Sponge

In search of asymmetric bets

[Sponge @Sol_Sponge]

Here’s why the $KOPN base case is $13.50 and bull case $42.

AI’s bottleneck isn’t compute — it’s the wires between GPUs (one of many bottlenecks).

Copper hit a wall. Silicon photonics is expensive. MicroLED is the dark horse: sub-pJ/bit, no lasers, CMOS-compatible.

$90B AI optics TAM by 2030.

[Sponge @Sol_Sponge]

Turn the printer back on

[The Kobeissi Letter @KobeissiLetter]

BREAKING: President Trump says the Trump Administration might buy equity stakes in US AI companies and that he will host a meeting with AI executives as soon as next week, per Reuters.

[Sponge @Sol_Sponge]

@Nerd_0ne I love the conviction

[Sponge @Sol_Sponge]

@Eric07644638658 I like it

[Sponge @Sol_Sponge]

My post last month on $KOPN

[Sponge @Sol_Sponge]

Here’s why the $KOPN base case is $13.50 and bull case $42.

AI’s bottleneck isn’t compute — it’s the wires between GPUs (one of many bottlenecks).

Copper hit a wall. Silicon photonics is expensive. MicroLED is the dark horse: sub-pJ/bit, no lasers, CMOS-compatible.

$90B AI optics TAM by 2030.

[Sponge @Sol_Sponge]

May has been the most insane month of trading. Taking foot off the gas here. What are we buying next?

Still holding:
$NOW
$PENG
$KOPN

[Sponge @Sol_Sponge]

$NOW

[Sponge @Sol_Sponge]

$NOW

[Sponge @Sol_Sponge]

$PENG just doubled its FY26 growth guide mid-year.

AI now 60% of revenue. Memory segment +53% YoY. First production CXL KV cache server shipping today.

They’re building the missing tier in the AI memory hierarchy.

$53 today. Base $95. Bull $185.

[Sponge @Sol_Sponge]

AI has a memory problem. $PENG has the solution. And it’s only a matter of time before the price reflects that.

Every long AI conversation builds a “KV cache” inside the GPU. Long docs, agent tasks, big context windows — the cache overflows fast.

When it overflows, GPUs recompute, stall, or refuse the request. That’s why inference is so expensive.🧵

[Sponge @Sol_Sponge]

$NOW

[Qualtrim @qualtrim]

ServiceNow just raised AI guidance.

AI/Now Assist grew by +130% YoY last quarter, with deals including 3+ Now products increasing by 70%.

New Guidance: $1.5B (raised +50%)

$NOW

[Sponge @Sol_Sponge]

@kozumplikivan Hey. Tried it out getting an API error

[Sponge @Sol_Sponge]

The narrative around software will change. They have the infrastructure and sticky model to introduce, manage and scale agents at an enterprise level. Majority of companies won’t be able to implement these solutions on their own and will continue to rely on SaaS.

The good companies will identify this opportunity and I think $NOW has the right leadership to pivot effectively 🍻

[Sponge @Sol_Sponge]

It’s been a very very good week. Need to keep our head on straight here. Keeping it simple and sticking to the two plays I feel most confident in -

$PENG
$NOW

Happy Memorial Day

[Sponge @Sol_Sponge]

What do we think will happen with $NVDA earnings?

[Sponge @Sol_Sponge]

Interesting $NOW

[PeloSwing🎯👸🏻 @PeloSwing]

Wonder Where $NOW is Headed

@BullflowIO

[Sponge @Sol_Sponge]

@Traderibo123 @Polymarket We can only wish brother

[Sponge @Sol_Sponge]

All roads lead to $PENG

[🏀📡 @pennycheck]

$PENG just announced the current president of network infrastructure at $NOK and former CEO of $INFN (before NOK bought them) has joined the board.

[Sponge @Sol_Sponge]

Mood today $NOW

[Sponge @Sol_Sponge]

Nice write up on $PENG

[John Galt @AtlasShrug1]

$PENG has a fairly convincing path to get to $2.5B in sales in 2028…at a reasonable 5x multiple that would be a $12.5B market cap vs. $2B today. The mistake is not anyone counting on “photonic memory”, the mistake is pricing $PENG as a systems integrator when it is really an “anti-hyperscaler” as Manuel Walz (Garp Investing substack) so aptly put it in what I believe is one of the best pieces on $PENG to be published. As he so eloquently explains, the company “builds the sovereign, on-prem AI factory that enterprises, governments, and telco operators can’t obtain from AWS, Azure, or GCP”. He goes on to superbly address their software moat (ICE ClusterWare abs MemoryAI) already being deployed at production scale in Korea w/ $SKM who is an investor and sits on the BOD along w/ the Chairwoman of $LITE and CTO of $AMD. He then rightly talks about how the $NBIS acquisition of Eigen AI validates the inference optimization layer at the “highest value layer of the AI stack”. And then how the $NVDA/$NOK $1B partnership creates a “distributed edge inferencing market worth > $200B by 2030 rhat $PENG is UNIQUELY positioned to serve.”

$SKM is embedding $PENG in their rollout of its national AI -native strategy and has preferred convertible into shares at $33 purchased 2 years ago way before the big memory bottleneck was appreciated and $PENG was not one of or perhaps the only company with a production ready CXL-based KV Cache server to unlock the memory wall. This is critical with inference workloads being bound predominantly by memory (70%) and only 30% by compute. This technology didnt appear out of nowhere, it builds on $PENG’s 30-year Smart Modular Technology heritage and…the strategic deal they inked with SK Hynix in Jan of 2025 giving $PENG access to cutting edge DRAM packaging the the competition cannot replicate.

There is a lot more in Manuel’s piece on Substack, and I think his stuff is highly underrated as he is an extremely insightful writer.

I also want to acknowledge @Temple_Eight for being early to write up this name and put it in front of me as an idea requiring serious consideration as a core portfolio position.

I will be very surprised if within 12-18 months this isnt a 10-figure marker cap, which implies a 5-6x from Friday’s close.

@ThematicTrader or @BryzonX feel free to chime in here if you have anything to add!

NFA. Good luck and happy trading.

[Sponge @Sol_Sponge]

I love @Polymarket

[Sponge @Sol_Sponge]

Internalize this post

[QC Capital @QC_Capitals]

Us in 2030 because we bought $NOW at $90