Sacra estimates that SiFive generated $38.2 million in revenue in 2023. At this stage, SiFive should be understood as an IP company still early in the monetization curve, where reported revenue significantly understates the amount of future value already contracted but not yet recognized.

The bulk of 2023 revenue came from upfront IP licensing fees and engineering services tied to active design programs. Customers pay to license RISC-V processor cores and development kits at the start of a project, with additional revenue coming from custom SoC design services that help customers integrate, optimize, and de-risk their chips ahead of tapeout. These sources drive near-term cash flow but tend to be lumpy, tied to the timing of large customer commitments rather than steady usage.

Revenue is split roughly 60% IP licensing, 30% custom design services, and 10% development boards and tools. With more than 400 design wins and over 100 paying customers—including eight of the top ten semiconductor companies—SiFive’s customer base skews heavily toward large, long-cycle chip programs rather than high-volume, near-term shipments. Average revenue per customer therefore masks wide variance, with a small number of large customers accounting for the majority of contract value.

SiFive was valued at $2.5 billion in its most recent funding round, a $175 million Series F completed in March 2022. Coatue Management led the round with participation from existing investors including Intel Capital, SK hynix, Qualcomm Ventures, Sutter Hill Ventures, Spark Capital, Osage University Partners, and Prosperity7 Ventures.

The company has raised approximately $365.5 million in total funding across multiple rounds since its founding. Early funding included seed and Series A rounds led by Sutter Hill Ventures and Spark Capital.

Intel Capital and Qualcomm Ventures are strategic investors given both companies use Arm designs in their core businesses and are investing in RISC-V via SiFive. The company has discussed a potential initial public offering but has not announced timing for going public.

SiFive’s product is not chips, but the blueprint for chips. Customers license synthesizable RTL and a full development kit (verification collateral, software stacks, documentation) so they can integrate the design into their own SoCs and manufacture at foundries like TSMC or Samsung.

What makes SiFive feel modern compared to legacy IP vendors is the “configuration” layer. SiFive Core Designer, as described here, turns CPU selection into something closer to configuring a cloud instance: engineers pick a family and tune cache, pipeline, safety features, and other parameters, then download a package ready for simulation or FPGA prototyping.

That matters because RISC-V customers are often explicitly trying to differentiate. They aren’t buying a CPU the way you buy a commodity part; they’re buying the right to make tradeoffs (performance per watt, determinism, safety, area, cost) on their own terms.

The portfolio segmentation also maps cleanly to where RISC-V is heading. Essential Series anchors embedded controllers and IoT. Performance P-Series pushes upward into high-performance general compute, with the P870 positioned as the “we can scale” flag planted into server-class coherence. Intelligence X-Series is SiFive’s attempt to meet AI workloads halfway by bundling scalar, vector, and matrix engines rather than pretending CPUs alone will win inference. The Automotive series offers ISO 26262-certified cores with safety features like lock-step execution for ADAS and autonomous vehicle applications.

SiFive operates a B2B IP licensing model that monetizes through upfront license fees and ongoing royalties. Customers integrate SiFive’s cores into their own system-on-chip designs and manufacture those chips at third-party foundries. SiFive does not fabricate silicon and does not sell finished processors, which allows it to work across a broad customer base without competing directly with them downstream.

Revenue is generated through two primary mechanisms. The first is upfront licensing fees, paid when a customer selects a SiFive core for a new chip program and receives access to the RTL, verification collateral, software enablement, and documentation. These fees vary widely depending on performance class, customization requirements, and the scope of the deployment. Large customers may license multiple cores across different product lines under broader agreements.

The second mechanism is royalties tied to customer shipments. Once a customer’s chip enters production, SiFive earns a per-unit royalty based on the number of processors shipped that incorporate its IP. These royalties can persist for many years, particularly in markets such as automotive, industrial systems, networking equipment, and aerospace, where chips remain in production across long product cycles.

A significant portion of current revenue also comes from custom design services. These engagements support customers during integration, optimization, and verification, particularly for teams adopting RISC-V for the first time or targeting aggressive performance and power envelopes. Services revenue carries lower gross margins than pure IP licensing, but it plays a strategic role by reducing execution risk for customers and increasing the likelihood that a design proceeds to tapeout and volume production.

