SK hynix: Co-Author and Potential ShareholderNarrow moat
Sandisk (SNDK) — moat facet
SK hynix co-wrote the specification for Sandisk's most important future product and holds convertible bonds over a stake in the company that makes its present ones.
SK hynix co-authored the specification for Sandisk's most important future product, and holds convertible bonds over a stake in the company that manufactures its present ones.
Both halves are documented. On 3 August 2026 Sandisk and SK hynix released the first High Bandwidth Flash technical specification through the Open Compute Project, six months after the consortium formed, with Google and Tenstorrent among the contributors.1 SK hynix is not a licensee here; it is the co-author, and whether HBF becomes an industry standard or a single-vendor product depends substantially on it.
Separately, SK hynix holds convertible bonds that can convert into substantially all the voting rights of an entity owning 14.19% of Kioxia, and its chief executive has said the company is "very carefully looking at how we can co-develop the NAND flash market for our customers and suppliers."2
So the second-largest NAND producer is simultaneously Sandisk's partner in defining a new memory class and a potential shareholder in Sandisk's manufacturing partner. SK hynix also owns Solidigm, which competes directly in enterprise SSDs, and holds 19.3% of NAND to Sandisk's 12.4%.3
There is a precedent worth remembering, covered on Kioxia's pages: SK hynix's veto as a Kioxia investor is what blocked the merger that would have consolidated this industry in 2023.
Rated narrow, and genuinely two-sided. A co-author gives HBF a chance of becoming a standard; a shareholder in your fab partner gives it leverage over your supply.
Watch whether SK hynix converts. It is the single ownership change that would most alter Sandisk's position, and Sandisk would have no vote in it.
SK hynix co-authored the first HBF specification through the OCP in August 2026, which gives Sandisk's most important future product a second major backer. The convertible-bond position over a Kioxia stake points the other way and has not been exercised.
Datacenter SSDs are where SK hynix, through Solidigm, competes most directly; the share holding up is the contest being won.
- ReportedOn 3 August 2026 Sandisk and SK hynix released the first High Bandwidth Flash technical specification through the Open Compute Project, six months after the consortium formed, with Google and Tenstorrent among the contributors. SK hynix is not a licensee here; it is the co-author, and whether HBF becomes an industry standard or a single-vendor product depends substantially on it.Coverage of Sandisk's 2026 Investor Day of 13 August 2026 and its High Bandwidth Flash programme. Sandisk has taped out the first High Bandwidth Flash memory die, shown on an HBF Roadmap slide next to a die photograph, with the first HBF inference product samples listed as coming soon in 2027; a tapeout means the design is finished and committed to a mask set, and is several steps short of a product, with wafers still to return from the fab, yields to climb, and the die to survive stacking into 8-high and 16-high configurations with a working logic die and controller, followed by thermal and endurance qualification, accelerator software work and customer qualification cycles. Per Sandisk's HBF fact sheet, the first generation targets 512GB per stack built from sixteen 256Gb die at 1.6 TB/s of read bandwidth, which Sandisk claims lands at up to 8 to 16 times the capacity of HBM at a similar cost, in a package closely matching HBM4's footprint, stack height and power profile, and which is non-volatile and spends no power on refresh; a second generation targets more than 2 TB/s and up to 1TB per stack at 0.8 times the first generation's power, and a third pushes past 3.2 TB/s and up to 1.5TB per stack at 0.64 times the power. The Investor Day deck frames the target workloads as mixture-of-experts LLMs, long context lengths and large KV caches, with the architecture developed using input from major cloud and AI customers, and lays out three deployments: HBF augmenting HBM around an xPU, replacing HBM stacks outright in a similar footprint, or sitting disaggregated holding decode weights and KV cache while a smaller HBM tier acts as cache. Sandisk showed an inference token output comparison from which it draws an 8x capex efficiency claim (one HBF GPU against eight HBM GPUs) and a 2x GPU efficiency claim (four HBF GPUs delivering