Skip to research
stock3dby Fynca
Explore in 3D

COMPANY RESEARCH / 005930.KS

Samsung Electronics

Low-power DRAM for edge AI

Published by FyncaResearch checked How we research

HUMANOID ROBOT

Where it fits in the robot

Samsung’s LPDDR portfolio provides energy-efficient working memory for AI systems from edge devices to larger compute platforms.

What the evidence establishes

Adjacent technology

Samsung manufactures LPDDR memory. This component entry does not represent Samsung SDI batteries or the separate Rainbow Robotics business.

A relevant component capability. Its use in humanoids is an inference, not a disclosed customer relationship.

Component roleLow-power DRAM for edge AI
EvidenceSamsung manufactures LPDDR memory. This component entry does not represent Samsung SDI batteries or the separate Rainbow Robotics business.
Primary sourceSamsung Electronics — supporting disclosure
Company or official disclosure; vendor claims are attributed to their source.
Checked. This is the research review date, not the source publication or quote time.

Keep in perspective

This is the electronics parent, distinct from Samsung SDI. Memory capability is indirect robot exposure, not proof of a robot supply deal.

Diversified business exposure. This describes the breadth or role of the business; it is not an investment rating. A robotics product or development announcement does not by itself establish material humanoid revenue, market share, or future returns.

Connections to the assembly

These links explain the technology relationship. The model is illustrative and does not represent an actual manufacturer’s bill of materials.

Working memory & storage

Memory holds the working data used by perception and control; nonvolatile storage retains software and models. Capacity, bandwidth, and power consumption affect performance.

Scope of the company links: Memory manufacturers only. Micron explicitly discusses humanoids; Samsung and SK hynix are adjacent low-power memory exposure.

SMART GLASSES

Where it fits in smart glasses

Announced Android XR intelligent eyewear at Galaxy Unpacked (July 22, 2026, London): Snapdragon AR1 Gen 1, built-in camera, speakers and microphones, designs with Gentle Monster and Warby Parker, built with Google; scheduled to launch fall 2026 in select markets.

What the evidence establishes

Smart-glasses evidence

Disclosed: Samsung’s Galaxy Unpacked release names Snapdragon AR1 Gen 1, the camera, speakers and microphones, and its frame partners.

An explicit smart-glasses product, design-in, supply role, program, application page or partnership, including wrist or ring input sold with glasses. For a device maker, it means the company’s own glasses, not a supplier relationship. This does not establish revenue, market share, volumes or future sourcing.

Listed as a device maker for its own glasses, not as a supplier of the parts inside.

Component roleAndroid XR intelligent eyewear with Google
EvidenceDisclosed: Samsung’s Galaxy Unpacked release names Snapdragon AR1 Gen 1, the camera, speakers and microphones, and its frame partners.
How we knowDisclosed The company’s own filing, release, annual report, product page or investor record describes the smart-glasses role. Where customers are unnamed, the listing says so.
Primary sourceSamsung Electronics — supporting disclosure
Company or official disclosure; vendor claims are attributed to their source.
Checked. This is the research review date, not the source publication or quote time.

Keep in perspective

Announced for a fall 2026 launch in select markets, with no price or ship date published as of Sept 23, 2026. This is the electronics parent, distinct from Samsung SDI and Samsung Display.

Diversified business exposure. This describes the breadth or role of the business; it is not an investment rating. A smart-glasses product, design-in or partnership does not by itself establish material glasses revenue, market share, or future returns.

Connections to the assembly

This company is mapped at sector level. The reviewed evidence does not establish a direct match to a specific illustrated part. Explore device makers & platforms for the broader relationship.

AI SERVER RACK

Where it fits in an AI server rack

Says its HBM4 (11.7 Gbps) is designed for the NVIDIA Vera Rubin platform and in mass production, and showed HBM4E, SOCAMM2 (in mass production) and the PCIe 6.0 PM1763 SSD at NVIDIA GTC 2026. In Q2 2026 it scaled up HBM4 sales and shipped the first HBM4E samples to major customers (results, July 30, 2026). NVIDIA’s FY2026 10-K names Samsung as a memory supplier and a wafer foundry.

What the evidence establishes

AI-rack evidence

Disclosed: Samsung’s GTC 2026 release says its HBM4 is designed for NVIDIA Vera Rubin and in mass production; its Q2 2026 results report scaled-up HBM4 sales.