The operating model is capital-light but engineering-intensive. Costs are concentrated in processor design, verification, safety certification, software tooling, and field engineering. Once a core family is complete and validated, it can be licensed repeatedly with minimal incremental cost, which supports very high gross margins at scale. The challenge is timing. Licensing revenue arrives early, while royalties depend on customers successfully shipping and ramping products, which can take multiple years.

SiFive competes across several overlapping layers of the processor and custom silicon stack. The company’s closest competitive pressure comes from incumbent CPU IP vendors, emerging RISC-V specialists, vertically integrated chipmakers, and internal silicon teams at the largest technology companies. In each case, the buying decision is less about a single core and more about ecosystem depth, execution risk, and long-term control over a product roadmap.

Arm Holdings is the dominant incumbent in licensable CPU IP. Arm’s strength is not limited to microarchitecture performance. Its advantage comes from decades of software investment, broad operating system support, mature tooling, and a vast third-party IP ecosystem that integrates cleanly into Arm-based SoCs. For many customers, especially in mobile, consumer electronics, and automotive, Arm represents the lowest-risk path from architecture choice to mass production.

Arm has also responded directly to RISC-V adoption pressure through commercial changes. Programs that reduce upfront licensing costs and broaden access to core IP narrow the pricing gap that historically favored open architectures. As a result, SiFive often competes less on headline cost and more on customization, architectural flexibility, and the ability for customers to avoid long-term dependence on a single proprietary ISA steward.

Intel, through x86, competes more indirectly. x86 remains entrenched in PCs, servers, and enterprise software environments where compatibility and legacy workloads matter. While Intel is also an investor in SiFive, its strategic position highlights the broader challenge SiFive faces in pushing RISC-V upward into performance-sensitive markets that have decades of optimization around incumbent architectures.

Within RISC-V, SiFive competes with a growing group of IP vendors offering licensable cores at various performance and price points.

Andes Technology is one of the largest commercial RISC-V IP providers, with a strong footprint in embedded systems and consumer electronics, particularly in Asia. Andes competes on cost, maturity in low-power designs, and volume deployment, which can make it attractive for customers shipping high unit counts with relatively modest performance requirements.

Nuclei System Technology focuses on embedded and automotive RISC-V cores and has moved aggressively into safety-certified offerings. Early progress on ISO 26262 certification gives Nuclei credibility in automotive programs where qualification timelines and documentation are decisive factors. In regions prioritizing domestic supply chains, Nuclei can also benefit from local alignment and government support.

These vendors compete with SiFive primarily in the lower and mid-range of the performance spectrum, where price sensitivity is high and architectural flexibility matters less than time-to-market and certification readiness.

A different category of competitor combines RISC-V CPU IP with shipped silicon, offering customers working hardware rather than licensable designs.

Ventana Micro Systems targets datacenter and infrastructure workloads with high-performance RISC-V cores delivered as chiplets and full silicon products. By shipping real hardware, Ventana reduces integration risk for hyperscalers and infrastructure vendors that want RISC-V performance without standing up a full internal CPU design effort.

Tenstorrent, led by Jim Keller, blends RISC-V CPUs with AI accelerators and full system designs. The company has secured automotive and AI-oriented design wins and manufactures at advanced foundries. Tenstorrent competes with SiFive by collapsing the value chain, offering customers a packaged compute solution rather than standalone CPU IP.

These companies shift competition away from licensing economics and toward system-level performance, packaging, and availability. For SiFive, the risk is that RISC-V adoption accelerates, but value accrues to integrated chip suppliers rather than to independent IP licensors.

The largest long-term competitive pressure comes from internal silicon development at major technology companies.

Apple, Google, Amazon, Microsoft, and Meta all operate large in-house chip design teams. As RISC-V matures, these companies can choose to treat CPU development as a core internal capability rather than a vendor relationship. Acquisitions of RISC-V-focused startups and the expansion of internal CPU teams reduce the addressable market for external licensors like SiFive.

When a hyperscaler internalizes CPU design, SiFive loses both direct revenue and an ecosystem amplifier. These companies often influence tooling, compiler support, and platform standards, which can shape adoption patterns across the industry.

At the same time, internal teams do not eliminate SiFive’s opportunity. Many companies want differentiated silicon without maintaining large permanent CPU design organizations. In those cases, SiFive competes by offering production-proven cores, predictable schedules, and deep enablement that internal teams may struggle to replicate efficiently.

