IBM's Quantum Timeline Gets Concrete as New AI Models Outperform Bigger Predecessors

IBM's stock falls 21% YTD despite strong Q1 earnings, new AI platform 'Bob', and a 2029 quantum computing target, leaving analysts puzzled.

The disconnect between IBM’s operational momentum and its stock price is becoming increasingly difficult to ignore. While the technology giant rolls out ambitious new platforms and posts solid quarterly results, its shares have shed roughly a fifth of their value since the start of the year — a divergence that has analysts scratching their heads.

On the product side, IBM has been busy. The company launched its enterprise AI platform, codenamed “Bob,” into general availability on April 29, targeting the full software development lifecycle. The system blends IBM’s proprietary Granite models with third-party offerings from Claude and Mistral, and internal testing across 80,000 IBM employees showed productivity gains of up to 45 percent on complex workflows. Developers stand to reclaim as many as ten hours per week, according to company data. Pricing starts at $20 per month for the Pro tier, with the Ultra version commanding $200 monthly.

Alongside the platform release, IBM Research unveiled the Granite 4.1 family of language models, available in three sizes. The mid-range 8B parameter model reportedly outperforms its predecessor — a 32B parameter version — in computational efficiency, marking a notable leap in what IBM calls “compute economy.”

A New Lab and a 2029 Target for Fault-Tolerant Quantum

The same day, IBM and the Massachusetts Institute of Technology inaugurated the MIT-IBM Computing Research Lab in Cambridge, Massachusetts. The facility replaces the earlier Watson AI Lab and aims to bridge quantum hardware with classical computing and artificial intelligence. CEO Arvind Krishna used the occasion to set a concrete timeline: fault-tolerant quantum computers capable of tackling problems beyond the reach of today’s AI systems should be feasible by 2029.

IBM is also expanding its physical footprint in the United States. A new hub in Chicago will create 750 jobs focused on cybersecurity, data science, and quantum computing, signaling the company’s long-term commitment to these disciplines.

Earnings Beat Fails to Lift the Stock

The first quarter delivered better-than-expected numbers. IBM posted earnings per share of $1.91, topping the analyst consensus of $1.81. Revenue climbed 9.5 percent to $15.92 billion. The board raised the quarterly dividend to $1.69 per share, giving the stock a yield of roughly 3.0 percent.

Yet the market response was decidedly cool. Shares dropped about 2.7 percent on the day, caught in a broadly weak trading environment. At current levels near €196, the stock sits roughly 21 percent below its start-of-year price and nearly 28 percent off its 52-week high. The relative strength index has slipped to 40, a technically oversold reading that suggests the selling pressure may be nearing exhaustion.

The stock has drifted dangerously close to its 52-week low of €189.44, and with the ex-dividend date set for May 8, traders are watching whether that support level can hold. The share price currently trades about 16 percent below its 200-day moving average, a technical signal that often points to further downside risk.

Analyst Optimism Amid Market Skepticism

Management has held firm on its full-year guidance, calling for revenue growth above 5 percent and an additional $1 billion in free cash flow. The Erste Group Bank recently lifted its 2026 EPS estimate to $12.41, a figure that sits slightly above the broader analyst consensus. The average price target across Wall Street stands at $296 — implying upside potential that the market, for now, appears unwilling to price in.

The gap between IBM’s operational trajectory and its valuation is widening. Whether the company’s quantum ambitions and AI productivity gains will eventually close that gap — or whether the market’s caution proves prescient — remains the central question for investors watching this storied technology bellwether.