Mach

Aerospace engineer | Aircraft engines by day, hooked on markets by night | Long AI since GPT-2

[Mach @MachStack]

AMD beat NVIDIA by 5.9% in published peak HBM4 bandwidth. It did so with 50% more stacks.

AMD won the headline. NVIDIA may have built the better configuration for an HBM shortage.

Why AMD's HBM4 advantage could backfire: https://t.co/KYRCE5PmAZ

[Mach @MachStack]

@StockMKTNewz @Fiscal_ai and if you look at profits the gap is even crazier

Nvidia $NVDA made 8.3x AMD's revenue last quarter, but 26x its net income

[Mach @MachStack]

I get the long-term argument, but I'm not sure America is a great comparison

America gave access to huge amounts of productive land and resources that Europe either didn't have or couldn't produce at anything close to the same scale. There was a clear economic reason to exploit that new territory

a lot of the space economy being pitched today feels different, you're often taking things we already do on Earth, compute, mining, energy, and trying to make them work somewhere with a massive cost disadvantage

maybe those costs collapse over time, but that's the part I'm trying to understand, what does space eventually do so much better than Earth that it creates an entirely new economy?

[Mach @MachStack]

Even the best AI agent today gets a 30% resolution rate on expert-built scientific workflows

Terminal-Bench-Science tested 70 tasks across life, physical, Earth, mathematical and engineering sciences, with 3 independent runs per task

Claude Opus 5 + Claude Code: 30.0%
GPT-5.6 Sol + Codex: 22.4%
Claude Fable 5 + Claude Code: 21.4%
GLM 5.3 + Claude Code: 8.1%

the cost gap is pretty crazy too

Sol basically matched Fable for $4.2K vs $14.2K, while actually using more tokens, 8.4B vs 6.4B
these aren't exam questions either, the agent has to actually do the work: analyses, simulations, proofs, code, etc. and each task has its own verifier

this benchmark, just like Humanity's Last Exam or FrontierMath, was specifically built to be hard for frontier models

so current models can have a much bigger impact today than a 30% score here might make you think

[Mach @MachStack]

@Dr_Singularity Welinder clarified this later, he thinks AGI will come pretty fast

the few years are for the physical constraints around real-time inference and cheap edge hardware, not AGI itself

[Mach @MachStack]

@sattyyouneed For the same reason people still buy Nvidia GPUs when cheaper alternatives exist

Being cheaper and good enough isn't the same as being the best. The US still holds the lead at the frontier

[Mach @MachStack]

@TimJayas Yeah, there have been some rumors about the new Fable but so far everything that's come out makes it look like it'll be below Astra

we'll see

[Mach @MachStack]

Most of that increase is related to memory procurement, Nvidia $NVDA is basically locking in supply years ahead

even with memory prices going crazy they only expect gross margins to go from 75% to ~74% next quarter, before bottoming around 71-72%

looks a lot more like preparing for massive demand than a bearish signal to me

[Mach @MachStack]

OpenAI spent the last few generations compressing intelligence, getting more capability out of fewer reasoning tokens

if these Astra leaks are real and Max is suddenly thinking way longer than Sol, maybe they're starting to spend those efficiency gains again on test-time compute

especially interesting now that OpenAI says monitoring Astra adds ~20% extra inference compute, and they've found longer reasoning traces are easier to monitor

[Lentils @Lentils80]

🚨 Major Scoop:

OpenAI has just expanded internal testing of GPT Astra, codenamed "mozaik-alpha-fdm", signifying release might be near

AND of course like previous times, we have the FIRST EVER public outputs of it for y'all😉

Both outputs are zero-shot on Max effort. Frontend *might* have finally been fixed, and the model gives a lot of attention to details

[Mach @MachStack]

AI is another wave that's going to spread into everything, just like electricity and the internet did.

The difference is the adoption curve is moving much faster.

Electricity added close to 0.5pp a year to productivity growth at its peak, the internet boom around 1.5pp, and both took decades to diffuse through the economy.

If AI really gets knowledge workers to 2-3x their current output and we compress that adoption curve into just a few years, it's hard to see aggregate growth staying basically the same.

[Mach @MachStack]

@PolymarketMoney Nvidia $NVDA is already generating around $700M in operating income per day, growing 19% QoQ

their Q3 guidance implies around $780M per day, so $750M by year end might actually be conservative

[Mach @MachStack]

@Dr_Singularity Every major technological shift has its luddites, the world will keep moving forward without them

[Mach @MachStack]

@DavidOndrej1 the fact that Europe has so few major tech companies should've already told you how this survey was gonna look

this continent is doomed

[Mach @MachStack]

GAAP operating income generated per day:

Nvidia: $700 million
Alphabet: $448 million
Microsoft: $446 million
Apple: $392 million
Amazon: $302 million
Meta: $206 million
Broadcom: $119 million
Visa: $76 million
Netflix: $46 million
AMD: $22 million
Uber: $21 million
Tesla: $4 million

[Mach @MachStack]

Zero productivity gains yeah, that's why developers are already getting to the point where they can't work without AI

have you ever stopped to think that a lot of tasks that don't look like coding can still be accelerated with code? it's only a matter of time before that spreads to other sectors

[Mach @MachStack]

AI is another wave that's going to spread into everything, just like electricity and the internet did.

The difference is the adoption curve is moving much faster.

Electricity added close to 0.5pp a year to productivity growth at its peak, the internet/computing boom around 1.5pp, and both had pretty slow adoption curves despite being revolutionary technologies.

It's not hard to imagine what happens when you compress the AI adoption curve into just a few years.

[Mach @MachStack]

@kimmonismus we've gone around 3 months without really pushing the frontier forward, it's time to move that bar again

[Mach @MachStack]

this wasn't an 8% pump just because of a double beat

the market got a reminder that Nvidia $NVDA's moat isn't something you tear down overnight. That's 30+ years of trial and error

anyone who's worked in engineering knows how massive the difference a few years of iteration can make

[Mach @MachStack]

@forgebitz don't worry, we've got every mom and grandma on the planet thanking AI for us