Trendforce reports that $AMD is actively trying to secure CW laser supply with multiple major procurement orders... Is probably just the start of the bottleneck? There's not much independent capacity in Western supply chains left other than $SIVE or $AAOI and maybe Macom. Especially after Lumentum/Coherent got locked up with multi-year agreements with Nvidia. (disclosure, own Sive and aaoi) Lumentum is already CW laser constrained and is likely buying off Japanese companies like Sumitomo/Furukawa if I had to guess per ER, and those are probably running at max capacity. From the Trendforce report, this is: "to ensure that its future capacity will not be constrained by NVIDIA and other major Cloud Service Providers (CSPs)." I wouldn't be surprised if other hyperscalers like Amazon, Microsoft, and others saw Nvidia / AMD signing LTAs, and are trying to secure capacity next. A lot of it is game theory on not getting choked out by competitors, and looks like AMD is tipping the first domino after Nvidia. But my opinion is that this just goes and show how invaluable this CW laser chokepoint is and the companies are inside it.

Per TrendForce data cited in a Bernstein report, global NAND makers have taken a harder line on mobile NAND price hikes in order to bring pricing in line with eSSD, while China's YMTC has not pushed for increases of the same magnitude — allowing it to pick up additional share.

Memory market update: - Counterpoint forecast a 20% QoQ increase for DRAM. - Trendforce are more conservative at 13-18% for conventional DRAM and 10-15% for NAND. - Morgan Stanley raised Q3 PC DRAM forecast to 15-20% increases QoQ from 3-8%. These aren't perfectly comparable baskets but the main conclusion is that tightness still persists. SK Hynix CEO and Chairman both confirmed supply tightness post-US IPO, potentially into the 2030s. - Samsung is reportedly seeking a Q3 DRAM ASP increase of up to 20% QoQ, with LPDDR increases potentially exceeding 20%. That would be their third consecutive quarterly hike after +90% in Q1 and +60% in Q2. This is above TrendForce's forecast but if price increases stick, $MU and SK hynix would likely follow. - TrendForce expects Q3 server DRAM contract prices to rise 13-18%. Multi year supply agreements limit price increases for some large volume US cloud customers. Which means customers without LTAs and incremental volumes sold outside those agreements would absorb the largest hikes. - SK hynix reportedly plan to scrap price caps in its new LTAs so that spot upside flows straight through, while $MU's SCAs carry floors and ceilings. - TrendForce estimates HBM will consume 22% of the top three suppliers' DRAM wafer input in 2026 while producing only 9% of DRAM bits. Those figures rise to 30% & 13% in 2027. This disproportionate wafer consumption is the main structural reason why conventional / server DRAM can remain scarce, even as capex increases. - On HBM4 itself, 2027 contract negotiations land around Q4'26 and Digitimes model HBM4 moving from ~$2/Gb in 2H26 toward $4-5/Gb next year. - Morgan Stanley suggest NAND remains undersupplied through 2027. - TrendForce expects mature SLC NAND pricing to rise 120-170% in 2H26, driven by shrinking mature-node capacity & MLC -> SLC migration. Positive read-through for $SNDK, Kioxia and $MU, while $SIMO offers controller leverage. - Worth noting Kioxia has filed for a US ADS listing, and $SNDK has raised NAND product pricing double digits while extending the JV with Kioxia to 2034. - $MU also recently broke ground on a ¥1.5 trillion Hiroshima expansion, targeting advanced DRAM & HBM shipments around summer 2028.

Well, looks like SLC NAND is forcasted to rise up 120-170% for H2 2026 per Trendforce. There's $MU (21%), Kioxia (20%) as largest share, but clearer beneficiaries appears to be: 1. Winbond (2344): ~15% of the SLC NAND market 2. Macronix (2337): ~11% market share 3. SkyHigh Memory via Puya Semi: ~14% market share Given Micron at $1T and others are a bit large relative to SLC NAND TAM. (source: Q1 2026 Winbond presentation/Trendforce).

🚀 Could #NVIDIA#'s next-gen chip specs change? Rubin Ultra's #HBM# configuration remains under review amid tight DRAM supply, while major CSPs weigh lower memory capacities.💡More analysis from #TrendForce#: 🔗

📢 AI computing demand remains strong. Capital spending by the world's nine largest CSPs is projected to rise about 90% YoY in 2026, surpassing US$886.7 billion.💡#TrendForce# on the latest AI server and CSP investment landscape: 🔗#Google# #Amazon# #Meta# #NVIDIA# #Microsoft# #Oracle# #ASIC#

📢 Taiwan's panel makers are reshaping their strategy. As #Innolux# and #AUO# divest aging factories, they're shifting from scale to high-value technologies such as Micro LED, AI, and CPO.💡More analysis from #TrendForce#: 🔗

👀 AI will continue to drive memory demand in 2027, keeping #DRAM# tight and prices firm. As #NAND# Flash capacity expands and consumer demand weakens, will second-half price corrections reshape the market? 💡More analysis #TrendForce#: 🔗

"In 2027, driven by strong demand from AI servers, DRAM bit demand growth will continue to outpace supply, extending the seller’s market and high-price environment." -Trendforce Structural Shift $MU $SKHY $DRAM