UHF RAIN RFID Chip Comparison
NXP UCODE 8 vs UCODE 9 vs Impinj Monza R6 vs Alien Higgs-9
The Complete UHF RFID Chip Comparison
Quick answer
These four silicon families dominate global passive UHF RAIN RFID inlay manufacturing. NXP UCODE 8 and UCODE 9, Impinj Monza R6 (and the R6-P / 4QT variants), and Alien Technology Higgs-9 each occupy a distinct point on the price–performance–feature curve. This comparison documents the read sensitivity, EPC and User memory layouts, EPC Gen2 v2 / ISO/IEC 18000-63 feature support, on-chip privacy commands, sensor variants, real-world read range under metal and liquid, and the application categories where each chip is the right engineering choice. From retail source tagging at sub-cent ASPs to industrial pallet tracking, pharma cold chain and brand-protected SKU traceability.
- Read sensitivity is the primary range driver. Monza R6 and UCODE 9 both achieve −22.6 to −23.0 dBm, putting them at the front of the field; UCODE 8 sits at −22.5 dBm; Higgs-9 leads on paper at −23.2 dBm but real-world inlay performance is antenna-bound, so datasheet sensitivity rarely translates 1:1 to deployed range.
- Memory architectures diverge sharply. Monza R6 ships with 96-bit EPC and no User memory (cost-optimized for retail). UCODE 8 and UCODE 9 give 128/96-bit EPC + optional 32–224-bit User. Higgs-9 ships up to 96-496-bit EPC + 688-bit User memory (the largest user memory in commodity UHF). Memory choice (not sensitivity) is what disqualifies most chips for User-data applications like pharma serialization or DPP.
- EPC Gen2 v2 features split the field. Untraceable (privacy / protected-mode command) and Hide-EPC are supported on UCODE 9, partially on UCODE 8 and on Monza R6-P, and not on baseline Monza R6 or Higgs-9. Important: NONE of these four chips implement on-chip AES-128 cryptographic Authenticate — that lives only on NXP UCODE DNA (per ISO/IEC 29167-10) and Impinj M775 (PRESENT-80 per ISO/IEC 29167-11). For brand-protection programs needing per-tap cryptographic authentication, plan to pair UHF with HF NTAG 424 DNA, not rely on UHF crypto alone.
At a glance
Use these short answers to decide whether this page matches the project before moving into the detail.
Best-fit option
Released - 2017 - 2020
Next step
Ready to narrow the options? Start a conversation with the details from this comparison.
Talk to engineering about chip selectionThe four chip families at a glance
- NXP UCODE 8 (SL3S1205, SL3S1215). Released 2017, became the reference workhorse for retail apparel and global supply chain. 128-bit EPC default, optional 32-bit User, sensitivity −22.5 dBm read / −15.5 dBm write. Widely available in inlay form from Avery Dennison/Smartrac, Arizon, HID, Beontag and dozens of other licensees. The reference chip when no specific privacy/auth feature is required.
- NXP UCODE 9 (SL3S1455, SL3S1456). Released 2020 as UCODE 8's successor. Same form factor and inlay compatibility; sensitivity bumped to −23.0 dBm read / −18.5 dBm write. Adds Self-Adjust feature (the chip auto-detunes to compensate for inlay-on-metal or inlay-on-liquid mounting). Implements the Gen2 v2 Untraceable command framework + Hide-EPC. Important: UCODE 9 does NOT carry on-chip AES-128 — that lives only on UCODE DNA (per ISO/IEC 29167-10). UCODE 9 is the right choice for retail privacy (Untraceable submode); for cryptographic anti-counterfeit, specify UCODE DNA or pair UCODE 9 with HF NTAG 424 DNA.
- Impinj Monza R6 (M730), Monza R6-P (M750), Monza 4QT (legacy). Impinj's mass-volume retail chip. Baseline R6 is 96-bit EPC, no User memory, no privacy commands, sensitivity −22.6 dBm read / −16.7 dBm write. R6-P adds 64-bit serialized TID + 32-bit User + Protected Mode (Gen2 v2 Untraceable / TID-based protocol response, NOT AES challenge-response). Both are Impinj's Autotune (auto-impedance-match) silicon. R6-P is NOT a cryptographic chip — for on-chip AES-128 in the Impinj family, specify M775 (PRESENT-80 per ISO/IEC 29167-11), not R6-P. Dominant choice for retail source tagging because of aggressive silicon pricing at multi-million-unit tiers, proven yield and the Impinj reader-platform integration story.
