The industry seems to be moving towards for specialised chips as opposed to general purpose chips. Do you think this will increase the value added for chip design firms? Also doesn't India have a sizeable footprint in chip design?
#1 You are spot on about specialized chips. Nvidia makes GPU chips for supercomputers, LLM, and gaming. Qualcomm focuses on chips for wireless technology like 5G, vehicles, and laptops.
AMD does GPUs, TVs, and accessory electronics.
There's so much and there's still many Japanese, Korean, and Chinese firms I could mention as well
#2.
Depends what you mean by "sizeable"
Yes India has some chip design from Tata, Wipro, and Infosys. several domestic firms and many startups.
But in terms of fabrication or assembly & test exports, India footprint is slightly under $700M, an order of magnitude less than Thailand, Philippines, Vietnam, or Malaysia.
Depends on what your objective is. If your objective is military and strategic independence, then yes, your focus should on acquiring domestic packaging, assembly and fabrication capabilities. But if your goal is growth and productivity why not double down on your comparative advantage which is chip design which seems to be more value added.
As is usually the case, the chip design industry emerged in India not because of government but despite it. Because all chip design requires is a bunch of nerds in an office unlike fabrication which requires a lot of land and specialised infrastructure.
On a side note, what is your opinion on Risc-five? Given we're moving to a world of increasing specialisation, closed source software tools for development might prove to be a bottleneck. In such a world, Risc-five is a welcome addition. Do you see any other promising open source tools cropping up?
I doubt they'll be successful. It's like him trying to stop Chinese civil servants from using Excel. I think if Risc-five takes off it create a cottage industry of chip designers in the middle and upper income countries.
Well all NIMBY resistance can be based on "reasonable" environmental concerns. Also people may not want to live near a chip fab.
1. Fabs use many chemicals and gases in their manufacturing processes, which could pose risks to residents if not managed well. These substances could release air and water pollution.
2. Fabs can be noisy which could disrupt nearby communities.
The issue in America is that NEPA is so powerful that many infrastructure, housing, and industrial production has slowed down considerably since the 1970s ever since the NEPA act has given communities the ability to raise legitimate concerns of construction related to environmental impact, public health and community well-being.
NEPA should exist and people who have bought their homes and lived in an area for decades should have a say on what their community should look like, but its gotten out of hand on how terrible America is at construction speed compared to Europe and East Asia. There should be a rebalancing, and that doesn't exist yet.
Excellent article!
Great Article!
Very informative. I had no idea how broad the scope of this industry was.
The industry seems to be moving towards for specialised chips as opposed to general purpose chips. Do you think this will increase the value added for chip design firms? Also doesn't India have a sizeable footprint in chip design?
#1 You are spot on about specialized chips. Nvidia makes GPU chips for supercomputers, LLM, and gaming. Qualcomm focuses on chips for wireless technology like 5G, vehicles, and laptops.
AMD does GPUs, TVs, and accessory electronics.
There's so much and there's still many Japanese, Korean, and Chinese firms I could mention as well
#2.
Depends what you mean by "sizeable"
Yes India has some chip design from Tata, Wipro, and Infosys. several domestic firms and many startups.
But in terms of fabrication or assembly & test exports, India footprint is slightly under $700M, an order of magnitude less than Thailand, Philippines, Vietnam, or Malaysia.
https://oec.world/en/profile/hs/integrated-circuits
Hence, why I still call India an "emerging player".
Depends on what your objective is. If your objective is military and strategic independence, then yes, your focus should on acquiring domestic packaging, assembly and fabrication capabilities. But if your goal is growth and productivity why not double down on your comparative advantage which is chip design which seems to be more value added.
As is usually the case, the chip design industry emerged in India not because of government but despite it. Because all chip design requires is a bunch of nerds in an office unlike fabrication which requires a lot of land and specialised infrastructure.
On a side note, what is your opinion on Risc-five? Given we're moving to a world of increasing specialisation, closed source software tools for development might prove to be a bottleneck. In such a world, Risc-five is a welcome addition. Do you see any other promising open source tools cropping up?
Risc-V is cool and seems better compared to the other open source stuff like OpenRISC or SPARC. I am not aware of any new ones past OpenRISC
What seems troubling is that Joe Biden want to limit China from using RISC-V
https://economictimes.indiatimes.com/tech/technology/us-lawmakers-press-joe-biden-on-plans-to-counter-chinese-use-of-open-chip-technology/articleshow/104911582.cms?from=mdr
I doubt they'll be successful. It's like him trying to stop Chinese civil servants from using Excel. I think if Risc-five takes off it create a cottage industry of chip designers in the middle and upper income countries.
Awesome read! Love the focus on a particular industry, very interesting.
How about water consumption for semiconductor production? Is there a reason for environmental concerns, or can it be dismissed as NIMBY-resistance?
Well all NIMBY resistance can be based on "reasonable" environmental concerns. Also people may not want to live near a chip fab.
1. Fabs use many chemicals and gases in their manufacturing processes, which could pose risks to residents if not managed well. These substances could release air and water pollution.
2. Fabs can be noisy which could disrupt nearby communities.
The issue in America is that NEPA is so powerful that many infrastructure, housing, and industrial production has slowed down considerably since the 1970s ever since the NEPA act has given communities the ability to raise legitimate concerns of construction related to environmental impact, public health and community well-being.
NEPA should exist and people who have bought their homes and lived in an area for decades should have a say on what their community should look like, but its gotten out of hand on how terrible America is at construction speed compared to Europe and East Asia. There should be a rebalancing, and that doesn't exist yet.