Sivers Semiconductors AB
Information Technology · Semiconductors & Semiconductor Equipment
Sivers Semiconductors sits at the intersection of three structurally growing markets - 5G mmWave densification, LEO satellite broadband, and automotive radar - each of which requires high-performance phased-array RF front-ends that Sivers has spent a decade perfecting.
The company's beamforming chipsets are among the few commercially available at 28 GHz and 60 GHz targeting the 5G infrastructure build-out, and operators densifying urban mmWave coverage need exactly this type of custom silicon. The LEO constellation boom (Starlink, OneWeb, Amazon Kuiper, Telesat Lightspeed) is a massive demand catalyst: each ground terminal requires a phased-array antenna module, and Sivers has announced design wins with multiple terminal makers.
Automotive radar, migrating from 24 GHz to 77/79 GHz for higher resolution ADAS, is another volume ramp where Sivers' mmWave expertise translates directly into silicon. The photonics unit adds a high-growth optionality leg - silicon photonics co-packaged with CPUs and GPUs is a hot area as hyperscalers push beyond copper interconnects, and Sivers' process IP positions it as a potential ASIC foundry partner.
Revenue per chip is modest today but unit volumes in 5G and automotive can be enormous; a single Tier-1 design win with a Nokia, Ericsson, or Aptiv could step-change the revenue trajectory. At sub-$100M market cap, the stock prices in nearly zero probability of large-scale adoption - making it a high-risk, high-asymmetry bet on mmWave penetration.
Sivers is a small-cap fabless company competing against RF giants - Qualcomm, MediaTek, NXP Semiconductors, and Infineon all have mmWave product lines with vastly superior sales forces, customer relationships, and balance sheets. 5G mmWave deployments outside dense urban US markets have been slow, and carriers globally have deprioritized mmWave in favor of mid-band 5G, compressing the near-term addressable market.
No major news in the last 7 days for SIVE - only listicles and opinion pieces, which we filter out by default. .
No key levels recorded for this ticker.
LEO terminal design wins are often announced without volume guarantees - design-in does not equal production ramp, and several well-funded mmWave terminal chip startups are chasing the same sockets. The company burns cash, and repeated equity raises have diluted shareholders significantly; without a large production contract, the cash runway is a recurring overhang.
The dual listing on Nasdaq provides US investor access but the primary price discovery happens in Stockholm in SEK, creating currency and liquidity fragmentation that suppresses institutional interest on the US-listed shares.
5B market cap (about 70x trailing sales) on the AI-photonics + space-defense narrative. The asymmetry has inverted, the stock now prices in large-scale adoption rather than near-zero. Both legs gained concrete validation. SPACE, pulling Sivers BFICs into a US defense prime serving Army / Navy.
(2) AI-datacenter photonics: a GlobalFoundries silicon-photonics collaboration (co-packaged + linear pluggable optics) and an O-Net / Enablence external-light-source partnership position the InP-laser leg for the co-packaged-optics buildout.
09/share of cash, the valuation now prices the ramp rather than the option, and the chart is parabolic (weekly RSI in the 90s, ~26x off the low). The structural story strengthened; the entry got expensive.