Microsoft Widens Local AI Footprint as Stock Trades Through Technical Headwinds
15.06.2026 - 16:36:38 | boerse-global.de
The software giant is quietly rewriting the rules of its operating system, shifting artificial intelligence from cloud servers directly onto users’ machines. Two recent moves underscore the ambition: Microsoft has opened its Phi Silica language model to run on conventional Nvidia RTX 30-series graphics cards with at least six gigabytes of video memory, and it is embedding a new transcription model, MAI-Transcribe 1.5, deep into Windows through insider preview builds. The latter supports 43 languages and processes speech entirely on-device, eliminating the round-trip to the cloud that previously defined such tasks.
These steps mark a deliberate departure from the exclusive Copilot+ PC strategy. Until now, Microsoft’s local AI features were locked to expensive hardware with specialised neural processing units. By letting developers test Phi Silica on commodity GPUs, the company is courting a far larger installed base of desktop PCs. For corporate IT teams, the impending integration of local dictation into Outlook and Teams means early testing is advisable; the exact hardware requirements and offline model behaviour are still being ironed out.
Investors, however, are focused on the here and now. After a bruising start to the year that wiped about 15% from the share price, the stock found a bid on the Phi Silica news, trading at €341.80 — a gain of roughly 1% on the session. That still left it well below the 50-day moving average of €353.30 and far from the 200-day line, which analysts place at €388 to €389. The secondary article notes a Friday close of €337.85, implying a year-to-date decline of around 16% by that measure. Technical damage is substantial, and the market is demanding evidence that a broader on-device AI push will translate into higher cloud revenue.
Should investors sell immediately? Or is it worth buying Microsoft?
The cloud business remains the anchor. In the last quarter, Microsoft’s total revenue reached nearly $83 billion, with the AI segment alone notching an annualised run rate of $37 billion. Paid Copilot subscriptions have also crossed 20 million users, growing rapidly. These figures underline why the company can afford to experiment with local AI without immediate monetisation: the strategy is ecosystem-building, not direct profit-taking. More users running local AI applications should, in turn, encourage developers and deepen the moat around Microsoft’s software stack.
The path back to prior highs will require more than technical manoeuvring. Both articles converge on the view that a sustained rally depends on proving the new hardware-agnostic approach accelerates cloud consumption. For now, the market is holding its fire, watching a company that is simultaneously laying foundation stones for the next decade while nursing a share price that has lost nearly a sixth of its value this year.
Ad
Microsoft Stock: New Analysis - 15 June
Fresh Microsoft information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
