Not only is the chiplet configuration able to perform, within reasonable limits, what the leading-edge superchips from Nvidia/TSMC can do, the chiplet configuration has proven in some instances to be more robust, as in more resistant to wear-and-tear. The reason for that is that the while the 3- or 5-nm super chip could process faster and more dynamically, it's truly a super chip in that just that one tiny piece is relied on to withstand the entire onslaught of processing load by itself, subjecting it to higher "abuse". The chiplet architecture, on the other hand, has several pieces of 7- or 9-nm put together so that the aggregate of these multiple chips handle the same processing load of the Nvidia/TSMC super chip. This means each chiplet, whilst handling only a single sub-function, has far bigger capacity to withstand an even bigger onslaught of processing load per chiplet for longer. Additionally, when a superchip crashes, the entire super chip needs replacing whereas with the chiplet architecture, only a single chiplet needs replacing when it breaks down, which is a lot cheaper as the chiplet is a lot more accessible at scale. The chiplet architecture isn't as dazzling and sexy as the super chip, but it more than adequately tides China over while the latter goes all out to develop their own super chip competitor which they're doing on different technological paradigm including, especially, ones that render obsolescent the entire lithographic tech currently monopolised by ASML.
Thanks for the informative supplementary insights! This adds another layer to the conversation about chiplet technology. Don't forget to subscribe our channel @alpha-uncovered
I cannot find online articles that support your claim that latest US naval ships were made in China. Seem like US Navy doesn’t even want to buy from S Korea nor Japan…
Thanks for for your support. We got clarification from Steven: "60 ships that are part of the U.S. Maritime Security Program and the Tanker Security Program, none was built in the U.S. and the last three tankers enrolled in the program were built in China. Furthermore, China is the only country that is capable of manufacturing all 18 major ship types within these programs, ranking first globally in 14 of them.”
There is nothing valud or important discussed here, inly sime generalities that aren't supprrable by specific wxamples if they ever decided to go that far. I'd caution prospective investors to look more closely at any supposed Chinese unnovarion and weigh its sugnificance before placing money. So for example in chipmaking the process is very complex. Any supposed innovation like chiplets isn't new and its basic concept what to put on a single die has seesawed back and forth throughout history.. And it by itself isn't likely to bust oast the limitations of sanctions. Photonic computing pathways today is still mostly an teresting idea but a very, very long way from implementation not helped by China's relative lack of laser technology. In all sciences and technologies China faces similar obstacles, combinations of lack of knowledge and access to technology owned and implemented by others. China has been firced to innovate in some areas like electromagnetism to compensate for lacking aerospace and fanjet engine technology. China might build a lot of warships but it's blikely none will still be sailing 30years from now like US naval ships and there are even questions how well they could perform today.
Not only is the chiplet configuration able to perform, within reasonable limits, what the leading-edge superchips from Nvidia/TSMC can do, the chiplet configuration has proven in some instances to be more robust, as in more resistant to wear-and-tear. The reason for that is that the while the 3- or 5-nm super chip could process faster and more dynamically, it's truly a super chip in that just that one tiny piece is relied on to withstand the entire onslaught of processing load by itself, subjecting it to higher "abuse". The chiplet architecture, on the other hand, has several pieces of 7- or 9-nm put together so that the aggregate of these multiple chips handle the same processing load of the Nvidia/TSMC super chip. This means each chiplet, whilst handling only a single sub-function, has far bigger capacity to withstand an even bigger onslaught of processing load per chiplet for longer. Additionally, when a superchip crashes, the entire super chip needs replacing whereas with the chiplet architecture, only a single chiplet needs replacing when it breaks down, which is a lot cheaper as the chiplet is a lot more accessible at scale. The chiplet architecture isn't as dazzling and sexy as the super chip, but it more than adequately tides China over while the latter goes all out to develop their own super chip competitor which they're doing on different technological paradigm including, especially, ones that render obsolescent the entire lithographic tech currently monopolised by ASML.
Thanks for the informative supplementary insights! This adds another layer to the conversation about chiplet technology. Don't forget to subscribe our channel @alpha-uncovered
@@alpha-uncovered Most welcomed.
I cannot find online articles that support your claim that latest US naval ships were made in China. Seem like US Navy doesn’t even want to buy from S Korea nor Japan…
Thanks for for your support. We got clarification from Steven:
"60 ships that are part of the U.S. Maritime Security Program and the Tanker Security Program, none was built in the U.S. and the last three tankers enrolled in the program were built in China. Furthermore, China is the only country that is capable of manufacturing all 18 major ship types within these programs, ranking first globally in 14 of them.”
necessity is the mother of invention
Very refreshing... Good talks
Thank you 🙂
This Economic World War is going to the scale of Armageddon!!!😂
Very interesting insights guys
Glad you enjoyed it
What's that apostrophe doing there?
There is nothing valud or important discussed here, inly sime generalities that aren't supprrable by specific wxamples if they ever decided to go that far. I'd caution prospective investors to look more closely at any supposed Chinese unnovarion and weigh its sugnificance before placing money.
So for example in chipmaking the process is very complex. Any supposed innovation like chiplets isn't new and its basic concept what to put on a single die has seesawed back and forth throughout history.. And it by itself isn't likely to bust oast the limitations of sanctions. Photonic computing pathways today is still mostly an teresting idea but a very, very long way from implementation not helped by China's relative lack of laser technology.
In all sciences and technologies China faces similar obstacles, combinations of lack of knowledge and access to technology owned and implemented by others.
China has been firced to innovate in some areas like electromagnetism to compensate for lacking aerospace and fanjet engine technology. China might build a lot of warships but it's blikely none will still be sailing 30years from now like US naval ships and there are even questions how well they could perform today.
Do you think Chinese fabs will make it to 5nm (or less) at scale for consumer electronics?