Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

Saturday, October 10, 2026

Back to today

From X

@convequity

Convequity on X

High-bandwidth memory generally costs about 5x more per GB than conventional DRAM. As AI systems support longer context windows and more complex reasoning, memory demand is spreading into conventional DRAM and NAND. The mechanism is a tiered memory system. HBM holds active model weights and the context needed now. External DRAM holds a broader pool of context that the model can retrieve, while NAND-based SSDs store less frequently used context and model checkpoints. This keeps more context accessible at a lower cost, with a latency trade-off. Memory producers are operating at approximately 80% gross margins and 60%+ EBIT margins. The industry's through-cycle average gross margin has been 35-40%, and past peaks rarely exceeded 55-60%. Supply explains much of this. Samsung, SK Hynix and Micron invested heavily in capacity during the 2021-2022 boom. That capacity came online in 2023 and 2024, when smartphone, PC and non-AI server demand was weak. All three cut production, recorded operating losses and set conservative capex plans. Samsung cut DRAM output by approximately 15-20% in 2023. The capacity response as of mid-2026: 1. Samsung has targeted a roughly 50% increase in HBM production capacity this year. 2. SK Hynix has accelerated its HBM4 ramp through the M15X fab, while reportedly redirecting some planned HBM4 capacity toward DDR5. 3. $MU has accelerated DRAM and HBM output in Taiwan. Meaningful wafer starts from its new US and Singapore facilities are not expected until 2027 and beyond. These additions are weighted toward HBM and high-end server DRAM, not broad conventional memory capacity. They appear insufficient to close the gap in the near term, so the unusually high margins look likely to persist for several more quarters.
@StableBread

Fajasy on X

Half the gold, platinum, and palladium in ore from McCreedy West, the producing mine of Magna Mining $MGMNF (TSX: $NICU), goes to Franco-Nevada $FNV for $1,200 per gold-equivalent ounce, 29.0% of the average gold price in Q4 2025. Magna inherited the deal, a precious metals stream, when it bought McCreedy West from KGHM $KGHPF, the Polish copper giant. A stream is a deal where a financier pays a miner upfront for the right to buy part of its future metal at a discount to market. Magna’s stream gives Franco-Nevada 50% of the gold, platinum, and palladium in ore from McCreedy West and “certain areas” of Levack and Podolsky, until the deposits run out or 2048. Franco-Nevada pays 60% of the monthly average gold price per gold-equivalent ounce (platinum and palladium converted into gold) from McCreedy West, capped at $1,200/oz once gold tops $2,000. In Q2 2026, the stream cut net revenue by US$3.9M, so revenue would have been ~18% higher without the stream. CEO Jason Jessup called the stream “a bit of an overhang” on McCreedy West, and in April said its terms are “definitely open for discussion” with Franco-Nevada. The higher gold climbs, the more the stream costs Magna, and I wouldn’t count on Franco-Nevada softening the terms. Crean Hill, Magna’s permitted nickel mine, carries no Franco-Nevada stream, and Jessup says R2, the high-grade zone Magna found at its flagship Levack mine, has none either. At Q4 2025’s gold price, Magna kept 64.5% of the value of McCreedy West’s gold, platinum, and palladium (50% + 50% × 29.0%). That means a stream-free ounce from Crean Hill or R2 is worth 1.55x as much to Magna as one from McCreedy West (100% / 64.5%). The stream also covers two areas of Levack, including the Morrison deposit. Franco-Nevada’s filings set the price for its Sudbury streams at $400/oz, rising 1% a year from July 2011, and its later increases applied to McCreedy West only, so Levack’s price would’ve reached ~$460/oz by Q4 2025 ($400 × 1.01^14). So at Q4 2025’s gold price, Magna would keep 55.6% of the value of Levack’s streamed ounces (50% + 50% × ($460 / $4,141.90)), less than the 64.5% it kept at McCreedy West.
@convequity 3 clicks

Convequity on X

$NET's AI hosting is limited by the physics of its edge footprint and by the role of its products. Carrier-hotel sites built for fast network traffic are a different proposition from purpose-built AI datacenters. Cloudflare’s own engineering disclosures acknowledge power constraints, and management distinguishes its strategy from the hyperscalers’ GPU spending race. Our assessment starts with a two-H100-per-server edge configuration. H100 PCIe and NVL-class cards operate around 350–400 watts per GPU, versus up to 700 watts for H100 SXM. Hyperscalers commonly pack eight SXM GPUs into a machine and connect those machines into dense clusters. The distinction is power, cooling, memory bandwidth and GPU interconnects — not simply whether a model can run. The existing edge footprint cannot acquire hyperscaler-class cooling and power density through server upgrades. Dense, liquid-cooled Blackwell and Vera Rubin systems require a different infrastructure commitment. We see little chance that NET’s existing edge footprint becomes a competitive home for those dense clusters. For sustained, large-model inference, our assessment is that sparse edge GPUs cost more per token than dense AI clusters. Geographic proximity saves network transit time, but it does not guarantee a faster first token or a cheaper response. Greater compute capacity, faster interconnects and efficient batching can outweigh the edge’s network advantage. We expect heavy training, frontier-scale hosting and the most demanding next-generation inference to remain concentrated with hyperscalers. Workers AI’s progress with larger open models is real. Hosting those models does not establish an economic advantage over dense clusters. AI Gateway also lets customers route requests to external providers, matching the model and infrastructure to the task. Cloudflare can still win substantially. More agents mean more requests to secure, more Workers executions, more orchestration and more demand for fine-grained access controls. Those workloads build on infrastructure and capabilities Cloudflare already has. On the SaaS Survival Scorecard, we score $NET at 2.5 for Model-Layer Leverage. That reflects its substantial work building an agent harness across inference, orchestration and execution, while stopping short of model post-training and reinforcement learning.

Investor Letters

All letters

TSOH Weekly Roundup (10/09/2026)

Paramount Skydance’s debt-heavy Warner Bros. Discovery acquisition, streaming bundle pricing and alcohol-industry diversification anchor…

WBD $78B
TSOH Investment Research
BNTC · long

Benitec Biopharma INC

Benitec Biopharma's BB-301 gene therapy is in a Phase 1b/2a trial for an ultra-rare muscular dystrophy.

BNTC $314M
Value Investors Club

Boyz in the Hood

Robinhood’s AI agents, 24/7 equity trading and 10:1 crypto perpetual futures encourage higher retail trading velocity, leverage and…

HOOD $98B
WHAT ARE WE DOING? “Contrarians At The Gate”

Publications we follow

Every publication on this site, hand-picked.

109 publications

Newsletters 92

YouTube 4

Blogs 12

Seeking Alpha 46 of 218

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

74 accounts
Nominate an account

Investor letters we follow

189 firms

Sectors

Ordered by how often our editor reaches for them.