An explicit AI-server or AI-rack product, design-in, supply role, program, reference design or partnership, such as a part named for rack-scale AI systems or a product sold for AI data centers. For a rack builder, it means the company’s own AI servers or racks. This does not establish revenue, market share, volumes or future sourcing.

Component roleHBM4 for NVIDIA Vera Rubin, SOCAMM2 and data-center SSDs
EvidenceDisclosed: Samsung’s GTC 2026 release says its HBM4 is designed for NVIDIA Vera Rubin and in mass production; its Q2 2026 results report scaled-up HBM4 sales.
How we knowDisclosed The company’s own filing, release, annual report, product page or investor record describes the AI-server or AI-rack role. Where customers are unnamed, the listing says so.
Primary sourceSamsung Electronics — supporting disclosure
Company or official disclosure; vendor claims are attributed to their source.
Checked. This is the research review date, not the source publication or quote time.

Keep in perspective

Its HBM3E qualification for NVIDIA Blackwell-class GPUs (Sept 2025) is reported by Korean press and not confirmed by either company in the sources captured. HBM revenue is not disclosed within a diversified group.

Diversified business exposure. This describes the breadth or role of the business; it is not an investment rating. An AI-server product, design-in or partnership does not by itself establish material AI-rack revenue, market share, or future returns.

Connections to the assembly

These links explain the technology relationship. The model is illustrative and does not represent an actual manufacturer’s bill of materials.

HBM stacks

Towers of 8 to 12 DRAM dies joined by through-silicon vias and placed beside the logic dies. Eight stacks give one GPU hundreds of gigabytes and several terabytes per second of memory bandwidth.

Scope of the company links: Three companies make HBM, and NVIDIA’s FY2026 10-K names all three (SK Hynix, Micron, Samsung) as memory suppliers. Micron is mapped first because its release names GB300 NVL72 and GB200 NVL72 (and the AMD MI350 Series) for its HBM3E. SK hynix reports HBM4 mass shipments from Q2 2026 and a 56.4% HBM revenue share in Q1 2026 (its prospectus), without naming customers. Samsung says its HBM4 is designed for NVIDIA Vera Rubin and in mass production; its HBM3E qualification for NVIDIA (Sept 2025) is reported only. No source gives a supplier’s share of a specific GPU.

CPU memory

Low-power LPDDR5X packages or modules beside an Arm CPU, or DDR5 memory modules with interface chips in x86 trays. The GPUs can reach this memory through the CPU.

Scope of the company links: Grace-class CPUs use LPDDR5X (up to 17 TB per GB200 NVL72, NVIDIA). Micron says its SOCAMM LPDDR5X module was developed with NVIDIA for GB300 and was in volume production (Mar 18, 2025); SK hynix reports significantly growing SOCAMM2 sales for AI servers (Q2 2026), and Samsung markets SOCAMM2 with no named platform. x86 trays use DDR5 RDIMMs or MRDIMMs whose interface chips come from Montage (RCD, MRCD/MDB) and Rambus (RCD, PMIC, SPD hub; a SOCAMM2 chipset from Apr 22, 2026); Renesas is sampling a third-generation MRDIMM chipset, with production expected in 2H 2027.

E1.S NVMe SSDs

Flash drives in the slim EDSFF E1.S form factor: NAND flash, a controller and DRAM that hold the operating system, checkpoints and cached data. Reference trays carry four; some vendors fit up to eight in internal bays.

Scope of the company links: E1.S NVMe drives sit in each compute tray (NVIDIA lists four per GB200 tray). The rack-specific qualifications found are Sandisk’s DC SN861 E1.S (certified for GB200 NVL72 as Western Digital, Oct 14, 2024) and Micron’s 9550 and 7450 (on NVIDIA’s GB200 NVL72 recommended vendor list, per Micron). Kioxia (NX1 E1.S for GPU servers) and Samsung (PM1763 for AI infrastructure) position drives for AI servers, SK hynix reports expanding eSSD sales for AI servers, and its subsidiary Solidigm, which is not separately listed, shows a cold-plate-cooled SSD. Silicon Motion, Phison and Marvell sell SSD controllers with no named rack program. A vendor list shows eligibility, not volumes.

Related research

Companies with overlapping sector exposure. Their products and evidence may differ; this is not a list of equivalent investments.

Suggest a correction · Research and market-data disclosures