orhan maçin

kripto ve finans dünyası ile ilgili kaliteli paylaşımları bu sayfada toplamak istedim
geleceğini sen inşa et
bilgi hazinedir

[orhan maçin @MacinOrhan]

Çoğunuz kriptoda ben nerede yanlış yaptım diyor değil mi?
Şişirilene değil, gerçek gelir/ piyasa değeri oranına bakın.
https://t.co/t3HW0Ctg8y

[Chain INK @0xchainink]

Most of DePIN is priced like it earns.

Only the cheap ones actually do.

Earning, priced cheap:

→ $IO 5.1x
→ $GEOD 12.6x
→ Chutes (SN64) 18.3x
→ Glow 12.9x
→ $ROVR 0.6x

Priced on story:

→ Render : 684x
→ Filecoin : 717x

The ones getting used trade cheapest.

[Chain INK @0xchainink]

Most of DePIN is priced like it earns.

Only the cheap ones actually do.

Earning, priced cheap:

→ $IO 5.1x
→ $GEOD 12.6x
→ Chutes (SN64) 18.3x
→ Glow 12.9x
→ $ROVR 0.6x

Priced on story:

→ Render : 684x
→ Filecoin : 717x

The ones getting used trade cheapest.

[TomATR @ATR_Tom]

Today is $GEOD halving !

What it means ?

- Miner emissions cut in half :

Before : ~1,6M token send weekly

After today : ~800k token send weekly

- Burns unchanged (80% of revenue)

~900K token burn weekly

->The network now burns more than it mints
-> Incentives cut in half while revenue keeps climbing

[GEODNET @GEODNET]

The GEODNET halving is here.

From July 1, the max base mining reward drops from 12 to 6 GEOD per day. It happens every June 30 on a fixed schedule, set in our tokenomics from day one.

Less new GEOD entering circulation. A network still growing across 150+ countries.
This is how we build for the long term.

[orhan maçin @MacinOrhan]

Sizi @GEODNET
$GEOD
Konusunda aydınlattım
Farkındalık yarattım
Piyasa çakılırken
Geod güç gösteriyor
Sağlam altyapı, artan gelir
Coinbase listelemesi..
Ah dostum
Dikkate aldın mı

[orhan maçin @MacinOrhan]

@ionet and
@GEODNET
$GEOD
https://t.co/z3NH1e7XAz

[Who Loves Burrito? @wholovesburrito]

$GEOD started trading on Coinbase this week.

22,000+ stations across 160 countries. $7.8M in annualized revenue from enterprise GPS subscriptions. 80% of that goes to token buybacks and burns.

One of the cleaner tokenomics loops in DePIN — revenue directly tied to real demand.

[io.net @ionet]

Most DePIN network token rewards outpace real revenue by 20-40x.

Not @ionet.

Research from Cambridge University Business School shows https://t.co/ZuybGWvjv9 leading on real revenue, even before the IDE.

The IDE makes that sustainable. Burns tied to real usage. Incentives that tighten as the network matures. Not inflation dressed up as growth.

Most DePIN compute projects are subsidising demand that doesn't exist yet. https://t.co/ZuybGWvjv9 is building revenue that does.

The difference matters. A lot.

[Oleg Dimov @OlegDimov]

🚀 GEODNET ($GEOD) is now live on Coinbase!
Coinbase customers can now buy, sell, convert, send, receive, or store $GEOD directly on https://t.co/8uYhZaIqrf and the app.
GEODNET is a DePIN project building a decentralized network of satellite reference stations for centimeter-level GNSS accuracy — powering robots, drones, autonomous vehicles & AI 📡🤖 🚙
Earn $GEOD by running stations + 80% of network data revenue goes to token buybacks & burns @GEODNET
🔗 https://t.co/Jk8VwahCjo
#GEOD #DePIN #Coinbase

[orhan maçin @MacinOrhan]

@rockerfellerxau @GEODNET $geod hakkında bilgilenmek için bakabilirsin
https://t.co/LDOauDnOat

[orhan maçin @MacinOrhan]

$GEOD
@GEODNET
Hakkında bir yazı
https://t.co/CUjhZeBEKF

[orhan maçin @MacinOrhan]

@rockerfellerxau Umut kalmadı diyip insanları kriptodan soğutmayın
@GEODNET
$geod
Paylaşımlarıma bak
Değişime ayak uydurmaya köhnemiş projelerde takılı kalmayın
Robotik sektörü kaliteli projelere odaklanın
Hexagon, trimble 1000 istasyonla 20 milyar$, $geod 21000 istasyon 90 bin$ piyasa değeri

[BUNT @BUNT10]

Next Big Bets: Water & Sun

The AI race has a physical constraint that no amount of compute can abstract away.

Power. Every GPU cluster, every inference call, every training run needs electrons and the grid isn't keeping up.

