Live data from Hacker News

AMD CEO sees chips from TSMC's US plant costing 5%-20% more

bloomberg.com

191–200 of 761 posts

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#191

Earlier quoted context omitted.

Ajinomoto (Japanese company) is nearly the sole manufacturer of build up film need for CPU manufacturing (for about 30 years). There's all kinds of stuff like this in supply chains. Low profit, high barrier to entry critical items.

Wild to learn this is the same company selling MSG (mono sodium glutamate) and build up film - https://www.ajinomoto.com/innovation/our_innovation/buildupf... - https://www.ajinomoto.com/brands/aji-no-moto

Not really shocking considering how Japan likes conglomerates.

Hyundai makes cars and military weapons and probably thousands of other things that aren’t even related to each other, don’t know if they still make computers.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#192
post #34

What I really want to know, from someone who does know: Is Intel cooked? Like, will they be able to manufacture chips that compete with TSMC? They used to be a crown-jewel of US tech. But it seems like every time I read the news, they are announcing a delay or shutting down some product.

I'm not into hardware but I remember when AMD was sneered at, and all real CPUs were Intel. Then Ryzen happened. My meta conclusion is that its super hard to tell when someone is done, and it can change quickly. Or not. Sometimes it if looks like terminal decline, it simply is terminal decline.

Everyone thought AMD was done. Intel is going through a difficult transition, but if they can make 18a /14a work and keep improving their GPU line we could be having the same conversation about AMD in 10 years.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#194
post #157
post #148

Earlier quoted context omitted.

But not because of its ISA. I mean, to first approximation everything is a "flop" in semiconductor architectures (or really in tech in general). The population of genuinely successful products is a tiny fraction of the stuff people tried to sell. In this particular case: ia64 leaned hard into wide VLIW in an era where growing transistor budgets made it possible to decode and issue traditional instructions in parallel…

VLIW came with an implication that static scheduling would win out. The deeply OoO chips you see now have a very different architecture to support that: Itanium was much more a DSP like thing.

Even in VLIW, DRAM fetches are slow, instructions have variable latency and write-before-retire register collisions require renaming. Itanium would have gotten there at some point. OO isn't an optional feature for high performance systems and that was clear even in the 90's.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#195
post #89

Earlier quoted context omitted.

Intel's fab issues are overstated in my opinion. They were stuck on 14nm for a very long time because they bit off too much with 10nm. People act like that means ALL research in nodes smaller than 10nm must have stopped, but that's simply not true as research into tech and materials needed for smaller nodes happens in parallel. It's also noteworthy that GAAFET being a complete redesign of major parts of the manufactu…

They just farmed out the compute section of Nova Lake to TSMC which is a sad statement (probably a good business decision, though).

They just farmed out the compute section of Nova Lake to TSMC which is a sad statement

Apple farms out its displays to Samsung, a competitor. It's just how business is done.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#196
post #194
post #157

Earlier quoted context omitted.

VLIW came with an implication that static scheduling would win out. The deeply OoO chips you see now have a very different architecture to support that: Itanium was much more a DSP like thing.

Even in VLIW, DRAM fetches are slow, instructions have variable latency and write-before-retire register collisions require renaming. Itanium would have gotten there at some point. OO isn't an optional feature for high performance systems and that was clear even in the 90's.

If you have that what's the VLIW getting you?

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#197
post #154

Earlier quoted context omitted.

What would be good for the rest of the world is if there were SOTA chips that were not produced by the US nor Taiwan. Frankly, even the ones produced in Taiwan are under US control. The world needs a healthy diversified CPU/GPU chip market. At least there is ARM on the CPU side, but it's not nearly enough.

Where could it be? The places with abundant energy are where these things establish. US is about at the lower limit. Korea, Taiwan, Japan. China has SMIC and Huawei. But Europe doesn’t have enough energy to run air conditioning. They’d struggle to add more industry. India has power shortages. Africa isn’t reliable. Australia? South America too unstable.

>But Europe doesn’t have enough energy to run air conditioning.

That's just silly.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#198
post #174

I see the point being made here, and yeah 5%-20% extra for what amounts to insurance against geopolitics isn't too bad, but doesn't this all fall apart when China catches up? That 5%-20% is worth it now because no one else can fabricate competing chips. In a competitive market, 5%-20% can be the difference between having the price edge or not. I understand why the USA wants TSMC to manufacture outside of Taiwan, but…

With China, the issue isn't really the quality of the product. It's the geopolitical issues. Japan, Korea, Taiwan, and other countries even Philippines and Vietnam don't want to depend too much on China. A lot of island disputes and so on and so forth. My guess right now is that China will never catch up because Europe, US, Australia, and many other developing countries will avoid depending on China critically. This…

It's a real shame that the U.S is alienating it's allies through aggressive economic policy. Maybe we'll find ourselves on the wrong side of that economically resilient policy.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#199

Earlier quoted context omitted.

And yet Itanium flopped.

By all accounts I can find Itanium performance was good, perhaps even great when writing assembly. It seems to reinforce the point that ISA doesn't really matter. But let's be clear: Of course ISA matters. It's just as trivial to make a bad ISA as it is a bad syntax. But does the ISA of modern superscalar processors matter? Probably a bit, but certainly not a whole lot.

It wasn't good vs peer competitors at the time (HP-PA, DEC Alpha, IBM RS/6000, even MIPS). And it was very expensive. Huge die. It was an expensive, strange thing, that didn't have the necessary 2X peer performance advantage to offset those issues.

Re: AMD CEO sees chips from TSMC's US plant costing 5%-20% more

#200
post #34

What I really want to know, from someone who does know: Is Intel cooked? Like, will they be able to manufacture chips that compete with TSMC? They used to be a crown-jewel of US tech. But it seems like every time I read the news, they are announcing a delay or shutting down some product.

Intel’s chances of being a foundry for others are close to 0. It does not matter how good their process technology is. The problem is that Intel was an IP thief in the 80s and 90s; being a foundry requires trusting Intel with the exact IP Intel was known for stealing and nobody wants to take the risk.
Post reply on HN