when it dips below $200, indicating a favorable risk/reward despite a somewhat unclear competitive narrative.
The stock was included in MS's US OpenClaw beneficiary basket, which saw bids today.
The company's stock also experienced downside movement.
The AI chips deal with Meta is projected to generate substantial revenue for AMD, with each gigawatt of deployment potentially adding significant EPS over five years.
The agreement expands on existing strategic partnerships and aligns roadmaps for silicon, systems, and software to deliver AI platforms optimized for Meta's workloads.
The first deployment will utilize custom AMD Instinct GPUs based on the MI450 architecture and 6th Gen AMD EPYC CPUs, running ROCm software and built on the AMD Helios rack-scale architecture, demonstrating advanced technology integration.
META/AMD: Meta and AMD Agree to AI Chips Deal Worth More Than $100 Billion to Deploy 6GW
The company is experiencing strong demand for its server CPUs and GPUs.
AMD/NVDA: Hearing Clev positive citing strong Server CPU (AMD) & GPU Demand
The MI355X chip, while physically generating more tokens per second than NVIDIA's H200s in simple tests, runs slower or fails in complex setups due to software being approximately six months behind, negating its hardware advantage.
The MI455X, intended for the company's first rack-scale system, is facing significant manufacturing delays, with mass production pushed back to Q2 2027.
The company is seen as a relevant name benefiting from improving OpenAI sentiment.
The stock experienced weakness along with other big year-to-date winners.
The stock rallied due to improved sentiment around OpenAI.
Management confirmed that the MI450 is progressing better than expected, despite street rumors, and insisted that rumors of delays for the 2H26 launch are unfounded.
A broader positive piece on CPU demand from Semi Analysis also contributed to positive sentiment.
OpenAI related names rallied on 1) the back of a positive CNBC article which said Altman said Chatgpt back to 10% monthly growth and called out +60% Codex w/w growth and 2) DA Davidson upgraded “OpenAI and its Orbit” leading up to model releases and fund ra...
The company is expected to deliver revenue and earnings upside in the near term, supported by strong demand across client and EPYC CPU segments.
NVDA +1%/AMD +3% up early on positive China H200 news.
Bernstein raised its price target to $225 and increased estimates due to stronger server assumptions.
The company is making progress in AI, and solid server trends are supporting the core business.
There is positive sentiment regarding CPU server strength.
AMD/NVDA: Hearing both Clev and Edgew out positive on both (CPU server strength for AMD)
The stock received a bid due to CPU server shortage and pricing, nearing all-time highs despite some OpenAI sentiment wobbles.
Analysts were positive on CPU server strength, with server unit demand expected to increase by over 25% and supply anticipated to tighten further in the second half of the year.
AMD +7.7% getting the CPU server shortage/pricing bid alongside INTC and sitting close to all-time highs despite some OpenAi sentiment wobbles (see ORCL, MSFT, and Softbank all down). Both Edgew and Clev were out positive today talking up CPU server strength with ser...
Bernstein's Q4'25 and Q1'26 models are roughly in line to modestly below consensus, and FY26E and FY27E are still below the Street's expectations.
There is limited visibility into AI progress, with OpenAI's Helios being the only named customer.
The valuation at approximately 35x P/E already prices in significant AI success.
ORCL -3.4% / MSFT -2.3%: OpenAI weakness as heard Semianalysis was out with a note saying Google is way ahead in locking down power supply in ‘27 and ‘28, the read-through being OpenAI won’t have as much compute available as they are not exposed to TPU. Don...
The company is preferred in AI hardware, alongside CRDO, if cloud capex visibility improves and GTC re-energizes sentiment.
The company's Trainium and AMD solutions have delivered only about 30% cost reductions and remain behind on absolute cost compared to Nvidia and Google/Broadcom's TPU.
Wells Fargo reiterates Overweight and names the stock its top 2026 pick, citing continued CPU share gains, a ramp in rack-scale datacenter GPUs, and a clearer path to $20+/share in EPS by 2029.
The firm sees AMD well-positioned to compound datacenter revenue at over 60% CAGR, supported by Helios rack-scale systems, MI450X GPUs, EPYC Venice CPUs, and Pensando DPUs, with datacenter GPU revenue modeled at $12.1 billion in 2026 and $32.0 billion in 2027.
The analyst highlights growing visibility into multi-gigawatt deployments, including OpenAI’s planned 6GW build with Oracle as the first hyperscaler partner, starting with approximately 50k MI450X GPUs in 3Q26 and extending into 2027.
Server CPU leadership is driving share gains toward a 50% revenue share target and mid-20% CAGR in server CPU revenue.
The company is raising CPU prices amid shortages, with server processor capacity nearly sold out.
Chinese customs authorities have informed agents that Nvidia's H200 artificial intelligence chips are not permitted to enter China, and Chinese government officials have instructed domestic technology companies not to purchase these chips unless necessary.
KeyBanc upgraded the stock to Overweight with a $270 price target, reversing a prior downgrade, as supply-chain checks indicate the company is largely sold out of server CPU capacity for 2026, with hyperscaler demand potentially driving 10-15% pricing increases in 1Q26 and server CPU growth estimated at over 50% this year.
The analyst anticipates a significant MI455 ramp in the second half of the year, including approximately 290,000-300,000 units targeted for the Helios rack-scale platform, supporting estimated 2026 AI revenues of $14 billion-$15 billion.
The company received an initiation at Buy from Chian Renaissance.