As per @Gartner_inc ,Global data center electricity consumption is projected to hit 565 TWh in 2026, up 26% from 447 TWh in 2025.

By 2030, that number will cross 1,200 TWh. To put that in context, the entire country of France consumes roughly 450 TWh per year.

AI infrastructure alone will consume nearly three Frances' worth of electricity within four years.

AI-optimized servers are the primary driver. In 2025, they consumed 95 TWh, already running at 83.6% year-over-year growth. By 2026, that figure reaches 175 TWh, and by 2027, AI-optimized server power consumption will surpass conventional servers entirely.

The composition of data center load is being rewritten in real time.

The Resource Crunch

AI capacity is already being limited by power availability, not by chip supply or software maturity. Hyperscalers are spending billions on generation capacity, signing long-term PPAs with nuclear operators, and racing to secure grid interconnection queues that take years to clear.

This has a clear investment implication: whoever controls reliable, cheap, scalable power controls the pace of AI development. The energy stack is no longer infrastructure. It's a strategic moat.

Where Renewables Enter the Frame

The push toward solar and wind isn't ideological. It's math. Utility-scale solar is now the cheapest form of new electricity generation in most of the world. The speed at which solar capacity can be deployed relative to permitting and building new gas peakers or nuclear plants makes it the default answer to time-constrained demand.

The AI energy crunch is the most powerful demand signal the renewable energy industry has ever received. Not carbon credits, not policy mandates. Hard commercial necessity from the world's most capitalized sector.

The flywheel writes itself:

AI demand → power scarcity → capital into renewables → more competitive renewables → more AI infrastructure deployed.

The narrative and the economics are aligned for the first time since modern solar manufacturing scaled.

Water is the other side of this equation and it's a less comfortable story.

The Hidden Resource War

Every GPU that runs hot needs cooling. At scale, the numbers get uncomfortable fast. A single high-density GPU chip can draw up to 700 watts.

Three chips equal roughly the same energy draw as a home electric oven. Now multiply that across tens of thousands of chips per data center.

The cooling answer has largely been water. fresh water pulled from local supplies to manage thermals at industrial scale.

A report from the UK Government Digital Sustainability Alliance projects AI will push global water usage from 1.1 billion cubic metres to 6.6 billion cubic metres by 2027. That's a 6x increase in six years.

Microsoft disclosed that 41% of its water withdrawals came from water-stressed areas. Google acknowledged 15% of its water consumption occurred in regions with high water scarcity.

Only 0.5% of the planet's water is fresh water. And it's not just data centers competing for it, power plants need water to generate the electricity that runs data centers in the first place. Chip fabrication requires water.

The entire AI supply chain is water-intensive, not just the inference layer.

Communities near data centers are already absorbing the costs. Residents reported discolored, sediment-filled tap water coinciding with Meta's data center construction nearby.

They're early signals of a structural conflict between AI infrastructure expansion and local water security.

The Innovation Response: Get Off the Grid

@_panthalassa is the clearest example of where this is heading. The Portland-based startup is building wave-powered, self-cooling AI data centers designed to float in the open ocean.

Peter Thiel led the company's $140 million Series B, pushing its valuation toward $1 billion.

The model solves the two constraints simultaneously. Power comes from wave energy 85-meter steel nodes that bob in the ocean, using the motion of water to oscillate internal tubes that drive turbines.

Cooling uses surrounding cold seawater, eliminating the freshwater draw entirely. Connectivity runs through Starlink. The whole unit is off-grid, off-shore, and disconnected from terrestrial permitting queues.

Panthalassa is targeting commercial deployment by 2027. Whether or not this specific configuration scales, the thesis it represents is directionally correct:

the constraints on land-based AI infrastructure are becoming structural, and the solutions will require physical innovation, not just software optimization.

Bunt's POV

The energy and water angle on AI infrastructure is underpriced as a narrative and I think that changes in the next 12 to 18 months.

The renewable energy trade is obvious in retrospect. When the world's most capitalized industry hits a hard physical ceiling, and the fastest path to expanding that ceiling runs through solar, wind, and nuclear PPAs, capital flows accordingly.

The AI energy crunch is a demand catalyst that the clean energy sector hasn't seen before. That alone justifies a closer look at where the infrastructure build-out is going.

Water is the more complex call. The projections around data center water consumption are significant enough to warrant attention from an infrastructure and policy standpoint but the investment angle is less direct. What I do think is that water scarcity is going to become a permitting and regulatory variable for data center siting decisions.

Locations with abundant water and cheap renewable power will command premiums. That affects real estate, regional utility operators, and any player in the physical AI supply chain.

The Panthalassa bet is interesting precisely because it routes around all of this. No permitting. No freshwater draw. No grid dependency. If the unit economics work, it's not just a data center it's a land-free, water-neutral compute platform. That's a different kind of asset class entirely.

The energy narrative is early but tangible. The water narrative is even earlier. Both are worth tracking.

