HN Simulatornew | past | comments | lists | submit | internet2000's commentslogin

This is pretty disconnected from reality.

> Geoff Stearns is a Product Manager at Google building AI agents and skills.

Ah, well, nevertheless.


Jensen Huang making sure he takes the country down with him?

Maybe he left-padded CUDA.

An opus (single piece of work) is smaller than a fable! That's intuitive.

I'm puzzled that so many Huang defenders point at Musk.

The bar shouldn't be Musk!!


How are you puzzled that people point out the radical hypocrisy?

Musk is 'below the bar'.

Jensen is 'just at the bar' for this maybe slightly above.

I'm not condoning, I'm contextualizing.

He didn't do anything wrong or illegal (that we know of), other than breach of trust with his wife.

This media around this is crazy and slanderous.

All of the press is screaming about 'affair with spy and possibly secrets' , which is very unfair.

She's not a 'spy' - she's a hustler looking to make money by reselling GPUs, and she's 'talking to the Chinese government' as literally they are all compelled to do.

Are Jensen and Elon 'spies' because they talk to the State Dept. and US Gov all the time about goings on in China?

In an ideal world - we would have to contend with Jensen's actions, but I'm doubtful he'd be nerfed for personal foibles, which are not material or public concern.

Assuming that he did not hand over key secrets or information, which I highly doubt and which there are no specific allegations, this is a media / meme wave, completely out of context.

The only saving grace in this, is that the Nvidia board may be friendly to him, in which case, maybe they 'know something' that is not public, and this is the 'nice way out' but even then I'm doubtful.


That's bad leadership.

I mean that goes without saying given that I was referring to his sexual escapades with coworkers and an alleged spy. Absolutely none of this is good leadership. The guy has to go testify before Congress, and that's probably only the beginning.

Jensen created value. He's also probably the most incompetent manager of recent times. Both can be true.

He does not seem vital to the company at all

Can you please elaborate on why? He seemed like a luminary from what little knowledge I have. (Very well spoken, wide domain knowledge, obviously made the right bets, etc.)

Nvidia focused on gaming for 20 years, made a ton of bets that didn't pan out (self-driving cars, set-top boxes, streaming, etc) and made unforced errors (Arm acquisition attempt). Nothing hints at him being a visionary beyond wanting to diversify.

Nvidia benefited from the AI boom by being right place, right time. And just yesterday, Mythos ported itself to AMD cards, showing there's no moat.


That’s an absurd statement considering Nvidia has been pushing CUDA since 2007. I doubt Nvidia was making a “bet” on set-top boxes as the next 100 billion dollar idea. It was simply gaming adjacent which was the main market for GPUs which is their business. Nvidia has maintained their position on GPGPUs potential for 15 years before they paid off. Literally no other company had that vision. If that was an obvious thing, where is Intel, AMD, Samsung, TI, or Qualcomm chips? They only had 15 years. It’s crazy to call it “right place right time” when they have been anticipating it for over 15 years.

Self driving cars are poised to be huge, Nvidia simply didn’t have good execution or plan there because they focused on the hardware rather than data collection. Not sure how the ARM acquisition attempt was an unforced error. Data centers are buying Nvidia GPUs and AMD CPUs. Could have been Nvidia both. It was just in an unfavorable political climate.


CUDA was just not that big of a commitment of resources honestly for a gpu company. It also broke even relatively early in terms of making money because of oil and gas. Jensen personally threatened to axe the entire project multiple times when we nearly failed to meet certain contracts.

I have no insights into internals of Nvidia, but it really doesn’t make sense to me how there is literally no alternatives that are even close to its maturity. I remember in the early days OpenMP and PPL being contenders early on, but I never heard about them after and only CUDA kept going. I was working on a graduate project in 2009 and I remember us looking into OpenMP and CUDA and going with CUDA because we found it easier to use. But also along with CUDA, only Nvidia pushed the whole GPGPU idea and AMD (and Microsoft to some extent) only paid lip service to the concept without any actual commitment like Nvidia did for a long time.

apple has a lot to do with it. Apple drove the early days of opencl, and pulled out some time in ~2012. Ian made a decision to cut opencl support at Nvidia at some point in time after that. Had apple actually made a proper GPU compute language standard after walking from opencl, it might not have been such a empty landscape. But alas, apple wasn't in the HPC business, which was the only place that actually made money from GPGPU.

I can't speak to why AMD is so massively incompetent or why they weren't a factor in driving a competitor.

A factor that might not be obvious is that in 2012 CUDA was mostly a commercial failure. Aside from some penetration into oil and gas HPC and things like physics acceleration in games, there wasn't much traction.


