Why SSPPU exists: the apportionment problem
Patent damages must compensate only for the value of the patented invention — not for the value of unpatented features, the manufacturing efficiencies of the accused infringer, or the goodwill of the accused product's brand. When a patented feature is just one component among hundreds in a complex product (a smartphone, laptop, or car), using the entire product revenue as the royalty base and applying even a small royalty rate produces a damages figure that reflects much more than the patent's actual contribution. The smallest salable patent-practicing unit (SSPPU) principle addresses this problem by requiring that the royalty base start from the smallest component that actually practices the patent — not the finished product. Cornell University v. Hewlett-Packard (N.D.N.Y. 2009) is the seminal SSPPU case: Judge Rader (sitting by designation) excluded damages testimony based on HP's entire server processor revenue, requiring apportionment to the processor's instruction dispatch unit — the actual component practicing the patent — which was a fraction of the server's cost.
LaserDynamics v. Quanta: SSPPU in the Federal Circuit
The Federal Circuit endorsed and developed the SSPPU principle in LaserDynamics, Inc. v. Quanta Computer (Fed. Cir. 2012). LaserDynamics' patent covered a method for optical disc identification (how an optical disc drive identifies whether a disc is a DVD or CD). The patent was practiced only in the optical disc drive — a component that represented about 2% of the laptop's total cost. LaserDynamics sought to use the entire laptop price as the royalty base, arguing the drive was inseparable from the laptop. The Federal Circuit rejected this: 'it is generally required that royalties be based on the smallest salable patent-practicing unit, with further apportionment where the patent does not account for all the value of the component.' The court required LaserDynamics to use the optical disc drive (not the laptop) as the royalty base and further apportion if the drive's value exceeded the patented function. LaserDynamics is the foundational SSPPU precedent for complex consumer electronics.
The relationship between SSPPU and the Entire Market Value Rule
EMVR allows the entire accused product to be the royalty base, but only if the patented feature drives consumer demand for the whole product. SSPPU and EMVR are in tension — using EMVR without sufficient justification is exactly the error SSPPU is designed to prevent. After Uniloc USA v. Microsoft (Fed. Cir. 2011), courts treat EMVR as a narrow exception, not a default. The Uniloc court rejected use of the entire Microsoft Office revenue as the royalty base for a feature patent on product activation, holding that EMVR requires evidence that consumers purchased the entire product because of the patented feature. Most modern patent cases in electronics, software, and other complex technologies do not qualify for EMVR — patented features rarely drive demand for the entire product. Instead, patent owners typically argue for an appropriate SSPPU combined with a higher royalty rate that reflects the component's importance, or use comparable license rates that already embed apportionment.
Further apportionment beyond SSPPU
Even after identifying the SSPPU, the damages analysis may require further apportionment if the SSPPU contains significant unpatented value. VirnetX v. Cisco Systems (Fed. Cir. 2014) established this principle clearly: identifying the SSPPU (accused software modules) was not the end of the analysis. The accused modules implemented both patented and unpatented features, so the damages rate had to be further apportioned to reflect only the patented functionality. Methods for further apportionment within the SSPPU: (1) technical apportionment by a technical expert who identifies the patented feature's functional contribution to the SSPPU as a percentage of the SSPPU's overall functionality; (2) consumer surveys assessing what portion of consumer willingness-to-pay for the SSPPU is attributable to the patented feature; (3) income approach: estimate what share of the SSPPU's profit margin derives from the patented feature versus other features. The result is an 'effective royalty rate' on the SSPPU that properly isolates the patent's value.
Comparable licenses as an alternative to SSPPU analysis
An important alternative to SSPPU-based apportionment is the use of comparable licenses for the same technology or similar technologies. If real parties have negotiated licenses for the same patent (or comparable patents covering the same feature), those royalty rates embed the market's actual apportionment of value to the technology. CSIRO v. Cisco Systems (Fed. Cir. 2015) held that comparable licenses can satisfy the apportionment requirement without a separate SSPPU analysis — if the license rates are for the same technology and reflect actual market valuation. The logic: if parties negotiating at arm's length agreed to a rate of X% of component revenues (or a flat per-unit fee), that rate already accounts for the technology's contribution. Courts accept comparable licenses as meeting apportionment when the licenses are sufficiently comparable — technically (covering similar or the same feature) and economically (similar scope, field of use, and market conditions). In practice, comparable licenses and SSPPU are often used together — the SSPPU narrows the base, and comparable licenses inform the appropriate rate.
Practical implications for patent litigation
SSPPU has fundamentally changed how patent damages are litigated for complex products. For patent owners: (1) identify the SSPPU early in litigation and select the component with the highest revenue that can be justified as practicing the patent; (2) use comparable license rates that were negotiated based on the SSPPU or equivalent component; (3) consider whether EMVR is supportable for specific high-value, feature-driven patents where consumer demand for the product genuinely depends on the patented feature; (4) obtain a technical expert to explain the patented feature's role in the SSPPU. For defendants: (1) argue for a smaller SSPPU by identifying smaller separable components that practice the patent; (2) move to exclude plaintiff's damages expert under Daubert if apportionment is inadequate; (3) challenge the technical comparability of plaintiff's license comparables; (4) present your own damages expert with a properly-apportioned base and rate. Daubert motions targeting inadequate apportionment are among the most commonly-filed and commonly-granted pretrial motions in patent cases.
