5G chips and modules power phones, wearables, IoT, and vehicles
Rising 5G adoption across industrial IoT, connected automotive, and smartphones is driving demand for application-specific 5G chips and modules. Chipmakers are increasingly targeting 5G-Advanced, satellite non-terrestrial network (NTN) connectivity, 5G fixed wireless access (FWA), and 5G New Radio (NR).
Many of these 5G chips and modules deliver 3rd Generation Partnership Project (3GPP) Release (R) 17 and R18 features with backward compatibility, and now support R19 for 5G-Advanced, while setting the stage for 6G development. Chipmakers also target 5G RedCap (reduced capability) for cost-effective 5G devices in IoT applications.
Many of these solutions focus on higher integration and flexibility options to simplify development as well as reduce power consumption and footprint. Part of the integration effort also includes AI to enhance performance.
Here is a sample of 10 5G chips and modules introduced over the past year with higher performance that drive 5G adoption across a range of applications.
Modems and platforms
Qualcomm Technologies Inc. launched the X105 5G Modem-RF, featuring the industry’s first 3GPP R19-ready modem, which lays the foundation for 6G development and testing. The Qualcomm X105 5G-Advanced platform improves data speeds and power efficiency while claiming breakthroughs in quad-band GNSS and integrated NR-NTN support. Its integrated fifth-generation AI processor uses agentic AI in the modem to detect, classify, and optimize data traffic types based on user scenarios.
Qualcomm claims several industry firsts with the X105 platform. It features the first 6-nm RF transceiver, achieving up to 30% lower power consumption and 15% smaller board footprint than the previous generation; the first quad-frequency (L1, L2, L5, L6) GNSS engine with integrated multi-constellation support; and improved location accuracy.
The 5G Modem-RF also supports video calling, video streaming, data, voice, and messaging over NR-NTN. It delivers 14.8-Gbits/s peak download speeds while achieving up to 13.2 Gbits/s with sub-6-GHz bands. Qualcomm said it addresses the need for more uplink capacity and data speeds, achieving 4.2-Gbits/s peak uplink throughput.
With an R19-ready architecture, the X105 Modem-RF system is designed to accelerate 5G-Advanced adoption across smartphones, FWA, mobile broadband, automotive, XR, PCs, robotics, and industrial IoT.
Broadcom Inc. announced an integrated 5G and Wi-Fi 8 FWA platform in partnership with Samsung Electronics Co. Ltd. The broadband-optimized reference platform integrates Broadcom’s BCM6776 Wi-Fi 8 system-on-chip (SoC) with Samsung’s B1320 5G modem. Claimed as an industry-first platform, it combines 3GPP R17 connectivity with the emerging Wi-Fi 8 (IEEE 802.11bn) standard.
The Broadcom BCM6776 is a highly optimized, tri-band Wi-Fi 8 SoC. Key features include a quad-core Arm network processor with support for two-stream 40-MHz channels in the 2.4-GHz band and four-stream 160-MHz channels in the 5- and 6-GHz bands, delivering a complete Wi-Fi 8 platform in a single chip.
Other features include integrated 2 × 2 2.4-GHz and 4 × 4 5-GHz and 6-GHz Wi-Fi 8 MAC/PHY/radio functionality, which simplifies system design and lowers cost; an integrated multi-gig PHY; and on-chip 2.4-GHz power amplifiers and support for third-generation digital pre-distortion, which reduces external components and improves RF efficiency. Dual PCIe Gen3 controllers enable simultaneous tri-band applications with a single additional chip.
The platform reduces bill-of-materials costs, simplifies board design, and reduces active power consumption by 50% compared with previous generations, according to Broadcom.
The Samsung B1320 is a highly integrated, power-efficient, 5-nm chipset solution that supports 3GPP R17. It can achieve 5G downlink speeds of 3.43 Gbits/s and uplink speeds of 1.17 Gbits/s. It supports a 4Rx/2Tx radio chain. The 5G modem with a quad-core Arm CPU and the BCM6776’s network processor can support full Wi-Fi to 5G throughput. The platform also includes a Power Class 1.5 RF transceiver, power management, and a GNSS receiver. It also supports NR-NTN and NB-NTN for satellite communications.
