Investing.com — Here are the biggest analyst moves in the area of artificial intelligence (AI) for this week.
Bernstein raised its price target on SanDisk to $1,250 from $1,000 and reiterated its Outperform rating, calling the stock its top short-term pick. In a more aggressive scenario, analyst Mark Newman laid out a blue-sky valuation of $3,000 — implying roughly 250% upside from current levels.
Newman’s core argument is that the market is significantly undervaluing SanDisk’s earnings power. The stock pulled back roughly 19% after Google unveiled TurboQuant, a memory-compression algorithm that sparked fears of weakening NAND demand.
But Newman believes those fears were “overdone,” arguing that the technology primarily targets high-bandwidth memory (HBM) used in AI inference and has limited bearing on NAND’s broader role in storage and AI infrastructure.
On near-term numbers, Newman estimates Q3 adjusted EPS of $14.18, slightly above consensus and above the company’s own guidance range of $12–$14, with Q4 reaching $25.30 against the Street’s $18.78, driven by a projected 40% sequential increase in average selling prices (ASPs).
In the bull case, where NAND prices rise 75% in each of the next two quarters rather than the 55% and 40% Bernstein assumes in its base, he sees full-year fiscal 2027 (FY2027) EPS of $224, the figure underpinning the $3,000 valuation at a 13x peak-cycle multiple.
The base-case target of $1,250 applies 11x to average earnings across FY2026–2029, which Newman estimates at $114 per share.
On valuation, the stock currently trades at just 9x consensus forward earnings, below the 10–13x range during prior upcycles, and at only 0.5x the Philadelphia Semiconductor Index on a forward P/E basis versus 0.8–1.0x historically.
“This suggests that the market does not believe the higher side of our estimates are reasonable and/or that there will be an imminent collapse in NAND prices,” Newman wrote, adding that he believes the market “is significantly undervaluing earnings power and sustainability of this cycle.”
He notes NAND prices only began inflecting meaningfully six to seven months ago, with the industry remaining structurally undersupplied. SanDisk reports fiscal Q3 results on April 30
Earlier this week, Stifel upgraded Texas Instruments to Buy from Hold on Tuesday, and lifted its price target to $250 from $215, as the brokerage expects an inflection in free cash flow (FCF) generation following six years of elevated capital expenditure (capex).
Analysts led by Tore Svanberg said TXN is “well positioned to capture share in the next analog upcycle and return to strong FCF generation” as capital intensity normalizes. With gross capex expected to decline from $2.39 billion in 2026 to $2.20 billion in 2027, Stifel projects FCF per share rising from $8.13 in 2026 to $10.61 in 2028.
CHIPS Act benefits, including an investment tax credit recently raised to 35%, are seen providing additional support, with total projected benefits reaching $1.55 billion in 2026.
The pending acquisition of Silicon Laboratories is flagged as a meaningful earnings contributor, with Stifel projecting over 10% EPS accretion by 2030 from revenue synergies and manufacturing savings as SLAB production shifts into TXN’s 300mm fabs.
Data center exposure is another key plank of the thesis. Stifel projects TXN’s data center revenue — currently around 9% of sales — to reach 20% by 2029, growing at a 35% compound annual rate. The analysts drew a parallel to component shortages earlier in 2026, arguing that power management chips are becoming a similar bottleneck for AI infrastructure buildouts.
“TXN is addressing this supply concern through its 300mm domestic capacity expansion, which provides a dependable source of foundational analog and power management chips,” they wrote.
On manufacturing, the transition to 300mm wafers — targeting more than 80% of revenue on that process by 2030 — is expected to deliver up to 800 basis points of gross margin accretion over time, though near-term depreciation pressure is seen offsetting those gains through 2026 before reversing in 2027.
Stifel’s bull-case target for TXN stands at $286.
UBS downgraded ServiceNow to Neutral from Buy on Friday, cutting its price target to $100 from $170, as analyst Karl Keirstead reassessed whether the company’s competitive position is as resilient to AI disruption as previously believed.
ServiceNow had been UBS’s only Buy-rated application software stock on that basis.