SiFive’s total addressable market expands as more companies decide to build custom silicon and as RISC-V becomes an accepted foundation for those efforts. The opportunity is not limited to replacing incumbent architectures in existing markets. It grows as new categories of compute adopt bespoke processors where flexibility, control, and long-term availability matter.

Growth in AI infrastructure increases demand for specialized hardware, but it also increases demand for general-purpose CPUs that manage data movement, orchestration, networking, and system control. As datacenter architectures move toward heterogeneous designs with accelerators, chiplets, and disaggregated systems, CPU selection becomes more modular.

This creates openings for alternative architectures in specific parts of the stack. Control-plane processors, storage controllers, networking appliances, and edge inference systems represent potential entry points where compatibility requirements are narrower than in general-purpose servers. Even partial adoption at hyperscalers or large infrastructure vendors can translate into substantial unit volumes once a design is standardized and deployed broadly.

SiFive’s opportunity in this segment depends on whether customers are willing to adopt RISC-V beyond experimental programs and into production infrastructure. Performance parity alone is not sufficient. Customers must also be confident in toolchains, software support, and long-term roadmap stability.

Embedded systems remain the most natural market for RISC-V adoption. These include industrial controllers, networking equipment, storage devices, and specialized appliances where workloads are well-defined and software stacks are tightly scoped.

These markets tend to favor customization, predictable behavior, and long product lifecycles. Once designed in, processors often remain in production for many years. That dynamic is favorable for royalty-based models, provided the IP supplier maintains long-term support and compatibility.

For SiFive, expansion in this segment depends on scale and repeatability. Winning many mid-sized programs across industrial and infrastructure customers can collectively rival the impact of a small number of marquee wins, while also diversifying revenue risk.

Automotive represents a structurally attractive expansion area because safety certification raises barriers to entry and increases customer willingness to pay for validated IP. As vehicles incorporate more advanced driver assistance systems, centralized compute architectures, and software-defined features, the number and complexity of processors per vehicle continues to grow.

Automotive design cycles are long, and qualification requirements are strict. Once a core is certified and adopted, it can generate volume for a decade or more across multiple vehicle platforms. This makes the segment particularly attractive for long-lived royalty streams.

SiFive’s ability to expand here depends on continued investment in safety-certified cores, documentation, and toolchain integration that meets automotive compliance standards. Competition includes both Arm-based solutions with deep automotive heritage and RISC-V vendors that focus specifically on safety-critical designs.

Aerospace and defense programs value architectural control, long-term availability, and supply chain resilience. These customers often operate on timelines measured in decades rather than product cycles. RISC-V’s open standard appeals in this context, particularly when paired with cores validated for high-reliability environments.

While unit volumes in these markets are smaller, pricing and margins can be attractive, and deployed systems tend to remain active for long periods. Successful entry here can provide stable, long-duration revenue streams and validate IP for use in other regulated industries.

Geopolitical considerations increasingly influence architecture decisions. Governments and companies in multiple regions are evaluating alternatives to reduce dependence on a small number of IP providers. This dynamic expands the number of programs that consider RISC-V as a default option rather than a niche choice.

However, this expansion does not automatically benefit SiFive. Local vendors, state-backed initiatives, and regionally aligned ecosystems can capture a significant share of this demand. SiFive’s advantage emerges when customers prioritize performance, certification depth, and ecosystem maturity over local sourcing alone.

Arm pricing pressure: Arm's expansion of its Flexible Access program directly targets SiFive's cost advantage by reducing upfront licensing fees for competing architectures. If Arm continues aggressive pricing while maintaining its software ecosystem advantages, SiFive may struggle to win new design opportunities despite RISC-V's open-source benefits.

Ecosystem maturity gap: RISC-V lacks the mature software tools, operating system support, and third-party IP ecosystem that Arm has built over decades. Customers face higher integration costs and longer development cycles when choosing RISC-V, limiting adoption to applications where the architecture's benefits clearly outweigh these disadvantages.

Customer concentration risk: With revenue concentrated among top semiconductor companies, SiFive faces significant exposure if major customers reduce RISC-V investments or develop internal alternatives. Intel's abandonment of its RISC-V project and Meta's acquisition of Rivos demonstrate how strategic shifts at large customers can eliminate substantial market opportunities.