Intel’s once-dominant position in the semiconductor industry has been severely challenged in recent years, with significant setbacks in both RISC architecture adoption and GPU development. The company’s wavering commitment to RISC-V technology, coupled with fragmented AI and GPU strategies, has resulted in substantial market share losses and financial turmoil. This report examines how Intel’s mishandling of RISC-V adoption directly impacted its ability to compete in the increasingly critical GPU and AI accelerator market.

In August 2022, Intel launched the RISC-V Pathfinder project, signaling what appeared to be a serious commitment to the open-source instruction set architecture. The initiative aimed to develop software support for RISC-V on Intel FPGAs, allowing customers to pre-test CPU designs before manufacturing via Intel’s Foundry Service. This move was seen as a milestone for both Intel and the broader RISC-V community, given Intel’s historical focus on x86/x64.

After announcing enhancements such as multi-core support and new development boards, Intel abruptly terminated the Pathfinder project less than a year later, directing developers to use third-party RISC-V software instead. This sudden reversal sowed uncertainty in the RISC-V processor market and called into question Intel’s commitment to broader architectural innovation beyond x86.

The cancellation of Pathfinder is particularly puzzling when viewed against Intel’s broader RISC-V investments. In February 2022, Intel announced a $1 billion “innovation fund” heavily favoring RISC-V technologies. Intel had framed this move as “complementary” to its x86 processor business, offering a foothold in low-power and embedded systems where Arm excelled.

Despite significant financial commitment, Intel failed to convert its resources into cohesive product development. This operational disconnect exemplifies the core issue: strategic investments without consistent, long-term execution.

While Intel faltered, other companies showcased RISC-V’s capacity for AI workloads. Esperanto Technologies’ modified RISC-V SoC, featuring 1,092 cores on half the die size of Nvidia’s A100 GPU, achieved superior power efficiency. Running 20 watts on neural network tasks, Esperanto’s ET-SoC-1 purportedly matched or outperformed the A100 in MLPerf recommendation benchmarks.

These advancements highlight RISC-V’s untapped potential for specialized AI accelerators—an area Intel could have dominated had it more fully embraced the architecture.

Intel’s GPU development faced similar woes. In January 2025, the company canceled commercialization of its next-generation GPU architecture, Falcon Shores. Originally envisioned as an XPU blending CPU and GPU tiles, it was later narrowed to a GPU only, then ultimately shelved. This further illustrates Intel’s pattern of scattered focus and suboptimal organizational structure—particularly damaging in an era dominated by AI and GPU-centric computing.

Additionally, Intel’s Gaudi3 AI accelerators underperformed financially, missing the $500 million revenue target for 2024. Interim co-CEO Michelle Johnston Holthaus conceded that while Intel remains a dominant host CPU provider, it “is not yet participating in the cloud-based AI data center market in a meaningful way.”

Intel’s acquisition of Habana Labs for $2 billion in 2019 underscored its AI ambitions, but poor integration and competing internal priorities hindered progress. While Amazon claimed Habana’s Gaudi chip was 40% more efficient than Nvidia’s competing solutions, Intel never fully leveraged Habana’s technology. Ex-employees cite bureaucratic delays and internal rivalries, describing Intel as pursuing multiple AI strategies without fully committing to any.

The organizational choice to house Habana under the data platform group (DPG) rather than the GPU division (AXG) further fragmented resources, limiting Gaudi’s market traction and technical evolution.

Intel’s setbacks in RISC-V and GPU markets share underlying themes: strategic inconsistency, siloed organizational structures, and an inability to navigate emerging technologies in sync. The parallel failures eroded Intel’s competitiveness in AI acceleration, allowing rivals like Nvidia to dominate the space. While Nvidia integrates RISC-V cores into its GPUs and SoCs, Intel failed to leverage its once-promising RISC-V foothold to bolster its AI and GPU roadmaps.

Intel’s share price dropped over 38% in 2024, significantly trailing competitors. A tepid AI chip market presence—contrasted with Nvidia’s server business boasting triple-digit growth—magnifies Intel’s struggles. To cope, Intel has cut 15% of its workforce and suspended dividend payments, moves that signal short-term financial triage rather than a robust, future-proof strategy.

Intel’s x86 heritage is rooted in Complex Instruction Set Computing (CISC), which propelled its dominance in PCs. However, the mobile era and rise of AI require power efficiency—a strength of RISC architectures. While Intel understands the technical nuances, it failed to transition effectively. Competitors like Apple and ARM capitalized on RISC-based designs for modern workloads, leaving Intel scrambling to catch up.

While Intel stumbled, the RISC-V ecosystem gained steam. Major players like Nvidia, Qualcomm, and Google showcased RISC-V advancements, highlighting its efficiency for AI. Nvidia, for example, integrates RISC-V cores into GPUs, aiming to ship a billion RISC-V cores by the end of 2024. Even Meta is testing RISC-V-based AI chips. Intel’s inconsistent strategy left it on the sidelines of this evolving market.

Intel’s failures in RISC-V adoption and GPU development stem not from a lack of strategic insight but from weak execution and wavering commitment. By halting RISC-V Pathfinder, mismanaging the Habana Labs acquisition, and juggling multiple AI strategies without focus, Intel has relinquished crucial ground to competitors in the AI acceleration arena. The interlinked nature of these failures exacerbates the company’s decline, as missteps in one area hinder progress in another.

As Intel works to reverse course, it faces a more challenging landscape. Nvidia is entrenched as the AI leader, while companies from Meta to Google push forward with RISC-V-based solutions. Without a unified approach that aligns RISC-V adoption with GPU and AI development, Intel’s path to regaining lost market share remains fraught with uncertainty.