NVIDIA is getting huge attention today because its latest earnings did more than beat expectations. They pushed back against one of the biggest fears in the market: that AI spending may be starting to slow.
NVIDIA reported $96.2 billion in revenue, up 106% year over year. Data center revenue reached $89 billion, up 117%, while adjusted EPS came in at $2.22. Gross margin stayed at 75%.
But the bigger story was the outlook.
NVIDIA expects fiscal 2028 revenue to grow about 70%. Analysts were expecting closer to 44%. That is a major gap.
The message is simple. Demand for AI computing is still very strong.
$NVIDIA said demand is running ahead of what it can currently supply. Memory costs and limited capacity are some of the main constraints.
The company also guided for $108 billion in Q3 revenue, while assuming zero data center compute revenue from China.
That matters because investors have spent months asking whether we are near “peak AI.”
This report gives them a different answer.
AI infrastructure spending is still growing fast, and NVIDIA remains one of the clearest companies showing just how strong that demand is.
The risks have not gone away. Memory costs could pressure margins. China restrictions remain. Competition is increasing. Investors still need to see whether massive AI infrastructure spending will eventually generate enough returns.
Hyperliquid’s $HYPE token burn highlights an interesting part of its token model.
Trading activity generates fees, and part of that value is used for $HYPE buybacks. The purchased tokens are then burned, reducing the amount in circulation.
So the model creates a direct link between platform activity and token supply.
More trading activity can mean more fees. More fees can mean more buybacks. More buybacks can mean fewer tokens in circulation.
The important part is whether this activity stays strong over time. That will tell us much more than a single large burn figure.
$16B of US Treasuries Are Already On-Chain One of the clearest examples of traditional finance moving into blockchain is happening with US Treasuries. Around $16 billion worth of Treasury-backed products now exist on-chain. These are not speculative assets. They represent exposure to short-term US government debt or funds holding those assets. In simple terms, an old financial asset is being represented in a new digital form. The reason this matters is simple. Once a Treasury product exists on-chain, it can potentially settle faster, move between compatible wallets, and interact with other blockchain-based financial applications. Who is leading this market? The largest individual products include: • Circle’s USYC at around $3B • BlackRock’s BUIDL at around $2.7B • Ondo’s USDY at around $2.1B to $2.15B $ONDO also has a strong position among DeFi-focused Treasury products, with its combined products around $2.6B. But the ranking is not the most important part. The bigger question is what people can actually do with these assets once they are on-chain. This is where DeFi gets interesting Most of the roughly $16B is simply being held to earn the underlying Treasury yield. Only around 0.7% is currently being used inside decentralized lending and borrowing markets. That gap is important. A tokenized Treasury sitting in a wallet is basically a digital version of a traditional investment. A tokenized Treasury being accepted as collateral, used in lending markets, or connected to other on-chain financial products becomes something much more powerful. This is also why the growth of regulated digital assets matters for exchanges and platforms such as Binance. As more traditional assets move on-chain, the line between traditional markets and digital markets becomes less clear. The first step was putting Treasuries on blockchain rails. The next step is making those assets useful across the wider on-chain economy. That second step could be much bigger than the first. #ONDO #FOMCWatch
Ethereum currently holds about $23.3B of the $44.7B tokenized RWA market.
That means roughly 52% of tokenized real-world assets are on Ethereum.
But what are these assets?
Mostly tokenized U.S. Treasuries, government funds, private credit, and gold. These are traditional assets represented as digital tokens that can move and settle on a blockchain.
Why does Ethereum have such a big share?
It has deep liquidity, mature smart contracts, strong security, and infrastructure that institutions already understand.
But Ethereum is not alone.
$BNB Chain has around $5.7B. zkSync Era has about $3.3B. Solana has roughly $2.6B. XRP Ledger is near $2.5B.
So the lesson is simple.
Tokenization is still early, but Ethereum has already become the main home for it.
The bigger question is whether it can keep that lead as more traditional assets move onchain.
The CLARITY Act delay sounds bearish for crypto. But the data tells a different story.
The U.S. Senate pushing the bill to September adds more uncertainty around crypto regulation. But institutional demand has not stopped.
Bitcoin ETFs have continued to see strong flows. Major investors are still entering regulated crypto products, while tokenization and digital asset infrastructure continue to develop.
The bigger issue is what the delay affects.
Smaller tokens, DeFi, staking, and newer crypto products still need clearer rules. Large institutions also want clear laws before making much bigger long-term allocations.
So I see the delay as a slowdown, not a shutdown.
If the bill moves forward in September, it could remove another major barrier for traditional capital.
If it gets pushed into 2027, institutions can still participate through products that already have clearer regulatory structures.
My view: crypto adoption is moving forward. The CLARITY Act could speed it up, but its delay has not stopped the trend.
Are Bitcoin ETF inflows finally putting pressure on sellers?
The data says demand is real, but calling it a seller exhaustion signal is still too early.
U.S. spot Bitcoin ETFs have seen roughly $51.8B in cumulative net inflows since launch. In early August, they pulled in around $850M in one week, showing that large buyers are still active.
But here is the important part.
Bitcoin did not immediately break higher. ETF flows later turned negative again, while BTC stayed around the low to mid $60K range. That tells us existing holders are still willing to sell into the demand.
So I would watch five things from here:
• ETF inflows staying strong for several weeks • Exchange BTC balances continuing to fall • Less selling from long-term holders • BTC reacting more strongly to new inflows • Less aggressive short positioning in derivatives
If these start moving together, the seller exhaustion argument becomes much stronger.
My view: ETF demand is clearly helping absorb supply, but one strong inflow week does not prove a supply squeeze.
The real signal comes when strong demand keeps arriving and sellers can no longer keep price contained.
That is when Bitcoin’s supply story gets interesting.
Bitcoin miners are still moving into AI and HPC, but the market is becoming much more selective.
In 2024 and early 2025, an AI deal could send a miner’s stock sharply higher. Now the reaction is much smaller. Across roughly 25 deals since June 2024, average same-day gains have fallen from about 24% to around 10%.
The reason is simple.
AI deals are no longer rare. Investors now want proof that these projects can actually be built and generate revenue.
Only around 25% of the AI and HPC capacity leased by miners is currently online. The rest is still being built or remains announced.
The cost is also much higher than traditional Bitcoin mining infrastructure. AI-ready facilities can require $8M to $15M per megawatt.
That is why the market is looking harder at funding, power availability, construction progress, and actual revenue.
The AI shift is still important for Bitcoin miners.