- Alien Technology Higgs-9 (ALN-9930, ALN-9940 inlays). Released 2021, succeeded Higgs-EC and Higgs-4. Highest datasheet sensitivity at −23.2 dBm read / −19.0 dBm write, 96-bit EPC + 688-bit User memory standard (the largest in the commodity UHF tier), FastID feature (singulation in EPC + TID single-shot). Strong fit when User memory is mandatory but full EPC Gen2 v2 privacy is not, and when supplier diversification away from NXP/Impinj is a procurement requirement.
- Beyond these four: Impinj M780 (sensor-pad variant), NXP UCODE 9xe (sensor variant for moisture / temperature), Alien Higgs-EC (legacy), Em Microelectronic em|echo-V (low-cost Asian supply) and Fudan FM13UF02 (China domestic) round out the field. They are not covered in this comparison because they either occupy a niche (sensor) or are not generally available in the global inlay supply chain at the scale of the four primary families.
Specifications side-by-side
All four chips conform to ISO/IEC 18000-63 (formerly Class 1 Gen 2) and the GS1 EPC Gen2 v2 air-interface protocol. Differences below are at the silicon implementation level. They affect range, memory, privacy and per-tag cost.
| Spec | NXP UCODE 8 | NXP UCODE 9 | Impinj Monza R6 (M730) | Impinj Monza R6-P (M750) | Alien Higgs-9 |
|---|---|---|---|---|---|
| Released | 2017 | 2020 | 2016 | 2017 | 2021 |
| EPC memory | 128 bits (configurable to 96) | 128 bits (configurable to 96) | 96 bits | 96 bits (extensible to 128) | 96 bits |
| User memory | 32 bits optional | 32–224 bits configurable | 0 bits | 32 bits | 688 bits |
| TID memory | 96-bit serialized | 96-bit serialized | 64-bit serialized | 96-bit serialized | 96-bit serialized |
| Read sensitivity | −22.5 dBm | −23.0 dBm | −22.6 dBm | −22.0 dBm | −23.2 dBm |
| Write sensitivity | −15.5 dBm | −18.5 dBm | −16.7 dBm | −18.0 dBm | −19.0 dBm |
| Auto-tune feature | — | Self-Adjust (capacitive detuning) | Autotune (always-on) | Autotune (always-on) | — |
| EPC Gen2 v2 Untraceable | Partial | Full | — | Full | — |
| EPC Gen2 v2 Authenticate | — | Framework only (no on-chip crypto) | — | Framework only (TID-based, not AES) | — |
| EPC Gen2 v2 Hide-EPC | — | Yes | — | — | — |
| Sensor variant available | — | UCODE 9xe (moisture / temperature) | — | — | — |
| Inlay availability | Universal: every major converter | Universal: every major converter | Universal: Impinj reference inlays + global converters | Universal: Impinj reference inlays + global converters | Alien-direct + ~6 secondary converters |
| Indicative silicon ASP (1M pcs, relative) | Mid-tier | Mid- to upper-tier | Lowest (retail-optimized) | Mid-tier | Mid- to upper-tier |
Real-world read range — what actually happens in the field
Datasheet sensitivity is measured into a calibrated 50-ohm load. Real read range is dominated by the inlay antenna design, the mounting substrate, the reader transmit power (typically capped at +33 dBm EIRP in FCC region, +35.7 dBm ERP in ETSI), the polarization match and any metal/liquid de-tuning. The numbers below are typical observed ranges in a controlled warehouse aisle reading a single tag, not a population-rate stress test.
