DDR5 Price Surge Bolsters HBM4 Price Hike Case as SK Hynix Q2 Earnings Approach
N.R. Finch
DDR5 prices jumped roughly 40% in a single quarter, outpacing HBM gains and handing SK Hynix real ammunition for HBM4 contract talks later this year — with Q2 earnings set to break records, the memory profit engine is shifting from HBM-only to a broader mix.
Why is DDR5 suddenly rising faster than HBM?
LS Securities estimates SK Hynix's commodity DRAM averaged $9.30 per GB in Q1 and $13.00 in Q2 — a jump of roughly 40%. HBM averaged $12.80 to $14.30 over the same period, up only about 12%.
This means → the 2026 pattern is the reverse of 2025. Last year HBM pulled the blended average higher; this year conventional DRAM is the faster mover.
In plain terms = HBM is still climbing steadily, but DDR5 is sprinting — the gap between the two is closing fast or already flipping.
What is driving the structural squeeze?
HBM and conventional DRAM share the same wafer-fab capacity, and producing a given amount of HBM consumes roughly three times the wafer area of an equivalent DDR5 output.
This means → every wafer redirected to HBM tightens DDR5 and DDR4 supply further — the pricing logic is self-reinforcing.
Retail DDR5 prices rose 123% in 2025 and are forecast to climb another 45% in 2026. A 32 GB DDR5-6000 kit that cost $80–100 early in 2025 now runs $364–529.
Xiaomi says per-device DRAM costs could rise 25% in 2026; some manufacturers are already trimming memory specs in laptops and smartphones.
What does this mean for HBM4 pricing?
Rising conventional DRAM prices directly lift the opportunity cost of allocating wafers to HBM — in plain terms = making DDR5 instead of HBM now pays much better, so HBM buyers must offer more.
This means → when HBM4 contract talks begin late this year, SK Hynix has a concrete basis for demanding higher unit prices — not a negotiating bluff, but a reflection of what the same wafer could earn elsewhere.
Tech commentator Aakash Gupta estimates Nvidia accounts for roughly 90% of SK Hynix's HBM output. SK Hynix sold out its 2025 quota by mid-2024, locked all 2026 orders by mid-2025, and is already pricing 2027 supply.
Samsung vs. SK Hynix — who gains more?
Samsung has greater exposure to conventional DRAM and is expected to capture more of the DDR5 price upside.
About 50% of SK Hynix's revenue sits in long-term contracts where price adjustments lag, diluting the lift from spot DDR5 gains.
In plain terms = SK Hynix's HBM business is a locked-in, long-term book — stable by design, but that stability means it doesn't ride short-term price spikes as fully.
How strong will Q2 earnings be?
A Yonhap Infomax consensus of 14 brokerages projects SK Hynix Q2 revenue at ₩84.1 trillion and operating profit at ₩64.1 trillion — both would top the Q1 record.
Samsung has already flagged preliminary Q2 revenue of ₩171 trillion and operating profit of ₩89.4 trillion, up 129% and 1,810% year-on-year respectively.
This means → the two firms' combined operating profit would exceed ₩153 trillion, and the growth engine has shifted from HBM alone to HBM + DDR5 + NAND firing together.
What is the key marker for the second half?
SK Hynix reports Q2 results on July 29; Samsung follows on July 30. The market expects SK Hynix's Q2 operating margin could top 75%.
This reflects a profit expansion driven primarily by higher selling prices, not volume growth.
Whether HBM4 negotiations can convert this round of DDR5 inflation into higher long-term contract prices is the single most important pricing event of the second half.
Content is for reference only, not financial advice.