News for hackers. Stuff that matters.
Re:China squandered its chance to build a home-grown semicon... by 5822130027 on Sunday September 01, 2019 @01:22PM (#20853413)
I do not think China ever prioritized home grown semis, since its viewed as "plumbing" for more lucrative things built on top of chips ( software, etc).

Why focus on it when South Korea, Taiwan was already doing it ?

As the article states, it's a 500 billion dollar industry.

Compared to energy, defense it's much less important ( but critical ).

Now that the US is threatening to shut them off, they are being forced to invest more heavily.

I saw a graph the other day, within 2030 or something, china was poised to hold a monopsony position in semis, something like 80% of global semi consumption is going to come from China.

It's just surprising that the US would kick their biggest semiconductor customer in the head.

Re:China squandered its chance to build a home-grown semicon... by opportune on Sunday September 01, 2019 @01:42PM (#20853530)
Is that consumption measured by units or total value? I'm sure China could compete for cheap smartphone chips and microcontrollers but it would take them a while to make something like a Xeon.
Re:China squandered its chance to build a home-grown semicon... by lnsru on Sunday September 01, 2019 @02:04PM (#20853690)
You are making assumption, that they will stick to x86 architecture. What if they don’t want this old thing with all the microcode and hacks inside? What if they start on clean page with RISC-V or alike and software for it? Or China will just buy company building competitive chips: https://www.tomshardware.com/news/qualcomm-server-chip-exit-...
Re:China squandered its chance to build a home-grown semicon... by snaky on Sunday September 01, 2019 @08:39PM (#20855916)
The problem with monopsony is not any particular chip or technology. You can make (with huge investment and a couple of decades) or buy any technology. The problem is that you have to make the whole Digikey catalogue (and then some) all by yourself. From capacitors - all of them - types, capacities, sizes, form-factors, voltages, grades, to ADCs (no, buying Analog Devices wouldn't solve this, and you couldn't buy it anyway) and RF components, and everything in between. All of them, to stay competitive in what you can build with that components. That's where Soviet Union lost the game, and any modern country, including the United States would loose too, considering how the field has expanded since 1991.
Re:China squandered its chance to build a home-grown semicon... by opportune on Sunday September 01, 2019 @02:09PM (#20853713)
I just meant a high-quality vectorized server CPU by referring to the Xeon, didn't mean to imply any architecture
Re:China squandered its chance to build a home-grown semicon... by esmi on Sunday September 01, 2019 @05:10PM (#20854910)
Posted a few months ago. Admittedly it’s not a Xeon but these things are becoming commodities.

> Ara runs at 1.2 GHz in the typical corner (TT/0.80 V/25 oC), achieving a performance up to 34 DP-GFLOPS. In terms of energy efficiency, Ara achieves up to 67 DP-GFLOPS/W under the same conditions, which is 56% higher than similar vector processors found in literature.

https://news.ycombinator.com/item?id=20126280

Re:China squandered its chance to build a home-grown semicon... by 5822130027 on Sunday September 01, 2019 @02:26PM (#20853819)
https://www.eejournal.com/wp-content/uploads/2018/10/Fig-1-S...

market share, which means total value ( $$ ).

> but it would take them a while to make something like a Xeon.

By a while do you mean 10 years max ? that's not a long time.

Re:China squandered its chance to build a home-grown semicon... by Aperocky on Sunday September 01, 2019 @02:18PM (#20853764)
Xeon is no longer the top of the pack. AMD is blowing them out of the server right now.