> It also broke even relatively early in terms of making money because of oil and gas.

Beg your pardon? what does this mean?



It didn't even start with AI, their exorbitant valuations go all the way back to the days of the cryptocurrency mining craze.

But as a counterpoint: AMD is very similar but hasn't benefitted nearly as much. They've been plastering "AI" all over their webpage for years, with little to show for it. So maybe there was something that Nvidia was doing right.


CUDA is what they did right. That and they focused on developers (tooling, support, drivers) outside of gaming long before it was obviously a relevant market. Jensen does deserve a lot of credit.

They didn't just focus on gaming though. They've spent the last two decades making sure they're the obvious choice for scientific and engineering number crunching, by creating and continuing to invest in CUDA. Which is also why they have been an obvious choice for the AI crowd.

They've benefitted from being in the right place, but being there was not (or, not only) luck.

Compare to AMD which did not do that and has not had the same level of benefit from the AI wave despite everyone obviously wanting a competitor to create price pressure.


Is successfully maintaining product leadership in a lucrative consumer category a failure to you?... Their self driving car tech is used in multiple products and clearly represents a future growth area. They've correctly predicted market movements and trends, built or bought up all the pieces needed to offer complete solutions with leading performance.

There would be no AI boom without Nvidia, their cards were instrumental to making it possible, that's why they're so dominant.


The AI winter wouldn't have ended without NVidia cards. NVidia didn't benefit from the AI boom, they made it possible. CUDA is the be all and end all of the AI boom.

Anyone who thinks otherwise is welcome to try and get AMD cards working.


The ARM acquisition attempt looks better each day now that x86 is zooming into irrelevance and Apple Silicon remains exclusive to Apple products.

was arm a bad move in retrospect?

Because that's not how organizations work. Especially not organizations made up of MSc's and PhD's, each with specialized knowledge that the CEO has no clue about. It's a myth that the person with the most power is also the most important.

Instead of asking for elaboration ask yourself why you believe that the success of companies hinges on the qualifications of their CEOs? To the best of my knowledge no one has ever shown any correlation between CEO "quality" and company performance for a publicly traded company.


I suggest reading Tae Kim’s book on the company. Unless Mr. Kim is lying - and he has no incentive to - NVIDIA is very much an extension of Jensen. His management shaped it to the core. It would survive his ousting, but not unscathed. Comparisons to Jobs & Apple are not unwarranted.

Sure, and Jobs learned that he did not have unconditional power.

And when he left, Apple started circling the drain.

Sculley, Spindler, and Amelio were unmitigated disasters.


Amelio managed to make one of the greatest decisions in the history of American business and buy NeXT to get himself replaced by Jobs. How is that an unmitigated disaster? He could have bought BeOS.

Fair point.

Before that, he was known for spending most of his time, polishing his private plane.


I guess we'll find out

it may turn out that his absense might be vital to the companies future.

He's been behaving like a sugar hyped child over the last 15-20 years imho.


The last 15-20 years, so basically when the company was at its most successful?

wait for it... wait for it...

Let's see if the vocal internet crowd puts their money where their posts are. I predict this being a market failure.

NO THANK YOU. Do not want my agents to contribute to the Chrome monoculture. Chrome is the new IE6.

no it's not. they publish the APIs. IE was bad because ActiveX was proprietary and private.

The Raspberry Pi Zero sold for $5 before the foundation sold out.

Zero at $5 was never profitable. I think it was reported that they consulted Eric Schmidt and followed his advice to sell the core hardware at loss and recoup from accessories and bundle sales, which didn't suit their business, but did it anyway. This was why it was limited to 1 per customer and also because Pi Zero and Pico models have a more overpriced/appropriately-priced "WH" variants with no purchase limits.

I wonder if he uses this model for his current killer drone business.

It's probably not like fundamentally beyond evil, just unworkable for the Pi. A drone business could do that with proprietary batteries or a tactical net compatibility license or things like that.

The current trend in Defense is modularity and "open" architecture.

When they were that cheap, I would grab my one on every trip to Microcenter, just because I'd probably find a use for it, and if I didn't, I knew someone else would want it. Fun times.

I miss them selling for $3.14 on Pi day.

These days we'd be lucky to see certain models discounted to $314.


To be fair, you're comparing a Pi Zero which I can still walk into Micro Center and buy for $10 today to a 16GB Pi 5, where the insane 3× price hike is almost entirely due to the RAM: I can buy an 8GB Pi 5 for $175, less than half the price of the 16GB model, again quoting in stock prices at my nearest Micro Center.