“Given that our confidence in that view has weakened and we’re hearing more anecdotes of non-AI apps software budget pressure, we’re moving to a Neutral rating despite the material YTD de-rating in the stock to 15x 2026 FCF,” Keirstead wrote.
The analyst flagged three concerns from customer conversations. While clients showed no desire to replace ServiceNow as a system of record, some expressed interest in using AI to custom-build workflow automation apps or handle service tickets more agentically.
Customer support was identified as the area most at risk from AI-driven headcount reductions, creating seat growth risk in the CSM segment, which accounts for roughly 10% of revenues. UBS also did not hear consistent buy-in for ServiceNow at the agent orchestration layer.
On budgets, Keirstead noted that over half of enterprise customer conversations now include anecdotes of containing non-AI software spend, as AI infrastructure investment crowds out core software budgets.
“We now also expect skinnier-than-normal beats in the coming quarters, more limited upside to the guidance for stable organic c/c sub revs growth of 19% in 2026 and we’re trimming our cRPO growth estimates to exit 2026 at 16% c/c, down from our previous estimate of 20%,” he wrote.
In another upgrade this week, Guggenheim Securities raised Datadog to Buy from Neutral, describing the company as “a primary beneficiary of AI-driven growth” in data volumes and IT complexity.
Analysts Howard Ma and Joseph DiBartolomeo said their prior hesitation stemmed from Datadog’s heavy exposure to OpenAI, which they believe still plans to migrate off the platform.
But checks indicate that migration is taking longer than expected, buying time for Datadog to build its AI customer base elsewhere. The analysts believe an 8-figure deal announced on Datadog’s Q4 earnings call — described as the largest new customer logo in company history — is with Anthropic, part of a broader group of 650 AI-native customers.
Guggenheim now projects 27% revenue growth in 2026 to $4.36 billion, roughly 6% above consensus. The forecast includes core business growth of 24%, OpenAI revenue growing ~16% versus a prior estimate of a 20% decline, and other AI-native customers growing 220% year-over-year, adding roughly $160 million.
The analysts see growth dipping to 20% in 2027 as OpenAI’s contribution shrinks before a reacceleration to 25% in 2028 as Anthropic and others fill the gap. They estimate Anthropic’s ARR on Datadog could exceed $350 million by 2028, and see lower churn risk compared to OpenAI, given the two companies’ different business profiles.
On competitive moat, the analysts pointed to Datadog’s purpose-built storage and query engines — including its Monocle metrics database and Husky event engine — as difficult to replicate, pushing back on concerns that open-source telemetry tools could commoditize the platform.
They also flagged the Bits AI suite, which automates incident detection and response, as a potential new monetization vector. “The Bits AI suite is leading in its category and could serve as a new monetization lever for Datadog, perhaps via outcome-based pricing,” they said.
In a separate note, Guggenheim downgraded GitLab to Neutral from Buy on Tuesday, removing its $50 price target and citing “more pronounced AI risk and lack of near-term catalysts.”
The central concern is net revenue retention, which has decelerated from 133% in FY2024 to 118% in FY2026. Guggenheim expects it to exit FY2027 at around 113%, below management’s own expectation of ~115%.
The firm attributes the pressure to budget dollars shifting toward AI coding tools, with management acknowledging that over 30% of ARR is under stress — roughly 20% from price-sensitive SMB and mid-market customers, and 12% from the public sector.
The deeper structural worry is GitLab’s seat-based model. “Under what is still almost entirely a seat-based model, GitLab doesn’t monetize incremental usage, only tier upgrades,” the analysts wrote.
As AI coding tools improve and developers use them to write larger shares of code, willingness to expand GitLab spending is likely to erode even without a reduction in developer headcount.
On the agentic side, the broker was skeptical of the Duo Agent Platform (DAP), GitLab’s new credit-based model. “Based on our initial customer conversations, interest has been limited,” the analysts said, noting that 70% of revenue comes from self-managed customers who must upgrade to access the product. Management has guided for minimal DAP revenue contribution in FY2027.
Related articles
5 big analyst AI moves: Upgrades for TXN and DDOG; Further upside seen in SNDK
Nvidia’s new Alpamayo project: What it means for Tesla?
Morgan Stanley CIO survey: Why AI hype isn’t boosting 2026 IT budgets