Frequently Asked Questions
What is the smallest salable patent-practicing unit (SSPPU)?
The smallest salable patent-practicing unit (SSPPU) is a damages apportionment principle that limits the royalty base in patent infringement cases to the smallest separately salable component that incorporates the patented feature. The concept was developed by the Federal Circuit to enforce the broader apportionment requirement — the rule that reasonable royalty damages must reflect only the value of the patented technology, not the value of unpatented features or the broader product in which the patented technology is incorporated. For example, if a patent covers a specific feature in a smartphone chip, and the chip is separately sold, the SSPPU would be the chip (not the smartphone). The patent owner should select the chip's revenue as the royalty base and apply an appropriate rate — rather than using smartphone revenue and a smaller rate, which could improperly capture smartphone value unrelated to the patent. The SSPPU principle comes primarily from Cornell University v. Hewlett-Packard (N.D.N.Y. 2009, Judge Rader sitting by designation) and was endorsed by the Federal Circuit in LaserDynamics v. Quanta Computer (Fed. Cir. 2012) and subsequent cases.
What is the Entire Market Value Rule (EMVR) and how does it relate to SSPPU?
The Entire Market Value Rule (EMVR) is the counterpart to SSPPU: it allows the use of the entire accused product's revenue as the royalty base, but only when the patented feature is the basis for customer demand for the entire product. SSPPU and EMVR represent opposite ends of the damages base spectrum. SSPPU: narrow base (just the practicing component), higher rate. EMVR: broad base (entire product revenue), lower rate. In theory, both should produce the same damages figure if applied correctly — the base and rate together must reflect the patent's actual value contribution. In practice, courts strongly prefer SSPPU over EMVR for complex products. The Federal Circuit in Uniloc USA v. Microsoft (Fed. Cir. 2011) sharply limited EMVR, holding that it 'cannot be used for support, as it may skew the damages horizon for the jury' unless the patented feature truly drives demand. After Uniloc, the accepted practice for multi-component products is to begin with SSPPU and then further apportion if even the SSPPU contains substantial value from non-patented features.
Does SSPPU always set the royalty base, or can further apportionment be required?
SSPPU is a floor, not a ceiling. Even after limiting the royalty base to the SSPPU, further apportionment may be required if the SSPPU itself contains substantial unpatented value. In VirnetX v. Cisco Systems (Fed. Cir. 2014), the court held that the accused software modules (the SSPPU) included both patented and non-patented functionality, and damages must be further apportioned to reflect only the patented portion. In CSIRO v. Cisco Systems (Fed. Cir. 2015), the court allowed use of royalty rates from comparable licenses for the same technology without further SSPPU analysis — because those licenses already reflected an appropriate apportionment of value to the patented technology. The key principle: the final royalty must isolate the patented technology's contribution to value. Whether accomplished by a narrow royalty base (SSPPU), a further-apportioned rate, or comparable licenses that already embed apportionment, the result must be economically grounded in the patent's specific value contribution.
How do standard-essential patents (SEPs) affect SSPPU analysis?
Standard-essential patents (SEPs) — patents that must be practiced to implement a technical standard (like Wi-Fi, LTE, or Bluetooth) — raise special SSPPU and apportionment issues. For SEPs, the SSPPU is typically the component that implements the standard (e.g., the modem chip). However, the royalty must be further apportioned to reflect only the contribution of the specific patent among all patents in the standard. FRAND (fair, reasonable, and non-discriminatory) licensing commitments apply to SEPs, and FRAND rates are determined using a modified Georgia-Pacific analysis. Key cases: Ericsson v. D-Link (Fed. Cir. 2014) — the SSPPU for Wi-Fi SEPs is the Wi-Fi chip, not the entire device; the royalty rate must reflect the SEP's contribution to the standard and the standard's contribution to the chip; comparable license rates used by SEP holders provide the most reliable evidence. In SEP cases, damages experts often use a 'top-down' approach: start from an estimated aggregate royalty for all SEPs in the standard, then calculate the asserted patent's proportional contribution.
What happens when no smallest salable unit exists — when the patented feature is inseparably part of a larger product?
When the patented feature is inseparably integrated into a larger product (there is no separately salable component that contains just the patented functionality), SSPPU analysis must adapt. Courts have recognized that SSPPU cannot always be applied mechanically when the patent covers a feature embedded directly in an end product with no intermediate component. In such cases, damages experts typically use: (1) a technically-apportioned royalty rate — a technical expert apportions the patented feature's contribution to the overall product's functionality or value, then applies that percentage to product revenues; (2) consumer surveys establishing what portion of consumer willingness-to-pay is attributable to the patented feature; (3) comparable licenses for similar technologies if those licenses already reflect apportionment. The 2016 Federal Circuit case Power Integrations v. Fairchild Semiconductor clarified that even without a separable SSPPU, parties must use some form of apportionment to tie the royalty base and rate to the patented technology's contribution. Expert testimony on apportionment is rigorously scrutinized — Daubert motions to exclude damages experts for inadequate apportionment analysis are common.