It’s 128 bit. I’d say it needs a bigger GPU and a 256 bit bus to be “M2 Pro” class.

Eh actually the 4060 TI is way better for LLMs :P With Nvidia its all about VRAM capacity.

Speaking as an holder of AMD stock since ot was $8, and an all AMD CPU user, IMO Lisa Su is either an absolute idiot or colliding with her cousin, the CEO of Nvidia.

All they had to do was lift vram restrictions on consumer GPUs (so their OEMs could double the VRAM up) and sick like four engineers on bugs blocking the AI space, and they would be dominating the AI space and eating Nvidia’s pie…

And they didn’t. Like, its two phonecalls, thats it.

Intel had monumental problems it has to solve and struggles, but AMD has tiny ones they inexplicably ignore. Its mind boggling.

Heh, part of the reason is that OpenAI stagnated and that most AI products are still dumb.

And people, especially bad actors, use them stupidly too. Like are they running a cot finetune with constrained grammer tuned just for their… Blah blah blah, nope. They’re just using cheap OpenAI API, maybe with a jailbreak prompt, if not just prompting the webpage, out of sheer laziness and ignorance.

The problem has always been software support. If Intel wants a piece of the AI pie, they need fantastic software support. AMD has always been a bit lackluster here, whereas Intel has done a pretty decent job in the past (esp. on Linux, their drivers rock), so they would need to double down if they truly want to get after it.

Actually AMD is pretty okay for running LLMs and other ML workloads. Many libraries now explicitly target rocm, you can just plop down vllm or the llama.cpp server and have it work with big models out of the box. There are some major issues (like flash-attention), but its quite usable.

Intel though? Their software is a mess. You have to jump throigh all sorts of hoops, use ancient builds of pytorch, use their own quantizations and such to get anything working, fix Python errors, and forget about batched enterprise backends like vllm. And this is just their IGPs and Arc, forget trying to use the vaunted NPUs for anything.

This could change if they actually had a cheap 48GB GPU (or a big APU) for AI devs to target… But they don’t. And no one is renting Gaudi to build in support because its not even availible anywhere.

EDIT: oh, and one weird thing is the volume of Intel software support is high. Like they have all sorts of cool libraries, they make contributions to open projects… But its all disjointed and fragmented. Like theres no leadership or unified push, just random efforts flailing around.

Its a lower midrange only launch like it appears to be, it will be extremely unprofitable. AMD may even eat large chunks of this market with the Strix Halo APU, which could be similar to the B570 with no need for a discrete GPU.

Theres actually a big and growing demand for ANY high VRAM GPU for the LLM crowd (that AMD is ignoring for inexplicable reasons beyond Strix Halo) but it appears Intel can’t even compete there. No 256 bit APU, their GPU is 192 bit so capped at like 24GB…

They tried all this:

1. Not sure about this, but it appears AMD is simply out designing them. Some concepts like the many-little-core SKUs seem promising, but ultimately the EPYC MCM design is fundamentally very good here. And… Delays. Delays are killing them here.

2. This was Xe-HPC, the Falcon Shores APU, the Falcon Shores GPU, Gaudi… They’re so late to everything it didn’t work and it appears they’ve basically given up on the whole line besides consumer inference products, which is also kinda meager atm. And even AMD is mightily struggling here, with hardware that is straight up bigger/faster than Nvidia.

3. An M Pro esque chip was also in the plans, but seemingly canceled? Or way behind AMD, at least. And OEMs have repeatedly rejected their GPU heavy designs like Broadwell eDRAM and the AMD collab chip, as they’re kinda idiots and Intel is at their mercy. And the laptop chips they are selling now are basically their best shot at an “M” chip and arguably one of their most decent products.

4. They tried, and no one bit. Who can blame them, given Intel’s history of delays?

Its all the delays! Its destroying them.

I mean I’d guess I’d press on with Xe if I were CEO, but if they can’t launch anything on time what does it matter?

