CoreWeave shares (NASDAQ: CRWV) closed Monday at $87.39, a decline of 2.49%, before edging up 0.18% to $87.55 in after-hours trading. The pullback came despite a bullish assessment from Truist, which maintained its Buy rating and $165 price target on the stock, citing the cloud GPU provider's recent price increases and improving contract economics.

Truist Maintains $165 Target Amid Pricing Power

Truist analyst Arvind Ramnani reaffirmed the $165 price target following CoreWeave's latest company conference, a level representing substantial upside from Monday's $87.39 close. The target reflects expectations for stronger operating leverage as pricing gains create room for higher operating margins. Price targets like this one represent a single firm's assumptions rather than guaranteed outcomes, and Truist's case rests on the company converting announced pricing gains into the operating margins the analyst projects.

The company, which went public on Nasdaq in March 2025, is among a group of specialized providers that lease Nvidia GPU capacity to AI developers and enterprises alongside the larger hyperscale clouds. It disclosed another 10% price increase during its Fully Connected 26 conference, following a 25% hike implemented in July. Taken together, the two moves have lifted pricing across affected services by roughly 37.5% compared with levels before the July adjustment.

The company operates a capital-intensive model that depends heavily on GPU deployment, utilization, financing costs, and long-term customer commitments. In that context, pricing power carries outsized importance: higher revenue per unit can improve returns on expensive computing infrastructure, particularly when hardware costs grow more slowly than customer prices. The latest increases also indicate customers continue to accept higher costs while demand for advanced computing capacity remains firm across training workloads, inference applications, and enterprise deployments.

Extended GPU Life Could Improve Asset Economics

CoreWeave also highlighted an Nvidia A100 contract running through 2029, extending the revenue life of older accelerator hardware. Nvidia introduced the A100 in 2020, meaning those GPUs could remain revenue-generating for approximately nine years. That period exceeds the roughly six-year economic-life assumption CoreWeave currently applies to parts of its computing infrastructure.

A longer useful life could strengthen asset economics by allowing the company to spread hardware investment across more years and continue earning revenue after equipment passes its assumed depreciation period. Older GPUs may deliver stronger lifetime returns where customers keep using them for suitable computing workloads.

The A100 contract also aligns with CoreWeave's broader infrastructure strategy as newer Nvidia systems join its expanding data-center network. Newer GPUs can handle advanced workloads, while older accelerators serve less demanding inference and enterprise applications. This approach matches hardware generations to customer requirements and extends revenue opportunities across existing infrastructure.

Prepayments and Inference Demand Underpin Growth

CoreWeave said 70% of the deals signed during the second quarter included customer prepayments, strengthening its contract economics. Upfront payments provide cash before services are fully delivered and reduce some of the funding pressure tied to infrastructure expansion. That structure matters for a company that continues to spend heavily on GPUs, data centers, networking systems, and supporting computing capacity.

Stronger upfront payments can also reduce reliance on external financing as CoreWeave expands capacity for large enterprise customers, though its investment model still requires sustained utilization and disciplined capital allocation across new infrastructure deployments. Higher prices, longer hardware lives, and prepayment structures could improve returns where customer demand supports the additional deployed capacity.

Truist also identified inference as a growing opportunity as enterprises shift from model development toward large-scale deployment. CoreWeave is expanding beyond basic GPU capacity through CoreWeave Forge and services built around enterprise computing requirements. Rising inference demand, combined with higher pricing, could lift margins and generate additional revenue across both newer and older GPU assets.

The conference disclosures give readers concrete checkpoints to track in coming quarters: whether the announced pricing increases flow through to reported margins, the pace at which prepayment-backed deals convert to recognized revenue, and how long older accelerators remain in productive service under contracts such as the A100 agreement running through 2029.

This article is based on reporting originally published by Blockonomi.