the same token output as eight HBM GPUs), both labelled as based on internal testing; the fact sheet adds a simulation result of HBF landing within 2.2% of unlimited-capacity HBM when reading pretrained weights for Llama 3.1 405B. These are vendor numbers on unreleased silicon and the chart carries no axis values. On timing, in August 2025 Sandisk said the first HBF samples were targeted for the second half of calendar 2026 with AI inference devices expected in early 2027; as of the August 2026 Investor Day the first HBF inference product samples are listed as coming soon in 2027, and mass production is reported for 2028. On 3 August 2026 Sandisk and SK hynix released the first HBF technical specification through the Open Compute Project, six months after the consortium formed, with Google and Tenstorrent among the contributors. — August 2026 · publ. 2026-08-18 · source ↗
- ReportedSeparately, SK hynix holds convertible bonds that can convert into substantially all the voting rights of an entity owning 14.19% of Kioxia, and its chief executive has said the company is "very carefully looking at how we can co-develop the NAND flash market for our customers and suppliers." So the second-largest NAND producer is simultaneously Sandisk's partner in defining a new memory class...Reporting on Kioxia's 27 August 2026 investment plan and SK hynix's position. Kioxia and Sandisk plan to invest $31.4 billion in new NAND capacity over the next six years, according to Kioxia chief executive Hiroo Ota, including $11.3 billion in a new fabrication plant at Kitakami; Ota said Kitakami will become the primary manufacturing hub for the partnership's cutting-edge NAND chips and expects significant growth there, and Kioxia hopes the Japanese government will subsidise approximately one-third of expansion costs. Separately, SK hynix holds convertible bonds that can convert into substantially all the voting rights of BCPE Pangea Cayman 2, which owns 14.19% of Kioxia; SK hynix chief executive Kwak Noh-Jung said the company had no fixed plans regarding the stake but that it is very carefully looking at how it can co-develop the NAND flash market for its customers and suppliers. — August 2026 · publ. 2026-08-27 · source ↗
- ReportedSK hynix also owns Solidigm, which competes directly in enterprise SSDs, and holds 19.3% of NAND to Sandisk's 12.4%. There is a precedent worth remembering, covered on Kioxia's pages: SK hynix's veto as a Kioxia investor is what blocked the merger that would have consolidated this industry in 2023.TrendForce coverage of the Kioxia-Sandisk alliance in the AI NAND market, January 2026. Data from TrendForce shows that in Q3 2025 Samsung led with a 32.3% NAND market share, followed by SK hynix at 19.3%, Kioxia at 15.3% — surpassing Micron — and Sandisk at 12.4%. Sandisk and Kioxia have maintained a partnership spanning over 25 years, and even after Western Digital spun off Sandisk in 2025 the collaboration has strengthened; the two operate the world's largest NAND flash production sites in Japan, including the Yokkaichi and Kitakami fabs. On the technology front they co-developed BiCS FLASH 3D NAND, now in its eighth generation at 218 layers, with production of the tenth generation at over 300 layers set to begin in 2026 and Kioxia planning to repurpose its recently opened Kitakami K2 fab for the new node; BiCS8's 218-layer TLC 3D NAND features 35% higher cell current, 60% faster NAND I/O and a 50% boost in bit density through CMOS directly bonded to array and On Pitch SGD technology. Through the joint ventures the two share the costs of expensive semiconductor equipment and R&D, achieving economies of scale to compete with the South Korean memory giants. There is a clear split in market focus: Kioxia mainly supplies NAND to major Japanese and global electronics makers, while Sandisk commands the consumer storage segment and holds a strong position in enterprise SSDs across North America and overseas markets; despite deep integration on the manufacturing side, the two continue to compete directly in channels and branded end products. ETNews reported that Sandisk planned a 100% NAND price hike in the new year, and Kioxia's market value topped 10 trillion yen on 27 January 2026, just over a year after its December 2024 IPO. — 2025-2026 · publ. 2026-01-29 · source ↗
- Sandisk Corporation Form 10-K, FY2026 (SEC EDGAR)
- Kioxia's $31.4bn plan and SK hynix's convertible position (August 2026)
- Sandisk tapes out its first High Bandwidth Flash die, targets 2027 samples