Mmm, I hope the fab survives on their own…

The sentiment was not bad. TSMC is a shining example of how fab subsidies can be a good idea, and Intel’s fabs going under is bad and basically irreplaceable. Like… I am still happy with my tax dollars taking the risk, and Intel was clearly trying to right the ship when CHIPS was conceived.

But theres clearly rot in Intel. Thats a big difference I guess, as TSMC was built from the ground up (in a time where that was possible) while Intel is already weighed down with its sins.

I feel like we need a phrase for this.

“It’s the algorithm, stupid!”

And if they can be sued for algorithmic recommendations at all, I feel like their resort would be going back to pure “user” recommendation like community managed feeds, following other’s recs and such. Which would kill SO MANY birds with one stone.

Maybe I’m even more cynical, but it feels like us evangelical christianity is being used as a tool. A gullible fool.

So all these old gheezers are vaguely pushing for a US Christian nationalism… But coronated someone who doesn’t really care about that beyond how they can elect him, and basically shoved their values out the the party’s window.

And US youth is increasingly less Christian. Including Trump supporting youth.

So… I’m not even worried about this long term, relative to everything else to worry about. There will be short term scares, but post Trump Trumpism is not going to be very religious.

So what IS their strategy now?

Some of Pat’s initiatives were good (stay the course with Xe and fabs, which take a long time to pan out), but they kept delaying everything!

Yet Intel is kind of screwed without good graphics or ML IP.

If they spin off the fabs, I feel like they are really screwed, as they will be left with nothing but shrinking businesses and no multi year efforts to get out of it.

Like… Even theoretically, I dont know what I would do to right Intel as CEO unless they can fix whatever is causing consistent delays, and clearly thats not happening. What is their path?

NFT artists kind of hide behind the oldschool art world to justify it…

I think it shows how stupid that world is.

The point of art is to move people, to make them think and feel something by conveying the artist’s thoughts, not be a store of value. And while I’ve been to museums with gorgeous pieces, if you’re paying millions for a painting, at some point most of that value is the gratification of hoarding it. That same money could buy you an incredible experience in today’s art landscape, but it’s not about the experience, is it? And NFTs are like the perfect deconstruction of that.

That being said, OP I am downvoting your post because eyeballs are exactly what crypto bros want, no offense :P

Stadia.

…Is he gonna hire Phil Spencer?

Will it go as well as X AI?

Grok is the laughing stock of the LLM world, and he also completely lied about his promise to make them open source. The latest Qwen release from Alibaba (open, locally runnable weights, tiny, Apache, freaking smart) has rendered his obscene hoard of H100s basically obsolete, with a miniscule fraction of the money.

It’s even simpler than that, it’s people being told what to think.

I think “people” speaking very generally used to not read a ton of news, heard stuff from the grapevine, and so on. “Elites” and news junkies had somewhat more monolithic sources.

And that’s not true anymore. Nearly every “average” person’s life is now dominated by a personalized feed, a podcast, TV, radio, chatroom, whatever, and it’s having an outsized influence compared to their observations of reality now.

It’s my belief that there’s basically nothing Biden could have done to alter this (other than regulating algorithms, and it’s far too late) and ultimately it’s the DNC’s fault for “taking the high road” and not playing the propaganda game.

I do absolutely agree.

It’s like “don’t feed the trolls,” the most important internet wisdom ever, has been long forgotten and is tearing down society.

I don’t know how to fix it. We should have regulated social media algorithms years ago, but now it’s way too late, and people’s information hygiene is not getting better. Any kind of news outlet is basically compelled to cover it for financial reasons less they fade into obscurity, and algorithms love it and aren’t going to change.

Unfortunately we basically voted him into office.

It’s the “Trump problem” now. Now he has the power to act on these insane opinions and threats, so ultimately they are newsworthy even if I largely agree with you.

That’s the strange thing, it’s completely open with tons of news outlets constantly calling it out, not an “taboo secret” like of like old facist governments. And in all liklihood, it will stay that way.

The filter bubble and American apathy is just that powerful, I guess?

That is an American pasttime