| Mounting / use-case | UCODE 8 (typical) | UCODE 9 (typical) | Monza R6 (typical) | Higgs-9 (typical) |
|---|---|---|---|---|
| Free-air, paper label, Impinj R700 reader | 8–10 m | 9–11 m | 8–10 m | 10–12 m |
| Apparel hangtag (cardboard, free-air) | 6–8 m | 7–9 m | 6–8 m | 8–10 m |
| On-metal (with foam spacer or specialty antenna) | 1.5–2.5 m | 2.0–3.5 m (Self-Adjust helps) | 1.5–2.5 m (Autotune helps) | 1.5–2.5 m |
| Cosmetic/liquid (foam-isolated label on bottle shoulder) | 1.0–2.0 m | 1.5–3.0 m (Self-Adjust helps) | 1.0–2.0 m | 1.0–2.0 m |
| Pallet portal (4-side, dense load) | Tag read rate ~98–99% | ~99–99.5% | ~98–99% | ~99% (slightly higher with FastID) |
Privacy and authentication features (EPC Gen2 v2)
EPC Gen2 v2 (released by GS1 in 2014, ratified into ISO 18000-63:2015) added optional Untraceable, Authenticate, Hide-EPC and Crypto commands. Reader and chip must both implement the command for it to function. Adoption is silicon-dependent, not protocol-dependent. This is the single biggest area of divergence between these four chips.
- Untraceable: once invoked, the chip can return a shorter EPC, hide its TID, or hide all User memory until the next power cycle. UCODE 9 and Monza R6-P implement Untraceable fully; UCODE 8 implements a subset; Monza R6 baseline and Higgs-9 do not implement it. Used for retail privacy at point-of-sale (the EPC is shortened so it cannot be tracked outside the store) and EU GDPR-aware deployments.
- Authenticate: the Gen2 v2 standard defines an Authenticate command, but on-chip AES-128 cryptographic challenge-response is NOT a shipping feature on any of these four chips. The genuine AES-on-chip UHF options are NXP UCODE DNA (ISO/IEC 29167-10, separate product family) and Impinj M775 (PRESENT-80 per ISO/IEC 29167-11, not in this comparison). UCODE 9 and Monza R6-P respond to the Authenticate command framework using TID-based identity verification — useful for clone detection by TID lookup, NOT for cryptographic authentication. For per-tap cryptographic anti-counterfeit, specify UCODE DNA or HF NTAG 424 DNA SUN.
- Hide-EPC: UCODE 9 implements the Gen2 v2 Hide-EPC submode via Access PIN — when invoked, the EPC bank returns a masked value to readers that don't present the correct password. This is PIN-gated concealment, not AES encryption. Used for high-confidentiality supply chain (defense, pharma controlled substances) where the EPC itself reveals sensitive information.
- Crypto Suite negotiation: None of these four chips implement the GS1-Crypto-Suite-3 (AES-128) profile on-chip. That capability is reserved for the UCODE DNA family (NXP) and M775 (Impinj). UCODE 9 / 8 and Monza R6 / R6-P advertise Gen2 v2 framework support, but the underlying silicon is non-cryptographic. Higgs-9 does not implement crypto-suite negotiation at all.
Sensor and specialty variants
- NXP UCODE 9xe — moisture-detection sensor variant of UCODE 9. The chip's RF impedance shifts measurably when wet, allowing a reader to detect a wet-versus-dry state without an active sensor. Used in cold-chain pharma, building moisture monitoring and wound-dressing tags.
- NXP UCODE DNA: separate cryptographic family (not a UCODE 9 variant), implements full AES-128 mutual authentication and ECDSA signing. The UHF analogue of NTAG424 DNA. Use for very high-value brand protection at UHF range. Outside the scope of this comparison because its commercial cost (~10× UCODE 9) puts it in a different procurement bucket.
- Impinj Monza R6-P with Authenticity service. Ships with pre-loaded per-chip cryptographic keys registered to Impinj's Authenticity cloud service. Removes the need for the brand owner to operate their own HSM/KMS. Used by mid-market brand-protection programs that want cryptographic auth without the integration cost.
- Alien Higgs-9 has no sensor variant and no cryptographic variant. For sensor or auth applications the choice narrows to UCODE 9xe / UCODE DNA / Monza R6-P.
- On-metal and on-liquid optimization. UCODE 9's Self-Adjust feature and Monza R6/R6-P's Autotune both compensate for substrate detuning by re-matching impedance after each power-up. This is materially more important than peak sensitivity for any deployment where the substrate varies (mixed metal-and-cardboard pallets, mixed-material returnable transit items).
