Beth Kindig

SK Hynix on HBF: “HBF is typically half the bandwidth of HBM, but more like 14 times higher capacity than HBM.” $MU $SKHY $NVDA

AMD $AMD says the first Helion racks are under development and are planned to be available in November. $NVDA $INTC $ARM

ASML: “We are increasingly receiving orders for 2028 and customers are also expressing interest in placing orders for 2029 and even 2030.” $ASML $TSM $INTC $NVDA

TSMC: “We have no intention to leave any food on the table for others.” $TSM $NVDA $INTC

Global smartphone shipments declined (22%) YoY in Q3 to 255.3 million, the worst decline ever due to rising memory prices, with IDC warning that the “worst is yet to come.” $AAPL $GOOG $MU

Anthropic’s Q3 revenue reportedly hit $16.8 billion, up 46% QoQ from $11.6 billion in Q2, with YTD revenue around $33.2 billion. $NVDA $AMD $AVGO $AMZN $GOOG

Broadcom $AVGO on its financing endeavors for Anthropic and OpenAI: “It's a great investment for us when you think about it, that every gigawatt of compute they deploy, they could achieve $30 billion of ARR, annual revenue per gigawatt." $NVDA $AMD

The deteriorating macro backdrop, from bonds to inflation, could create a meaningful headwind for stocks, but that doesn’t necessarily mean the secular bull market is over. Below, we map out the headwinds at play, and what key factor could keep the bull market on track after a correction. https://t.co/x9DxAvVrFs $SPY $QQQ

Last week, we laid out the case for a continuation of this secular bull market, but with a correction coming first. In this week’s analysis, we examine the deteriorating macro backdrop that could pressure the broad market – and one key factor that could limit the severity of a correction. $SPY $QQQ

Nvidia’s $NVDA developer base has reached 11 million, more than tripling from 3 million in 2021. $AMD https://t.co/3LCIwpR3TI

We believe this bull market has considerably more room to run. However, before the next leg higher begins, we expect a correction that will shake many investors out of the trade – and the main AI proxy, $SMH, appears to be confirming this view. https://t.co/OWQT7pcPo6 $SPY $QQQ

TSMC $TSM reported Q3 revenue of $46.72 billion, up 41.1% YoY and beating estimates for $45.7 billon. For Q4, TSMC guided for revenue to be $52.3-$54.3 billion, up 58% YoY and above estimates for $50.5 billion. $NVDA $AMD $AVGO $AAPL

Since 1880, the average secular bull market lasts 11.2 years, with an average gain of 675%. Compare this to the current bull market, going on 17.4 years old and up 1,073%, as of September 2026. https://t.co/cQKdnGPgQp $SPY $QQQ https://t.co/LCrBvdjocj

Over the past three years, Micron $MU is outperforming Nvidia $NVDA, AMD $AMD and Broadcom $AVGO by ~3-4X, with a return of 1,230%. https://t.co/CBaDe7ybWb

Nvidia’s CUDA downloads have reached 1.2 billion as of Q2, up more than 37X from 32 million in 2021. $NVDA https://t.co/Qia7yRj7OY

Citi projects the agentic CPU TAM to rise 14X over the next four years, scaling from $11.6 billion in 2026 to $157.4 billion by 2030, accounting for more than half of the total CPU TAM. $NVDA $AMD $ARM $INTC https://t.co/GjKrCMvKOL

BofA raised its global memory TAM forecast to $2 trillion by 2030, up from $1.8 trillion, with annual run rate already exceeding $1 trillion as of Q3. $MU $SKHY $NVDA $AMD

GlobalFoundries $GFS has partnered with TSMC $TSM to expand fab capacity in New York for production of silicon interposers for TSMC’s CoWoS packaging under a $2 billion, five-year deal. Production is expected to start ramping in early 2028. $NVDA $AMD $AVGO

Nvidia-backed cloud startup Lambda is aiming to raise $4 billion in a final funding round before an IPO in 2027, with the new round said to value the firm at $14.5 billion. $NVDA $AMD

Broadcom $AVGO is in early talks to arrange more than $50 billion in financing to help OpenAI fund custom silicon purchases, following reports of the up to $60 billion package for Anthropic from earlier this month. $GOOG $MSFT $NVDA

AMD’s $AMD Helios racks are estimated to cost $7.33 million, per Bernstein, closely matching Nvidia’s $NVDA Vera Rubin at an estimated $7.52 million while far ahead of Google’s $GOOG TPU v7 Ironwood at $2.69 million.

The Next Big Buying Opportunity May Be Closer Than You Think

[music] >> Okay, everyone. We'll go ahead and get started here. Hope you have a good evening. Good afternoon, wherever you are. >> [snorts] >> And um >> [clears throat] >> we'll use the same format, just kind of go through the broad market. The risk we see, the long-term potential that we see, uh justifications for and I'm going to kind of get into a bit of a thematic issue within the macro. Just to kind of give you some stuff to think about. And then we'll get into some stock charts. So, starting off with the trend. This is the S&P 500 on a very large scale. It's a monthly scale. That's the 2000 top, 2009 low. So, the large decade-long secul…

In a full 96-accelerator rack system, the Maia 300 will deliver up to 7X the throughput of the Maia 200, while consuming about 167kW of power. $MSFT $AMZN $GOOG $NVDA

Although the current secular bull market is an outlier in both duration and returns, we believe some of the strongest gains may still be ahead – though not without volatility. Below, we map out why the best in AI is yet to come and the one key level the market must hold for our strategy to remain unchanged. $QQQ $SPY https://t.co/OWQT7pdndE

TSMC $TSM reported September revenue of NT$511.9 billion, up 54.6% YoY, bringing Q3 revenue to NT$1.49 trillion ($46.7 billion), up 51% YoY. $NVDA $AMD $AVGO

Nvidia $NVDA says that its share of IT capex across the top 5 hyperscalers rose from 11% at the start of 2023 to 38% in Q2 2026, expecting this to rise to 44% by the end of 2027. $AVGO $AMD https://t.co/D1qmX4LCBu

OpenRouter’s monthly token traffic has risen from 2 trillion in January 2025 to 379 trillion as of July 2026, with open-weight models accounting for 75% of token generation. $NVDA $MSFT $META $GOOG $AMZN https://t.co/edzFs1TBaB

The data center bottleneck may no longer be demand, but permission. While planned MWs represent data center capacity in the queue, permissioned MWs represent capacity that could realistically be built. Here’s why that distinction matters for investors. ⬇️ https://t.co/kRk2hIMBgI $ORCL $APLD $BE

In the race to ramp up, Nebius $NBIS and CoreWeave $CRWV have similar gross debt to annualized Q2 revenue, at 3.7x and 3.4x, respectively. The difference lies in how much liquidity each has to offset debt – Nebius has net debt of 0.2x annualized revenue, compared to CoreWeave’s 2.9x, giving it higher flexibility to build capacity. https://t.co/PzeUUk4VxG

Nvidia $NVDA says that for a 100MW data center, its Vera CPU offers more than 2X higher throughput and memory bandwidth versus x86 CPU competitors. $INTC $AMD https://t.co/qqm4eQNXe6