AI Semis / AMD / INTC: KeyBanc raises NVDA/AVGO CoWoS estimates and sees CPU Pricing Firming / Upgrades AMD and INTC on CPU refresh/pricing
A third-party firm is reportedly lowering numbers for the company due to softer December CPU momentum, which is more backward-looking than the positive outlook from KeyBanc.
The company is experiencing tight CPU supply due to rising month-over-month demand.
The company confirmed the MI400/Helios in 2026 and MI500 in 2027, indicating continued product development and future growth in AI infrastructure.
GS says CES 2026 confirmed sustained strength in AI infrastructure, with Nvidia, AMD, and Marvell citing robust demand and incremental upside from physical and agentic AI. Nvidia unveiled a context memory storage platform and reiterated a strong Rubin ramp...
There were some positive CPU checks at Edgewater.
The stock is struggling with its event path and what could turn it around, as it is currently tied to OpenAI.
AI Semis/NVDA/AVGO/AMD/CoWoS: Morgan Stanley Raises CoWoS Bookings for NVDA/AVGO, Trims AMD/MSFT Maia; Supply Constraints Still Drive ’26
Morgan Stanley’s Charlie Chan update sCoWoS assumptions, raising NVDA bookings to ~875k (from ~800k) with incremental Non-TSMC allocation and ~40% tied to Rubin, while lifting AVGO to ~280k (from ~235k) and trimming AMD to ~110k (from ~130k), including ~65k...
Morgan Stanley trimmed its CoWoS bookings for AMD, including CoWoS-L for MI400, suggesting a slight reduction in demand expectations for their advanced packaging solutions.
Management expressed bullish sentiment, citing very positive demand driven by exploding inference and token growth, which has customers constrained by compute resources.
AMD highlighted strong interest in its Helios MI455X rack-scale platform, which remains on track for launch later this year, and the MI500 is still tracking for 2027.
OpenAI's emphasis on the supply-demand squeeze for compute, despite tripling its own compute, validates AMD's increasing engagement with hyperscalers and AI labs.
AMD: Lisa sounded bullish with demand message very positive —exploding inference/token growth has customers compute-constrained, and AMD highlighted strong interest in its Helios MI455X rack-scale platform with launch still “exactly” on track for later this...
The recent pullback offers an appealing entry point ahead of the Helios rack-scale launch, with overblown fears around competition and the AI bubble.
Morgan Stanley’s Charlie Chan notes his latest checks point to 3nm capacity running tighter than expected, with TSMC’s 2025 outlook still at 110–120kwpm and 2026 rising to 140–150kwpm, supported by AZ phase-2 output and a 4/5nm conversion, yet still insuffi...
Futures flat but lots of movement underneath the surface with NVDA -4% and AMD -6%. The sell OpenAI levered names vs Buy GOOGL/TPU exposed names was in full force yesterday and likely continues today after TheInformation calling out META possibly buying TPU...
A massive $1B Department of Energy partnership helps offset near-term seasonality in GPU demand.
Ongoing momentum with major hyperscalers and potential for new customer announcements.
Potential near-term seasonality in data center GPU demand as customers wait for next-generation rack-scale shipments.
Citi says DRAM is poised to become the next key component to secure long-term contracts across the AI supply chain, mirroring the trajectory seen with NVDA, AMD, and AVGO. The firm believes Micron will benefit from a tightening supply backdrop and sustainab...
Buyside was about $20 in FY30 while sell side at 18.50 and most view this as conservative. The one nit being that FY28 guide of 10.65 is actually lower than the street at 11.21. Infra revs were raised from 18/32/73/114/144 to 18/34/77/129/166 from FY 26 - F...
BofA cuts INTC to Underperform from Neutral, maintaining a $34 PO, saying the stock’s $80B run now fully prices in balance-sheet improvement and foundry optionality. The firm argues INTC still lacks a competitive AI roadmap and continues losing CPU share to...
The risk/reward profile for AMD is highly favorable and skews heavily to the upside.
AMD's chiplet strategy allows for flexible reuse of compute tiles and interposes, leading to faster iteration and better wafer utilization than Nvidia's monolithic approach.
AMD +5% with some nice follow through after the OAI news still think rlr skews massively to the upside here: Wrote a lot on this one the but seems to be playing past couple of days, out nicely already: One interesting tidbit that friend sent along that I've...
NVDA / AMD: Mizuho More Positive on Asia ODM Checks; NVL72 Forecasts Raised
Q2 revenues beat expectations and the Q3 guide met buyside bogeys and beat the street, even without including China revenue.
The print supports the bullish thesis of a second-half acceleration and significant GPU upside in 2026.
The stock might experience some near-term consolidation because 2026 GPU estimates have risen very rapidly in a short period.
GPU revenue estimates are finally trending upward and investors are increasingly bullish on the long-term story.
UBS raised its price target, citing upside bias to Q2 results and high conviction in the data center ramp.
Semis/IT Spending: JPM CIO Survey Shows Strong AI and Cloud Spending Intentions; NVDA, AMD, AVGO, MRVL Among Top Beneficiaries
Research through 2026-11-02 · 2025-06-27 to 2026-11-02 · 122 source reports.
| Range | Prices through |
|---|---|
| 1W | |
| 1M | |
| 3M | |
| 6M | |
| 1Y | |
| 2Y | |
| 3Y |
Research through 2026-11-03. Artifact validity does not establish factual accuracy or research freshness.