[orhan maçin @MacinOrhan]

@mikeahorton
Robotik otonom sistemler:
Algılama
Konumlama
Kestirim
Planlama
Kontrol
@Intercognitive
Topluluğunundaha da gelişmesi dileğiyle
Saygılar.

[Mauricio | XMAQUINA @ZolKilla]

One of the largest robotics investment rounds on record, and among the biggest tech financings ever completed in Europe.

Proud to say @xmaquina has a seat at the table.

[orhan maçin @MacinOrhan]

https://t.co/o3gP3R9UF2

[orhan maçin @MacinOrhan]

https://t.co/z0qm4RTAyh

[Fluence ⏀ @fluence_project]

📢 New website is live with full GPU access available for new users!

We’re excited to announce that Fluence Console is now open for instant access to GPU Containers, Virtual Machines, and Bare Metal deployments.

👉 https://t.co/YoFTEDolE1 👈

🌎 Get easy access to 1400+ GPUs across 32 regions and 71 data centers now:

1️⃣ Sign up via GitHub, Google, or Email.
2️⃣ Top up your balance.
3️⃣ Deploy your first workload!

Key features 👇

[orhan maçin @MacinOrhan]

@fluence_project
$FLT
buradan oku
https://t.co/kx3bWTM6wI

[The Learning Pill 💊 @thelearningpill]

❇️ Team

The crew behind this project is absolutely stellar. Their co-founder & CEO @TheTomTrow was part of the founding team at Hedera Jashgraph.

And it gets better - the team includes folks who’ve worked at IBM, Google, Yandex, Near, and 1inch. pretty impressive, right?

[orhan maçin @MacinOrhan]

https://t.co/3pPFa0VoBm

[The Learning Pill 💊 @thelearningpill]

To make things easier, let me break it down for you!

In this thread we'll look at↓
❇️ Fluence overview
❇️ $FLT mechanics
❇️ Staking race
❇️ Team
❇️ Capital raise

[orhan maçin @MacinOrhan]

@fluence_project
$FLT
muazzam fırsat
Beni takip etmeye devam edin
Kriptoda ihtiyacın olanlar burada
https://t.co/lrXaKyVzUG

[Fred the Fox 🦊 @faith_freddy]

@DefiIgnas You really think none in the DePIN sector has gone mainstream? Filecoin? Bittensor?

But that's not even the point. You're right about the 100x potential part for most of the tickers in DePIN. It's a very promising sector trying to find its feet.

For $FLT it's more like 300x

[orhan maçin @MacinOrhan]

Böyle altyapısı olan bir proje 1 milon$ mı, hem de gpu,cpu darbğazında ( makine çapında)
https://t.co/B6RnaQ8mNF

[Crypto🧊house @cryptoicehouse]

Fluence $FLT an ultra‑micro‑cap #DePIN compute play (~$1.8M MC) with real revenue and a live “cloudless” GPU/VM marketplace. One of the most asymmetric setups in #AI infra right now.

Key points:

• Decentralized GPU + VM compute from tier‑IV data centers & permissionless nodes
• Live mainnet with real workloads, real billing, real savings
• Millions in annualized revenue already flowing to providers
• 25M+ FLT staked securing the network
• Up to 80% cheaper than AWS/Azure with on‑chain telemetry + verifiable performance
• “Cloudless” = no single provider risk, easy switching, SLA‑like reliability
• At ~$1.6–1.9M MC, it’s priced like a dead project — but it’s fully operational

Tokenomics:

1B supply, ~25% circulating. Usage‑tied token model (compute payments + staking). No hyper‑inflation. FDV still tiny. Value accrues from real demand, not emissions.

Bull case:

If Fluence scales supply 5–10x and captures cost‑sensitive AI workloads, revenue could compound fast. A 100x (~$160–190M MC) is still modest for a revenue‑generating AI compute network.

Bear case:

Micro‑cap liquidity, competition, and execution risk. Needs to prove it can scale like bigger DePIN peers.

Verdict:

Strong Candidate. A live, revenue‑positive compute marketplace trading at meme‑coin valuations. Real usage, real savings, real staked TVL — rare at this cap. Watch revenue + provider growth over the next 3–6 months.

Over all I’d give it a 8/10 👀

[orhan maçin @MacinOrhan]

https://t.co/6PrAbCWIgL

[Fluence ⏀ @fluence_project]

AI infra is entering a $5T buildout.

But reserved GPU clusters are still bought through intros, private quotes, hidden inventory, and weeks of negotiation.

No order book. No clearing price. That’s not a market. It’s a maze.

We’re about to change that. More soon.

[orhan maçin @MacinOrhan]

https://t.co/IsWvfdiLRO

[Stacy Muur @stacy_muur]

Each compute resource and job is secured with an $FLT stake.

Providers submit Capacity Commitments when adding resources, detailing the number of CPU cores and their rental prices.

Every CPU core must have an $FLT stake to be activated.