Analysts made several consequential moves this week on companies exposed to AI-related demand, signaling growing conviction that vendor revenue and platform strategies will outpace consensus in coming quarters. The changes include tactical adds, Buy starts and upgrades that reflect a combination of near-term earnings momentum and longer-term structural shifts in how networks, security and compute are consumed as AI workloads proliferate.
Evercore adds Arista to Tactical Outperform ahead of Q1
Evercore ISI analyst Amit Daryanani placed Arista Networks on the firm’s Tactical Outperform list as the company approaches its first-quarter report scheduled for May 5. Daryanani is looking at a beat-and-raise setup, forecasting that Arista will top consensus revenue and earnings per share estimates of $2.61 billion and $0.81, respectively. He expects AI-driven demand and steady enterprise traction to be the principal near-term catalysts.
Beyond the March quarter, Daryanani anticipates Arista will guide second-quarter revenue north of the $2.78 billion consensus and will at minimum issue a qualitative lift to its roughly $11.25 billion full-year target. The analyst highlighted Arista’s successful penetration beyond its historic customer base, noting incremental wins at Anthropic and Google. He also pointed to the March launch of XPO as a potential accelerator while network clusters transition from 800G to 1.6 terabit architectures.
Berenberg starts Palo Alto at Buy, views risk-off sentiment as entry point
Berenberg initiated coverage of Palo Alto Networks with a Buy rating and a $215 price target, framing the recent software-sector sell-off as an attractive buying opportunity for one of cybersecurity’s market-leading platforms. Analyst Rahul Chopra argued the sector has been pulled into a broader narrative that AI will cannibalize traditional software, pressuring shares and leaving Palo Alto trading about 30% below its 2024-25 average multiple despite no material deterioration in organic growth trends.
Chopra sees AI not as a threat but as a demand amplifier for cybersecurity. As firms integrate AI across workflows, the resulting increase in data flows, cloud applications and identities expands the attack surface and the need for comprehensive protection. He emphasized that the rise of agentic AI expands the total addressable market for cybersecurity vendors since AI agents themselves require protection.
The analyst also stressed Palo Alto’s data advantage - proprietary, near-real-time threat intelligence aggregated across a large customer base - which he argued is difficult for AI model providers to replicate. At the core of Berenberg’s bullish case is Palo Alto’s "platformisation" strategy. Through this approach the company replaces multiple point products with integrated security across network, cloud and security operations. Berenberg estimates that Palo Alto has completed about 1,550 platformisations to date, up from 850 in mid-2024, with management targeting 2,500 to 3,500 by fiscal 2030. Chopra considers the market’s consensus to understate both the pace and economics of this strategy and models 6-14% upside to FY30E consensus revenues.
Chopra highlighted the attractive economics of platformised customers. Palo Alto reports a net revenue retention rate of 119% for this cohort with low single-digit gross churn. Customers consuming all three platforms spend about $4.1 million per year on average, compared with $86,000 for single-platform customers. He flagged the company’s recently completed acquisitions - identity security firm CyberArk for $19 billion and observability platform Chronosphere for $3 billion - as meaningful catalysts that help expand Palo Alto’s estimated addressable market to $206 billion. Berenberg concluded that current multiples do not adequately reflect improving portfolio quality, platformisation success and Palo Alto’s strong Rule-of-40 metrics within the software peer group.
DA Davidson upgrades AMD to Buy on resurgent CPU demand
DA Davidson moved AMD from Neutral to Buy and raised its price target from $220 to $375, citing what analyst Gil Luria described as a structural increase in CPU demand and greater clarity around AMD’s role in the expanding data center market. Luria sees meaningful upside to AMD’s estimates starting with the March quarter, ahead of AMD’s May 5 earnings release.
Luria noted Intel’s stronger-than-expected results as an early sign of a broader recovery in the CPU market, suggesting those results foreshadow similar upside at AMD. He highlighted Intel management commentary pointing to double-digit industry growth through 2027 and the shifting CPU-to-GPU mix as agentic workloads evolve. Specifically, he referenced Intel’s observation that the GPU-to-CPU ratio, once roughly 8-to-1 for pretraining, is moving closer to parity for agentic workloads — a trend that reasserts the importance of CPUs in the AI stack.