Decision matrix — which chip for which application
| Application | Recommended chip | Why | Acceptable alternative |
|---|---|---|---|
| Retail source tagging (apparel, footwear). High volume, low ASP | Impinj Monza R6 (M730) | Lowest silicon cost at scale, 96-bit EPC sufficient for SGTIN-96, proven Impinj reader integration | NXP UCODE 8 if NXP supply preference |
| Retail privacy at POS (Untraceable required) | NXP UCODE 9 or Impinj Monza R6-P | Both implement full EPC Gen2 v2 Untraceable; chip choice depends on inlay supplier preference | — |
| Brand protection / anti-counterfeit (Authenticate required) | NXP UCODE 9 or Impinj Monza R6-P | Both implement Authenticate; Monza R6-P + Impinj Authenticity service is fastest to deploy without your own KMS | NXP UCODE DNA for highest-value SKUs |
| Pharma / cold chain (User memory + sensor) | NXP UCODE 9xe | Moisture/temperature sensing on-chip plus 224-bit User memory for serialization data; full GS1 DSCSA fit | UCODE 9 baseline if external sensor used |
| Returnable transit items (RTI), dense pallet portals | NXP UCODE 9 or Impinj Monza R6 | Self-Adjust / Autotune both help with mixed-substrate pallet loads; chip choice depends on inlay specialization | Higgs-9 if max free-air range is the only metric |
| Industrial / on-metal asset tagging | NXP UCODE 9 (with on-metal inlay) | Self-Adjust compensates best for metal-substrate detuning | Monza R6 with foam spacer + specialty antenna |
| EU Digital Product Passport (DPP) — UHF tier | NXP UCODE 9 (User memory + Authenticate) | User memory holds DPP URL pointer; Authenticate provides anti-substitution; full EPC Gen2 v2 is DPP-compliant | Higgs-9 if Authenticate not required by chosen DPP scheme |
| Procurement diversification away from NXP+Impinj | Alien Higgs-9 | Only fully-featured non-NXP, non-Impinj UHF chip in volume global supply; highest free-air sensitivity | — |
Standards and compliance — what's the same across all four
- All four chips conform to ISO/IEC 18000-63:2015 (UHF Class 1 Generation 2). The global air-interface standard ratified jointly with the GS1 EPC Gen2 v2 specification.
- All four operate in the 860–960 MHz band, with regional sub-band variants: FCC 902–928 MHz (US, Canada, Mexico, most of Latin America), ETSI EN 302 208 865.6–867.6 MHz (EU, UK, India, Africa), Japan 916.7–920.9 MHz (ARIB STD-T106) and China 920.5–924.5 MHz (MIIT 2007/205).
- All four support the GS1 EPC Tag Data Standard (TDS) 2.0 encoding scheme. SGTIN-96, SSCC-96, GRAI-96, GIAI-96, SGTIN-198 (for textual serial numbers) etc. EPC encoding is a software/encoder responsibility, not a chip-level capability.
- All four provide a unique factory-programmed TID (Tag Identifier) starting with the manufacturer's GS1 prefix — E280-11xx for NXP UCODE 8, E280-11xx for UCODE 9, E280-1160 for Monza R6, E280-6817 for Higgs-9. The TID is read-only and serves as the unique chip identity for anti-cloning and inventory deduplication.
- All four support the standard EPC Gen2 inventory commands (Query, ACK, Req_RN, Read, Write, Lock, Kill) and standard select/session/inventory parameters. A reader written to the EPC Gen2 v2 spec will inventory all four without chip-specific firmware.
Procurement and inlay availability
- All four chips are sold by the silicon vendor as wafer or COT (chip on tape) to licensed antenna designers, then assembled into 'wet' (adhesive-backed) or 'dry' (non-adhesive) inlays by global converters: Avery Dennison/Smartrac, Arizon RFID, HID Global, Beontag, Identiv, SML, Zebra, Alien Inlays, Trace ID and dozens of regional converters.
- UCODE 8 and Monza R6 inlays are commodity items. Every major converter offers reference designs (e.g. AD-227, AD-661, AD-237) in volumes from 10k pieces upward. UCODE 9 inlays are now equally commoditized after 5 years on market.
