You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
If you look closely at the slides in the video, you'll notice that some of the Intel performance slides are marked as "projected or estimated" with "an accuracy of +/- 10%". This seems a little loose given that this is - I assume - working silicon. Curious also that Intel obfuscated the die...
Wouldn't it be fun if just for a change the people who make the forecasts about early new tech like AI had to publish the error bars alongside their forecasts ?
The idea that there can be an absolute, leakproof export control ban on high end chips to China seems completely impractical in the modern world. Yes, you can - and probably should - have export control regulations as we've always had for advanced military or dual use technologies. And you can...
While I agree with the basic premise and that ongoing "subscription" education already matters more than degrees and that this is now technically possible, there are a few practical reasons why I don't see university education becoming obsolete.
Firstly, those of us claiming that we're readily...
"Never interrupt your enemy when he's making a mistake" as Napoleon famously put it. I'm not saying China's an "enemy" here - merely recalling an instructive piece of advice.
If China continues to heavily subsidise everything they export, surely eventually the losses become overpowering - and...
There's not sufficient evidence in any of these reports to say that this is anything more than an accident that happened at TSMC. Nothing at this stage to say who or what caused this or why it happened.
I certainly don't think that as - I assume mainly - professional engineers we should be...
Who else can ultimately pay other than the US consumer ? No imports = no tariffs = no US customers.
For sure, it can only hurt the US consumer by reducing choice and competition and reducing pressure to lower prices. The fact that Telsa has dramatically lowered prices (almost unheard of in auto...
Putting to one side the USA/China disputes, the one certainty I can see is that this - as well as many other policies (and yes, that includes the CHIPS Act) - will create more inflation in the US. Which will inevitably spread to other countries. Quite how Biden got his name onto an "Inflation...
Not sure what your graph is supposed to prove there. Looks like Zimbabwe ! Suggests there's catastrophic inflation in China if the money supply is increasing that fast - or that the conversion of Chinese M2 to US$ is very, very wrong !!! Either way, that's not the graph of a healthy economy -...
The US has won every technology race it's been in so far. They beat the German dictatorship. And the Soviet/Russian one (*). You've made no case why it would be any different this time. Frankly, it won't.
(*) Forgot to mention the Japanese one. And perhaps the Italian one too. And most recently...
Intel is already profitable ! Historically fantastically profitable. They could run the x86 business as a hugely profitable cash cow for years if they went fabless now. But they are not pursuing this "clear the decks" strategy.
The real question is whether they'll see a return on the multi...
The PE firms aren't typically putting up much - if any - of their own money. It's borrowed from banks/shadow banks/etc - often against the assets (in this case I assume the fabs/fab equipment). The downside risk is typically very limited for them. They take their big fees regardless.
Well, the only way the money could be cheaper from PE funding is if they can offset their debts (the bank loans raised to fund the venture) against the profits. Which they always do. As well as paying themselves huge fees. In other words, the "cheaper money" is a result of lower taxes on the...
I'm not sure this necessarily is competing with customers.
Let's roll the clock back to the mid 1990s. Mobile phone companies like Nokia are just starting to integrate DSP, MCU and ASIC onto a single wireless application processor chip. TI (who produce the chip for Nokia) learns from...
Most interesting.
So you can shop for approval either direct at the FCC or at one of its authorised agents like Intertek who have the authority to issue licences.
At a rough count Intertek has around 60 locations in mainland China.
Given what we know about the independence of Chinese...
I'm sure you're correct there. Every self-respecting country has its own regulator.
But the US market is huge (and ~100X Singapore's).
On the other hand, some regulators are more respected and valued than others. German DIN standards used to be the gold standard in Europe (as you'll no doubt...
I think his question is partly about the implied element of "self-certification" in the current process.
I don't understand how the FCC certification process actually works, but it might be worth someone who does clarifying that.
I'm assuming there must be at least 2 signoff levels needed:
1)...
OK, but what's the limit on what the subsidy should be then ?
Let's keep things simple and assume govt X is subsidising a fixed percentage of company Y's leading edge (let's call it around 3nm) logic wafer fab and no ongoing handouts. What's the rough % estimate where the subsidies outweigh the...