Qualcomm also delivers 5G to wearables with its Snapdragon Wear Elite Platform. The personal AI platform delivers personal, always-on, intelligent wearable computing devices. It works across WearOS by Google, Android, and Linux with a neural processing unit (NPU) for on-device AI and an advanced suite of ultra-low-power connectivity solutions.
Claiming a first‑of‑its‑kind, multimode connectivity architecture, the Snapdragon Wear Elite Platform integrates six advanced technologies:
- 5G RedCap delivers advanced, low-power cellular connectivity for connected intelligent applications.
- Micro‑power Wi‑Fi enables always‑on Wi‑Fi connectivity at significantly lower power for intelligent wearables, according to Qualcomm, which require continuous AI context synchronizing, richer data exchange, and seamless coordination with nearby devices and cloud services.
- Bluetooth 6.0 enables proximity‑aware interactions.
- UWB provides secure, precise, proximity‑based interactions.
- GNSS delivers advanced AI processing for precise location context.
- NB‑NTN enables satellite‑based, two‑way messaging and critical communications from intelligent wearables when cellular and Wi‑Fi coverage are unavailable.
Delivering advanced power efficiency and faster speed, the platform delivers a 5× improvement on single-core CPU performance and up to 7× faster GPU for app launching, multitasking, and smoother rendering. This enables the Snapdragon Wear Elite to run complex, 2 billion–parameter AI models locally, faster, and more securely.
The advanced power management enables 30% more battery use than the previous generation. When recharging is needed, rapid charging powers a device up to 50% in approximately 10 minutes.
5G SoCs and RF chips
In late 2025, MediaTek introduced its Dimensity 9500 SoC, delivering new levels of on-device AI, console-class gaming, and power efficiency. According to MediaTek, the 9500 is designed to “supercharge the next wave of flagship 5G smartphones.”
The Dimensity 9500 features a third-generation All Big Core CPU design combining a 4.2-GHz ultra core, three premium cores, four performance cores, and four-lane UFS 4.1 storage. It delivers up to 32% higher single-core and 17% higher multicore performance than the previous generation, while the ultra core achieves up to 55% lower power consumption at peak performance, for longer battery life and greater productivity, and claims up to 30% more power efficiency while multitasking in games and social audio call apps.
The platform also features a new cache and memory architecture, claiming the industry’s first support for four-channel UFS4.1. It doubles read/write speeds and accelerates large AI model loading by 40%. In addition, the second-generation Dimensity Scheduling Engine delivers smoother responsiveness and sustained efficiency even under heavy workloads, according to MediaTek.
The Dimensity 9500 integrates the Arm G1-Ultra GPU, providing up to 33% higher peak performance and 42% improved power efficiency, and introduces higher frame rate interpolation, up to 120-fps raytracing, for console-level, quality gaming.
MediaTek said integrating the ninth-generation MediaTek NPU 990 with Generative AI Engine 2.0 doubles compute power and introduces BitNet 1.58-bit large model processing, which reduces power consumption by up to 33%. The Dimensity 9500 is also the first integrated compute-in-memory architecture for its newly added Super Efficient NPU.
Doubling its integer and floating-point computing capabilities enables 100% faster, 3 billion–parameter LLM output; 128K token long text processing; and 4K ultra-high-definition image generation, while reducing peak-performance power consumption by 56%, according to the company.
The Dimensity 9500, with the MediaTek Imagiq 1190, supports RAW-domain pre-processing, up to 200-MP capture, 30-fps continuous focus tracking, a new portrait engine, and cinematic 4K 60-fps portrait video. It also provides the latest MiraVision Adaptive Display technology.
The platform’s connectivity suite includes Bluetooth call enhancement, Wi-Fi fast transfer, and multi-network intelligence. The AI-powered communication technologies are reported to deliver up to 10% lower power in 5G and 20% lower power with Wi-Fi, while 5CC carrier aggregation increases bandwidth by 15%. In addition, AI positioning and network selection technologies offer 20% higher accuracy, as well as 50% lower network latency thanks to AI congestion prediction, according to the company.