That dynamic led Luria to conclude that agentic AI workloads are driving unprecedented server CPU demand and that AMD is positioned to benefit materially, including the ability to raise prices across its portfolio as demand likely outstrips supply. DA Davidson lifted its 2026 revenue forecast by $2 billion and gross profit by $1.5 billion, stating that recent data make AMD’s long-term targets more attainable.
KeyBanc raises Intel price target after strong quarter
KeyBanc elevated its Intel price target to $110 from $70 while keeping an Overweight rating following Intel’s robust first-quarter results and a second-quarter revenue guide that topped expectations. Intel shares jumped more than 23% on the print, reaching a new all-time high.
KeyBanc pointed to strong server CPU demand led by agentic AI as the primary growth engine. Intel’s Data Center and AI segment expanded 22% year-over-year, and the firm’s gross margin improved nearly 500 basis points to 41% aided by price increases and improved yields on its 18A manufacturing process. KeyBanc also noted progress in foundry yields for 18A and commitments expected for the 14A process in the second half of 2026. Intel’s advanced packaging backlog grew as well, with multi-year buildout plans for back-end facilities to satisfy committed demand in 2027.
DA Davidson begins coverage of Reddit with Buy, citing under-monetization and data relevance
DA Davidson’s Wyatt Swanson initiated coverage of Reddit with a Buy rating and a $200 price target, arguing that the year-to-date decline in Reddit shares — roughly 33% — has opened a compelling entry point. The stock’s drop reflected investor concerns about slowing user growth, negotiation leverage with large AI companies, and macro uncertainty.
Swanson judged Reddit to be significantly under-monetized versus peers and highlighted the platform’s structural strength as a text-based community discussion forum that serves as a high-quality data source for large language models. He cited a study referenced in his note identifying Reddit as the most-cited domain across major AI platforms such as ChatGPT, Google AI Overviews and Perplexity.
Looking toward near-term results, Swanson pointed to Similarweb traffic data suggesting Reddit could beat daily active user consensus by about 1.3% for the March quarter, aligning with its historical average beat since going public. He added that April year-over-year traffic performance was the strongest observed in 12 months, implying potential upside as expectations were marked down after a choppier March.
Swanson also discussed Reddit’s existing and prospective data licensing deals with AI firms. Contracts with Google and OpenAI are scheduled for renewal in early to mid-2027; his base case assumes Reddit can negotiate renewals at 15-30% higher rates. While most catalysts are expected to materialize in the second half of 2026 and into 2027, Swanson concluded that Reddit’s significant YTD underperformance represents a buying opportunity.
What these moves collectively indicate
The analyst actions this week paint a consistent picture: sell-side research is increasingly positioning select technology and platform companies to benefit from AI-related demand across networking, compute and data licensing. Several themes recur across the notes: AI workloads are altering infrastructure mixes, elevating server CPU demand; platform consolidation drives higher customer spend and retention; and proprietary data sets underpin differentiated monetization opportunities for certain digital platforms.
From a market-structure perspective, the investment implications differ by subsector. For networking and switching vendors, like Arista, analysts are focused on order cycles tied to higher-capacity links and enterprise/AI cluster rollouts. For cybersecurity firms, platform consolidation and recurring revenue dynamics are central to valuation and long-term upside. In compute, the renewed emphasis on CPUs as integral parts of agentic AI stacks is reshaping demand assumptions. For content and community platforms, the intersection of unique data assets and licensing deals with AI players frames monetization potential.
Bottom line
The recent analyst activity underscores growing bullishness in names exposed to AI adoption, with fresh Buy ratings, upgrades and price-target increases grounded in company-specific data points such as platform adoption statistics, customer spend differentials, margin improvements and traffic metrics. While timing and execution remain important, the consensus among these analysts is that the AI-led demand backdrop is broadening addressable markets and improving economics for select vendors across networking, security, compute and data licensing.