Test report DSG-7240 · Rev C · tested October 12, 2026
Memory & StorageDevice under test
SK hynix hands Onto Innovation partner award for HBM defect detection
SK hynix has given Onto Innovation a partner award for defect-detection work on HBM, the stacked DRAM used in AI accelerators. The award marks a qualified-toolset win.
- Read
- 3 min
- Words
- 534
- Node
- 28nm
- Operator
- Grace Kim
Spec summary
- SK hynix awarded Onto Innovation a partner award for HBM defect detection
- Onto Innovation was formed in the 2019 merger of Nanometrics and Rudolph Technologies
- HBM is the stacked DRAM category driving AI accelerator memory demand
- SK hynix is the leading HBM supplier
SK hynix has named Onto Innovation a partner award winner, citing the U.S. metrology and inspection vendor's work on defect detection for high-bandwidth memory (HBM, stacked DRAM modules used mainly in AI accelerators).
The award, announced via StreetInsider, signals that Onto's inspection systems have become part of the qualified toolset inside SK hynix's HBM production lines. For a chipmaker, supplier awards of this kind typically follow sustained delivery performance, tool uptime, and measurable yield gains — not one-off purchases.
Onto Innovation sells process control equipment: systems that find and classify defects on wafers during manufacturing. In HBM, that job is unusually hard. The memory stacks multiple DRAM dies on top of each other, connected through thousands of vertical conduits called TSVs (through-silicon vias). A defect in one die, or a particle trapped during stacking, can kill the entire package.
That's why inspection matters more in HBM than in conventional DRAM. Each added layer multiplies the number of interfaces where faults can hide, and the finished module sells at a steep premium to standard memory. Catching a bad die before it gets bonded saves the good dies stacked alongside it.
Why does an award from SK hynix matter?
SK hynix is the leading supplier of HBM, the memory category that GPU makers — above all NVIDIA — need in volume for AI training and inference hardware. Demand has run ahead of supply through the AI buildout, and HBM makers have raced to add capacity and move to next-generation stack heights.
For an equipment vendor, a partner award from the top HBM producer works as a public qualification. It tells other memory makers — Samsung, Micron, and the fast-growing Chinese entrants — that Onto's tools performed where the scrutiny was toughest. Equipment orders in semiconductors follow exactly this kind of track record.
What does this change for Onto Innovation?
Commercially, it consolidates Onto's position in the fastest-growing corner of the memory market. HBM capacity expansion means more wafers inspected, more tools shipped, and more recurring service and software revenue on the installed base. Process control intensity — spending on inspection per wafer — runs higher for advanced packaging than for standard fab flow, which benefits specialists.
It also anchors Onto in advanced packaging more broadly. HBM is the flagship application, but the same die-stacking and bonding steps spread into logic products as chipmakers package compute and memory together. Tools proven on HBM stacks carry credibility into those programs.
The competitive backdrop
Onto competes in wafer inspection against larger rivals, among them KLA, which dominates process control overall. Recognition from SK hynix sharpens Onto's pitch that its systems hold their own at the leading edge — specifically in the bonded-wafer and stacked-die inspection steps that HBM demands.
For SK hynix, the award is routine supplier management made public. For Onto's investors and customers, it's confirmation that the company's HBM toolset is inside the industry's most-watched production ramp.
Onto Innovation, headquartered in the United States, was formed from the 2019 merger of Nanometrics and Rudolph Technologies. The company did not disclose award terms, and SK hynix has not published details of the toolset covered.
via Google News: HBM memory (Source)
More from Grace Kim
Show full bio
Market editor covering regulation, policy and their impact on hardware layer of artificial intelligence at Die Signal.
285 articles
Same lot · LOT-C1C8
- DSG-263065nmSK hynix honours Onto Innovation for HBM defect detection work
- DSG-876928nmSK hynix Names Onto Innovation Best Partner for HBM Defect Prevention
- DSG-40957nmSK Hynix Taps Intel EMIB-T for Next-Gen HBM Packaging
- DSG-212845nmNiche Memory Chip Bet Puts SK Hynix Ahead of Samsung in Value
- DSG-67193nmSK Hynix Eyes Opportunity Inside Intel's $100B U.S. Fab Plan