Paradi Lab

Ok, this is big for $Penguin Solutions(PENG.US) & High Performance Compute:

NextSilicon just announced that the Spectra supercomputer (the system $Penguin Solutions(PENG.US) built + integrated) has achieved full acceptance under Sandia National Labs’ Vanguard programme.

$Penguin Solutions(PENG.US) worked w/ Sandia + NextSilicon to deploy & support the platform so it could run real national security workloads.

E.g. HPCG, molecular dynamics + MonteCarlo sims.

And since the Vanguard programme is infamously selective, it basically validates $Penguin Solutions(PENG.US)'s integration capabilities in HPC environments.

Which should help with:

1. follow-on govt work

2. commercial HPC/AI customers looking beyond vanilla GPU clusters

So, $Penguin Solutions(PENG.US)'s strength in high-performance memory + systems integration (incl. CXL) is probably an advantage here when balancing new accelerator designs?

Disclosure: I don't have a $Penguin Solutions(PENG.US) position.

The fact that we've got a $Cerebras(CBRS.US) 2x leveraged ETF...

Is exactly why we can't have nice things.

GS data shows that Retail are using more and more leverage.

Which is why I think immediate releases of leveraged ETPs by firms like Leverage Shares is extremely predatory.

Profiting off of Retail mania hours post Cerebras IPO.

From my own personal surveying over the years, I'm confident in saying that 99% of leveraged ETP buyers hold for days/weeks at a time.

And not day trading those products, which is what they're designed for.

Tbh, for Retail, I don't advocate for leverage full stop.

SpaceX are going to raise up to $75B on IPO next month.

Which is 3.5x bigger than $Alibaba(BABA.US), the current biggest U.S. IPO.

Very grateful to have crossed 40K followers this week. Truly blown away by the support.

As you guys know, everything I make from X will be donated to charity.

-> GOSH are a London hospital that cares for kids with rare conditions/diseases. Very important work done by selfless individuals.

Did £4K to kinda match the 40K followers lol. Unfortunately I'm not rich enough yet to donate £40K in one go.

Regardless, a long-term goal of mine is to do monthly 5-figure donations to my chosen charities.

(I didn't make £4K from the rev share programme btw. Just less than £500)

Passing up on $Cerebras(CBRS.US) at ~$100B MC.

The bull case is compelling since they're a genuine architectural-paradigm story in AI compute.

Highlighted by a stellar list of customers using their chips:

- OpenAI: $20B+ contract

- $Amazon(AMZN.US): AWS data centers

- $Meta Platforms(META.US): Llama API for devs

- Perplexity: sonar model

- $GSK(GSK.US): biomed models

- Mistral AI: Le Chat

- Notion: enterprise search

- AlphaSense: enterprise search

All of that diversifies them away from G42 in the UAE which have been the historic rev driver.

Regardless, not too sure when I'll look to initiate a position, maybe like the $200 mark?

Kinda feel there's some downside risk rn when you layer on top:

- $NVIDIA(NVDA.US) Vera Rubin volume ramp (H2 2026)

- potential Nvidia absorption of Groq

- CUDA moat

So current valuation looks like all the execution upside is being aggressively priced-in.

Kioxia (285A) just put out sensational earnings & guidance.

Very bullish validation of the NAND supercycle.

I'm viewing this dip as a perfect buying opportunity to cost-avg up.

Q4 FY25 (Jan-Mar 2026):

- Revenue: ¥1,002.9B (QoQ +85%)

- OP: ¥599.1B (margin 60%; QoQ +314%)

- ASPs >2x'd QoQ; bit shipments down ~10% QoQ as planned for inventory control.

- 8th-gen BiCS production now exceeds 5th-gen levels.

Cash flow and balance sheet are pristine post-refinancing.

Q1 Guidance (Apr-Jun 2026):

- Revenue: +75% QoQ

- OP: +117% QoQ; margin 74%

- NP: +112% QoQ

"AI inference servers driving overall NAND demand growth… interest is intensifying."

"CY27: Demand anticipated to exceed supply."

Memory keeps going up and up with every quarter...

Pls DYOR.