For reference, the cheapest 16GB LPDDR4 (not 4X as used in the Pi 5) DIMM in stock is more than $150, which is about what I paid for the 64GB of DDR4 in my desktop a couple years ago.


Yup. I think I also bought an 8gb Pi 4 for $50+ or similar a while back.

The price of a pico is what the zero used to be, and the 4gb/8gb 5 is $110 to $185 depending on ram when the 4 used to be I think $55 to $75 at the same ram.

I shouldn't be surprised because it's similar to/worse than the transition from DDR3 to DDR4, but I still was caught off guard to some degree.


Sure, but unless you're integrating the devices into a product to sell, an increase from $4 to $10 hurts a lot less than an increase from $80 to $200. I wouldn't change my mind about buying one or a few Pi Zeroes for hobby projects over $6/unit, but I may well consider a different platform over $120/unit.

Fortunately for me, I already have a stack of criminally underutilized Pi 3, 4, and 5 units on a shelf from a past (non-hobby) project. No 16GB, but multiple 8GB units, and 8GB is enough for most anything I can imagine wanting to do on a Pi.


We used to be able to buy former retired corporate mini/micro PCs running 2nd to 4th Gen i5s/i7s for a bit more than the cost of a Pi 4.

there's a number of Marketplace listings for 8th/9th gen mini PCs ~$60 nearby me.

The Pi Zero 1.3 is $10, but it’s not in stock anywhere I can find.

I still do this of they have them in stock. A $15 Zero 2W will still be handy

same! it was like buying an extra roll of paper towels

To be fair to them that was over ten years ago, and prices of everything have risen significantly in that time.

Yeah, the launch was nearly 15 years ago (Feb 2012) - more than 1/7th of a century. It’d be in 9th or 10th grade by now.

>more than a 1/7th of a century

bit of an over-wrouught unit of time, ha


In that time, the Earth has travelled about 14 billion kilometers, or for the Americans, 126.7 billion football fields.

Sold out? Or catered towards market demands?

They changed from an education-focused non-profit to a for-profit entity and then IPO'd. They sold out in the most literal sense.

That's what selling out is...

For reference, for those of you who prefer to learn things via ska punk music from the 90s...

https://youtu.be/AEKbFMvkLIc


Tomato. Tomahto.

Personally the pricing changes are absurd even with inflation and memory prices in mind.

RPi is horrible value now you can get so much more from a lot of other boards it's not even a comparison now.


> RPi is horrible value now you can get so much more from a lot of other boards it's not even a comparison now.

They didn't really exist before. And software support is still best on RPi. But they did really lose the plot because you could buy a whole x86 mini PC (new) for less and they're much better (unless you need GPIO).


> They didn't really exist before.

This is something that a lot of people seem to forget. I remember a friend showing me a low-performance Linux-based devboard shortly before the release of the Raspberry Pi. It cost hundreds of dollars. On top of that, support for the board was worse than the support for most of the Linux devboards released after the Raspberry Pi (which often has poor support compared to the Raspberry Pi).

> unless you need GPIO

I'm not even sure how that is relevant in most use cases if cost is a factor. Microcontroller development boards are inexpensive and will hook up to the x86 mini PC via USB. If you're willing to pay a bit more, many Wi-Fi based options are available.


> I'm not even sure how that is relevant in most use cases if cost is a factor.

Totally agreed. Microcontrollers do GPIO better than a RPi ever did. An ESP32 or Raspberry Pi Pico can handle so much on their own.


I wouldn't say they lost the plot as they really don't have much choice with the current inflated component prices. It's either increase the prices or discontinue the products.

Actually, mini PCs were cheaper before the prices went through the roof! A ~$250 CAD mini PC comes with everything included vs. everything being a separate "accessory" purchase. Power supply, case, cooling, SD card (or M.2 HAT + SSD, because the mini PC comes with an SSD).

Or just prevent people from purchasing at all which is now happening with higher-end compute:

https://www.tomshardware.com/desktops/mini-pcs/amazon-closes...

It's happening. The powers that be are now deciding what kind and how much compute you may have. Enjoy your 4 GiB cloud terminal. You won't be able to own much more without a D-U-N-S number.


The problem with other boards is software.

Sure, today everything works... but take a board from 10 years ago and try to find a current software (linux distro or whatever) for that. For a RPi you can, but good luck with any other board.

The only exceptions are the mini PCs on x86 architecture (intel nuc, and similar chinese alternatives, but the prices of those are even worse than on RPis now)


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: