🚨Ethereum on exchanges just hit a record low—only 14.5m ETH left out there. Since late 2023, over 6 million ETH has been pulled off exchanges, thanks to ETFs and companies stacking up.
Themes Lab — 6/10 🧪
Capital's moving fast across 180+ themes. Semis are on fire — memory and foundry leading the charge, up 69% in 3 months. Market's feeling RISK-ON.
TOP MOVERS (RS)
• Memory — RS 92
• Semi Equipment — RS 75
• Analog Semis — RS 78
• AI Server
@TheBigBerbowski Saw that too. It’s basically saying the old network setups can’t handle the data explosion anymore, so they’re selling a faster, more scalable solution for data centers.
What Problem is $CRDO solving?
FY2025 10-K, Item 1, Industry Overview:
“Data generation has increased dramatically over the past ten years, creating new and complicated challenges in both circuit and system design.”“Back end scale up networks have grown from an appliance scale of 8 GPUs to rack scale of 72 GPUs in calendar 2024 and there is strong interest in growing this to scale in the coming years. Scale up networks are already 10x denser than scale up networks, a full 100x denser than front end networks.”13
CEO Bill Brennan, Q4 FY2026 Earnings Call, June 1, 2026:
“As AI clusters scale from tens of thousands to hundreds of thousands of GPUs, connectivity is no longer just about bandwidth. Reliability, power efficiency, signal integrity, and telemetry have become critical architectural requirements.”14
The problem Credo is solving is both critical and growing.
- AI cluster density has grown 9x (from 8 GPUs to 72 GPUs per rack), and
- Scale-up networks are 100x denser than front-end networks
As network density is increasing, reliability, power efficiency, and signal integrity aren’t just nice to haves, they are critical requirements.
Credo’s entire product portfolio is a direct solution to this problem.
FAANG is dead... Welcome to MANGO.
- Meta
- Anthropic
- Nvidia
- Google
- OpenAI
Think about what that means…
The companies that defined the last 15 years of tech just got replaced by the AI boom.
FAANG built the apps. MANGO builds the intelligence behind them.
$NVDA makes the chips. $GOOGL, OpenAI, Anthropic, and $META build the models.
Every major AI project depends on at least one of them.
And here's the interesting part:
The two hottest names in MANGO, OpenAI and Anthropic, aren't even public.
The LIT Sunday News is out now!
After Friday's rough sell-off, get ready for the week ahead and be prepared.
Free and paid members can check it out here: https://t.co/KOMzqBKCQL
Hold off on buying the dip for $DRAM or $SOXL, even though they dropped hard Friday. 📉
$DRAM still has a gap to fill down to $53.67 and might test a trendline around $51 first before bouncing.
$SOXL could slide to a support trendline near $170. If that breaks, next stop i
@Mr_Derivatives I get the point, but institutions also brought liquidity and stability. The real question is whether that trade-off was worth losing the wild upside.
$BTC will need to more than double here to get back to ATH’s.
$ETH will need to more than triple here to get back to ATH’s.
Did the institutionalization of crypto spell the actual doom of it?
Silver hit its 7-month uptrend and the 200-day moving average yesterday. Yeah, it's a rough sell-off, but prices were even lower back in February and March. Reckon we'll get through this one too...