Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

Latest News

Market capOver $10B
All Over $50MOver $300MOver $2BOver $10B
All Under $1BUnder $10BUnder $100BUnder $500B
SectorsAll
Filter by sector
Thursday, October 15, 2026· 5091 stock ideas· 217 writers·
15,653 views· 7,394 clicks

[scuttleslops] MRVL, APP, AMZN, Chips

Marvell is positioning networking, optical connectivity and server-management chips as AI data centers link GPUs, custom accelerators and racks, lifting five-year targets.

MRVL $241B
scuttleblurb 5 clicks

From X

@StableBread 1 click

Fajasy on X

I tried to work out who’s behind the AI data labeling companies that approached Nubeva Technologies $NBVAF (TSXV: $NBVA). CEO Randy Chou didn’t name the data labeling companies (firms that turn experts’ knowledge into training data for AI labs) or the customers behind them. “If a data labeling company’s coming to us, somebody behind them is asking,” Chou said, adding that the customer typically isn’t limited to cybersecurity companies. OpenAI is one potential candidate, since my previous write-up concluded that the frontier model Nubeva got access to in September was most likely OpenAI’s, through its Daybreak program for vetted defenders. OpenAI launched GPT-6 Astra on September 3, 2026, the day Nubeva announced its access. It’s the first model OpenAI rated Critical for cybersecurity capability. But access to OpenAI’s model wouldn’t prove OpenAI is also behind the data labeling companies, since any lab building cyber models could want labeled malware. Among data labeling companies, three of the largest are publicly recruiting or selling binary analysis expertise, the skill of figuring out what malware programs do: → Scale AI: Its product manager posting calls the cybersecurity portfolio “a product line that does not exist yet” and says the hire will recruit “malware analysts.” → Turing: Sells CyberBench, security tasks for training and testing frontier models, with malware analysis and reverse engineering among its ten categories. TechCrunch called Turing “a key coding provider for OpenAI” in March 2025. → Mercor: Recruits security experts, preferring reverse engineering or binary analysis backgrounds, and counts OpenAI, Anthropic, and Meta $META as customers. OpenAI cut ties with Scale in June 2025, after Meta bought 49% of Scale and hired its founder, Alexandr Wang. So if OpenAI is behind the approach, Turing and Mercor are the likeliest data labeling companies, since both sell to OpenAI and already sell or buy binary analysis expertise. Either way, these data labeling companies are worth $2.2-29B, so even Turing, the smallest, could pay the top of my $12-19M sale range for Nubeva for less than 1% of its last valuation.
@convequity 1 click

Convequity on X

AI swarms could make network security more important. Anthropic's October 12 research ran an AI model across 15 cheap Android TV boxes at 0.39 tokens per second. The small models still couldn't reliably complete the full replication cycle. I think we're still early. Inference on this hardware is painfully slow. But the idea is interesting because human-level intelligence isn't necessary for something to reproduce and spread. Think of insects: limited individual intelligence, yet populations can spread widely. An AI worm could carry its model with it, adapt to unfamiliar devices and use compromised hardware to run the next copy. A model too large for one device could be split across several. Better processors and more memory in future cameras, TV boxes and other connected devices would make more local inference possible, wherever attackers can gain access. That creates a security problem beyond installing software on individual devices. You're not going to put an endpoint security agent on every camera or TV box. You need to identify those devices through the network, restrict what they can communicate with and isolate them when compromised. A perimeter firewall alone cannot inspect traffic that never passes through it. I think $FTNT could benefit here. Its agentless device discovery and network segmentation fit this problem, especially across distributed sites. More capable AI on unmanaged devices could increase demand for those controls. The opportunity also extends to networking vendors that integrate security into switches and wireless access points, where they can control how devices connect and communicate.
@stockgutter

SG on X

I am a skeptic of the current quantum companies but cannot ignore that $IONQ just doubled its cash hoard from $2B to $4B, selling shares/prefunded warrants at $40 (above market) covered 2:1 by warrants with a strike at $69.46. Not a bad move imo. While I agree with critics that the warrants are a bit oppressive, shitcos will have to pay up one way or another to raise $$ in size. There are worse structures. No position in IonQ rn, but considering selling long-dated calls in the 2x inverse ETF $IONZ - I think the underlying will participate strongly in any junk bounce towards year end.
@convequity

Convequity on X

$NVDA's falling unit share in China's data-centre AI chip market overstates how much business Nvidia has lost to Huawei. Three things explain why. 1. IDC counts units, and did not say whether that count includes the Blackwell parts barred from China and reportedly smuggled in. Chinese demand also reaches Nvidia outside China. Xiaomi or ByteDance rents a Blackwell cluster in Malaysia from a local cloud provider. That is Nvidia revenue from a Chinese buyer that never appears as a Chinese unit. So Nvidia's share of what Chinese buyers spend on AI compute is higher than its share of counted units. 2. Huawei's volume growth is real, and memory explains how. Huawei cannot get enough stacked high-bandwidth memory, the HBM that top-end AI chips need. The 910C ran on an older grade, HBM2E, that Huawei had been able to mass procure. The new 950PR uses lower-bandwidth single stack memory instead of stacked HBM, so the 950PR does not depend on HBM supply. Each chip moves data more slowly, and Huawei can make many more of them. Also, we expect 950DT to ramp up quite quickly alongside 960DT coming earlier in 1Q27 instead of original 4Q27 timeline due to Huawei's own HBM variant that is able to ramp up production quicker. 3. So Nvidia's drop is not Nvidia failing to supply China. In our view, it has three causes: US export rules that bar its best parts, Beijing promoting domestic chips and requiring approval to import Nvidia's compliant ones, and a 950 series that beats the 910 series it replaces and is production-wise attractive for model serving. For Nvidia, where Chinese compute spend goes matters more than the China unit count, and much of it still goes to Nvidia hardware, in China or in Malaysia.

Investor Letters

All letters

Free Recaps of the best new stock ideas, delivered every morning.

Read past recaps

My 2028 numbers

ASML’s €11.7 billion quarter benefited from €3.3 billion in installed-base sales and operating leverage, while the MATCH Act threatens…

ASML $660B
Best Anchor Stocks 22 clicks

Have a source we’re missing? Send your pick

Recent recaps

Latest recap

Wednesday, October 14 70 stock ideas
  • October 2026 Random Ramblings Yet Another Value Podcast
  • Visa: Unpacking The 7 Powers And Forward Valuation Ahead Of Q4 Earnings Seeking Alpha
Read the recap

All recaps

Publications we follow

Every publication on this site, hand-picked.

129 publications

Newsletters 107

YouTube 4

Blogs 17

Seeking Alpha 49 of 226

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

88 accounts
Nominate an account

Investor letters we follow

189 firms

Sectors

Ordered by how often our editor reaches for them.

Anything we’re missing?

Send us a pick