Qorvo Inc. recently introduced a family of RF switches for 5G radios, designed to simplify multiband radio architectures. The family spans from 50 MHz to 10 GHz. These devices reduce component count, improve signal integrity, and enable more efficient RF system design across 5G infrastructure, as well as industrial, drone, and test applications.
Designers face increasing challenges maintaining isolation and signal integrity without adding size, loss, and complexity, as 5G radios expand to support wider bandwidths and more frequency bands, Qorvo said. Many designs rely on cascaded switch architectures or multiple narrowband devices, which increase insertion loss, degrade linearity and signal integrity, and add board space and design effort, according to the company.
The wideband, high-isolation RF switch family addresses these challenges, eliminating the need for cascaded switches in 5G radios. “This approach reduces insertion loss, maintains signal linearity, and simplifies design, improving receiver performance in applications such as digital pre-distortion feedback,” Qorvo said.
The family comprises three devices:
- The QPC6144 for high-isolation switching is a SP4T wideband switch that delivers greater than 65-dB isolation in a single device, eliminating the need for cascaded switches.
- The QPC6122 (SP2T) for compact wideband switching and QPC6188 (SP4T) for flexible wideband routing provide wideband absorptive switching across 50 MHz to 10 GHz, enabling a single platform approach to RF routing. These devices reduce component count and simplify design. They also maintain low insertion loss and strong linearity across wide bandwidths for calibration paths, general signal routing, and multiband operation.
Sivers Semiconductors announced earlier this year the general availability of its highly integrated 7- to 15-GHz Alpine0715 (Daybreak) multi-beamforming transmit/receive front-end IC for emerging 5G/6G FR3 applications and multifunction defense arrays. The FR3 frequencies are used for 5G-Advanced and 6G network deployments, delivering sub-6-GHz and wider bandwidths at mmWave frequencies.
The Alpine0715, housed in a 7.95 × 7.0-mm WLCSP package, offers high linear output power, low noise figure, high efficiency, and integrated beam-table memory in a single device, the company said, and the reconfigurable multi-beamforming architecture enables smaller, lower-cost antenna arrays
Featuring four dual-polarization TX and four dual-polarization RX elements, the Daybreak beamformer IC delivers 27-dB gain and 20-dBm OP1dB TX performance and 32-dB single-channel gain and <3-dB NF RX performance.
5G modules
At electronica India 2026, Cavli Wireless announced two 5G modules targeting automotive, IoT, and M2M applications. The AQ62 5G Autograde is an automotive-grade 5G NR sub-6-GHz module designed for next-generation telematics and connected vehicle applications, and the CQM211 5G NR Advanced sub-6-GHz module, compliant with 3GPP R16, is designed for IoT and M2M applications.
The CQM211 5G NR module delivers multi-gigabit 5G connectivity with LTE Cat 16 fallback and flexible integration options for new and existing hardware platforms. The module targets high-bandwidth applications such as Industry 4.0 equipment, smart surveillance systems, routers, gateways, mobile Wi-Fi, and other edge applications requiring high-speed connectivity.
Key specs include a Cortex-A7 processor clocked at up to 1.8 GHz, 2× carrier aggregation over NR, dual-band GNSS, a PCIe Gen3 interface, and DBI display support. It is available in LGA and M.2 form factors.
The AQ62 5G Autograde module for connected automotive applications delivers an embedded software environment for developing advanced applications such as telematics control units, ADAS systems, OBD solutions, and hybrid tracking applications. The AQ62, built on Yocto Linux Kernel 5.15 with SDK support, provides high-speed 5G NR connectivity with LTE Cat 18 fallback. It supports 3GPP R16.
The 5G NR automotive module features a quad-core Arm Cortex-A55 processor, L1+L2/L5 GNSS, and dual-SIM dual-standby support. The module also provides optional NG-eCall and NTN band support. The AQ62 is available in five regional variants: worldwide (WW), North America (NA), Eurasia and Japan (EAJ), India (IN), and Eurasia (EA).
Designed to accelerate Android-based IoT applications, Quectel Wireless Solutions announced two Android 16 smart modules: the SE505FE entry-level 5G smart module and the SH602FA high-performance 4G smart module. The modules provide integrated wireless connectivity and multimedia capabilities in compact form factors.