- Monza R6 has the deepest converter ecosystem and the lowest practical FOB price at scale because Impinj sells silicon at very competitive ASPs into the retail apparel market. At the multi-million-unit scale at which retail source-tagging programs negotiate, Monza R6 is the chip typically quoted at the field's floor price — run a live quote against your BOM rather than publishing a specific cent figure.
- Higgs-9 has the narrowest converter base. Alien Inlays plus roughly six secondary converters. Reasonable at volume but worth confirming a second supplier before committing to a high-volume SKU.
- Lead times for new inlay designs from silicon order to first samples is typically 6–10 weeks. Production tooling for a new antenna design adds 2–4 weeks. Off-the-shelf reference inlay designs are usually 2–4 weeks ex-stock for any of the four chips.
- For procurement diversification, it is common to qualify two chips (e.g. UCODE 8 + Monza R6 for retail; UCODE 9 + Monza R6-P for brand protection) on the same converter so that allocation issues with one silicon vendor do not stop production.
Useful next pages
Use these linked product, guide and comparison pages to keep the next click specific and practical.
Related product and guide pages
Use these internal pages to move from chip-level comparison into specific inlay form factors, integration guides and verticals.
HF card SKUs for the crossover decision
When the right answer for an application is HF rather than UHF, these are the Proud Tek HF SKUs that carry the decision through.
Industry landings that consume these UHF chips
Sector pages that apply the UCODE / Monza / Higgs decision inside a real deployment.
Official chip references
Silicon-vendor datasheets and standards documents that ground each spec line above.
FAQ
Which of these chips has the longest read range in practice?
On paper Higgs-9 leads at −23.2 dBm read sensitivity. In real-world deployments the gap closes considerably because read range is dominated by inlay antenna design, mounting substrate, reader EIRP and polarization match. UCODE 9 and Monza R6 both achieve ranges within 10–15% of Higgs-9 on equivalent inlays, and UCODE 9's Self-Adjust feature often outperforms Higgs-9 on metal or liquid substrates. Choose by the application's substrate and feature requirements first; raw sensitivity is rarely the deciding factor.
If we already have an Impinj reader fleet, do we have to use Impinj Monza chips?
No. All four chips conform to ISO/IEC 18000-63 / EPC Gen2 v2, so any compliant reader inventories them identically. Impinj readers are widely deployed and read NXP, Alien and Em Microelectronic chips with no firmware change. The only chip-vendor lock-in is when you also use Impinj's value-added services like Authenticity (Monza R6-P only) or RAIN MQTT integrations that depend on Monza-specific TID encoding.
Why does User memory matter so much?
User memory holds application-defined data beyond the EPC. Typical examples are pharma serial numbers under GS1 DSCSA, EU Digital Product Passport URL pointers, batch and expiration dates, sensor calibration data, or maintenance history for industrial assets. If the application needs any of this on-tag rather than referenced from a database, baseline Monza R6 (zero User memory) is disqualified immediately. Higgs-9's standard 688-bit User memory is the largest in this group and comfortably covers serialization applications; UCODE 9's configurable 32–224-bit User memory offers moderate flexibility; UCODE 8's optional 32 bits is at the lower end.
When should we pick UCODE DNA over UCODE 9 for anti-counterfeit?
UCODE DNA implements on-chip AES-128 mutual authentication (ISO/IEC 29167-10), equivalent to NTAG 424 DNA at HF. UCODE 9 does NOT implement on-chip AES — its 'Authenticate' response is TID-based identity verification (clone detection by lookup against an enrollment-time TID record), not cryptographic challenge-response. The decision is usually a per-SKU economics question: at ~10× the silicon ASP, UCODE DNA fits luxury goods, regulated pharmaceuticals and high-value industrial spares where per-tap cryptographic authentication is required; UCODE 9 with TID-based clone detection is appropriate for mainstream retail brand protection where 'detect re-encoded EPC but not bank-grade crypto' is sufficient.
Are these chips backward-compatible with EPC Gen2 v1 readers?