Here, we’ll focus on the 5G module, which provides Android-powered 5G connectivity to a wider range of IoT applications. The module is built on the MediaTek MT8863T chipset with a 2× Cortex-A76 and 6× Cortex-A55 processor architecture and Mali-G57 GPU. It supports 5G sub-6-GHz and LTE Cat 4, together with Wi-Fi 6, Bluetooth 5.2, and GNSS. Typical applications include AR/VR devices, digital signage, smart gateways, tablets, industrial PDAs, robotics, and intelligent video platforms.
The SE505FE series modules support multiple-input/multiple-output technology. Quectel said the concurrent use of multiple antennas at the receiver end on the same frequency band minimizes bit error rate and optimizes data speed, which improves communication quality.
The module supports multi-constellation GNSS (GPS, BDS, GLONASS, Galileo, and QZSS) receiver for fast and accurate positioning and offers a range of interfaces, including LCM, camera, touch panel, microphone, UART, USB, I2C, and SPI. It supports cameras up to 32 MP at 30 fps and 1080 × 2520 video recording and playback at 120 fps.
The module is housed in a 42.5 × 56.2 × 2.95-mm LGA package. The SE505FE series is available in three variants: SE505FE-EM, SE505FE-NA, and SE505FE-WF (supports Wi-Fi and Bluetooth only).
Quectel also expanded its mid-tier 5G family with RedCap 3GPP R17 variants: the RM255C-GL for global markets and the RG255C-NA for North America.
5G RedCap has emerged as the technology to address the mid-tier connectivity gap between high-speed eMBB and narrowband IoT as 5G expands into industrial and IoT domains, Quectel said.
The RM255C-GL is a 5G sub-6-GHz RedCap module in the M.2 form factor, developed in accordance with 3GPP R17. It supports peak data rates of up to 223-Mbits/s downlink and 123-Mbits/s uplink, with 5G sub-6-GHz standalone operation and LTE Cat 4 fallback. It is backward-compatible with R15 and R16.
The RM255C-GL is pin-to-pin-compatible with Quectel’s RM5x0N and RM50xQ 5G series and both pin-to-pin and size-compatible with selected LTE-A and 4G modules, including the EM06, EM12 series, EM160R-GL, and EM05, which helps simplify the migration from LTE to 5G RedCap.
The module integrates Qualcomm IZat Gen 9VT location technology with optional multi-constellation GNSS support (GPS, GLONASS, BDS, Galileo, QZSS, and NavIC) for fast and accurate positioning. It also supports a range of interfaces, including USB 2.0, PCIe 2.0, and PCM, and offers USB drivers for Windows 10/11, Linux, and Android, extending the module into a wider range of RedCap applications.
The RG255C-NA 5G RedCap sub-6-GHz, housed in an LGA module, is designed in compliance with 3GPP R17. Purpose-built for industrial-grade applications, it supports key 5G capabilities including URLLC and network slicing. The module delivers peak data rates of up to 223-Mbits/s downlink and 123-Mbits/s uplink. It operates in 5G sub-6-GHz standalone mode with LTE Cat 4 fallback and is backward-compatible with R15 and R16.
The module is compatible with Quectel’s compact EG2x 4G module series, simplifying design upgrades while meeting mid-speed application requirements for higher capacity, low latency, and enhanced reliability in industrial and commercial deployments, the company said.
The RG255C-NA also integrates Qualcomm IZat Gen 9VT location technology, supporting multi-constellation GNSS including GPS, GLONASS, NavIC, BDS, QZSS, and Galileo. The integrated GNSS receiver streamlines product development and provides fast and accurate positioning, Quectel said.
The module also features a suite of industry-standard interfaces, including USB 2.0, PCIe 2.0, PCM, UART, SGMII, and SPI, along with USB driver support for Windows 10/11, Linux, and Android.
The RM255C-GL and RG255C-NA have completed the required regulatory and industry conformance certifications. The RG255C-NA is certified to FCC, IC, and PTCRB standards, while the RM255C-GL has achieved CE, RCM, FCC, IC, GCF, and PTCRB certifications.
The post 5G chips and modules power phones, wearables, IoT, and vehicles appeared first on EDN.