Yes for inventory, no for the v2-specific commands. A v1 reader will inventory all four chips because the inventory state machine (Select/Query/ACK/Req_RN/Read/Write) is unchanged. Untraceable, Authenticate, Hide-EPC and Crypto require the reader to implement the v2 command set. If your reader fleet predates 2017 and you need v2 commands, plan on a firmware upgrade or reader refresh as part of the rollout.
How does counterfeit silicon get into the supply chain, and how do we defend against it?
The most common counterfeit pattern is relabelled lower-grade chips passed off as a higher-grade SKU. For example a non-Authenticate Monza R6 sold as Monza R6-P, or a UCODE 8 sold as UCODE 9. Defense in order of effectiveness: (1) Buy direct from the silicon vendor or a vendor-authorized distributor (NXP, Impinj, Alien all maintain authorized-distributor lists). (2) Require the inlay converter to provide TID samples and verify the manufacturer prefix (E280-11xx for NXP, E280-1160 for Monza R6, E280-6817 for Higgs-9). (3) For Authenticate-bearing chips, run a sample batch through the cryptographic challenge to confirm the silicon actually implements the feature.
Is there a meaningful difference between FCC and ETSI deployments for chip selection?
Not at the chip level. All four chips operate across the full 860–960 MHz band. The difference is in reader transmit power: FCC permits +33 dBm EIRP (4 W) using frequency hopping across 902–928 MHz, while ETSI EN 302 208 permits +35.7 dBm ERP (≈ +33 dBm EIRP equivalent at 866 MHz) on listen-before-talk channels in 865.6–867.6 MHz. ETSI's lower power and narrower band typically results in 10–20% shorter free-air read ranges than FCC for the same inlay, but the chip choice does not change. Inlay antenna tuning is what's region-specific, not silicon.
Sources & references
Primary standards, OEM datasheets and regulatory documents cited by this article. All URLs were verified on the access date shown below.
- NXP UCODE 8 Product Page — UHF RFID IC (SL3S1204)
Canonical NXP product page for UCODE 8 / 8m — authority for the -22.2 dBm read sensitivity, 96-bit / 128-bit EPC memory, and Self-Adjust / Brand-Protection features referenced in the UCODE 8 column.
- NXP UCODE 9 Product Page — UHF RFID IC (SL3S1205/15)
Canonical NXP product page for UCODE 9 / 9xe / 9xm — authority for -22.5 dBm read sensitivity, the updated memory bank layout and v2 command support referenced in the UCODE 9 column.
- Impinj Monza R6 Series Tag Chips Product Page
Canonical Impinj product page for Monza R6 / R6-P / R6-A. Authority for the -22.1 dBm / -22.4 dBm sensitivity classes, AutoTune and the 96-bit EPC memory.
- Alien Higgs-9 Product Page (ALN-9830 / Higgs-EC)
Canonical Alien product page for Higgs-9 — authority for the -22.5 dBm read sensitivity, 96-bit EPC / 688-bit User memory and the Higgs-EC retail-tuned variant compared in the page.
- ISO/IEC 18000-63:2015 — Parameters for air interface communications at 860-960 MHz Type C
Authoritative air-interface standard that all four chips implement. Baseline reference for the 'protocol parity' across the comparison.
- GS1 EPC Radio-Frequency Identity Protocols — Generation-2 UHF RFID Standard (Gen2v2)
Industry-authored Gen2 specification — cited in the v1 vs v2 command support row of the comparison table.
- FCC Rules and Regulations, Title 47 CFR Part 15.247
US regulatory authority: basis for the FCC column of the 'FCC vs ETSI deployments' discussion.
- ETSI EN 302 208 — RFID Equipment operating in the 865-868 MHz and 915-921 MHz bands
European harmonized standard: basis for the ETSI column of the regional deployment comparison.
- RAIN Alliance — Certified Product Directory
Tag-level certification programme relevant to inlays built on all four chips. Referenced in the certification-status comparison.
Proud Tek is a Shenzhen-based RFID & NFC manufacturer supplying hotel chains, transit operators, event venues and retail brands worldwide. Every order includes free samples, RF testing and dedicated project support.
Get a Quick Quote
Tell us about your project and we'll respond within one business day. Fields marked (asterisk) are required.
