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!
Report: Nvidia building unit to design custom AI chips for cloud customers - SiliconANGLE
siliconangle.com
Nvidia is set to spend billions to become a chip design powerhouse. There is no doubt they will apply their vast and deep knowledge of software automation to this task and could easily become the chip design powerhouse in the industry by applying their own most advanced AI to this challenge. Nvidia has the money, experience, industry connections to dominate this market. Nvidia could easily attract the partners and customers to build an AI/automation powerhouse of unrivaled speed, size and productivity at lower prices due broad size and reach. It is very easy to see that Nvidia could build custom AI/ML systems to almost any customers needs. These factors combined with their reach in the industry could create an entirely new ecosystem to speed progress in the industry. Nvidia could become one of the largest industry disruptors in history. Any thoughts on this would be appreciated.
This is why Gunslinger wants to go after them... and he should. He has a foundry. Fire the empire builders, meeting goers, regurgitators, dead wood and hire based on knowledge and creativity. Nvidia is expanding... and hopefully will face massive competition.
The chances of Nvidia going to Intel are near zero. TSM has mastered fab and Intel has zero chance of catching up in people, process, money or ecosystem. When TSM's foundries in the US are running, they will become an even bigger powerhouse than they are now. TSM has already hit economies of scale in their ecosystem at all levels. I wish Intel the best, but the gap is getting bigger not smaller.
The wall of IP Nvidia has will make it very difficult to compete on design, let alone the experience and access to tooling and talent that TSM has built up for years. I hope the best for Intel, but they have many obstacles to overcome.
Why should Intel become successful now after years of decline, what has changed? Gelsinger is better than most, but Intel needs a great visionary who can compete and execute.
Why should Intel become successful now after years of decline, what has changed? Gelsinger is better than most, but Intel needs a great visionary who can compete and execute.
I'm not a Gelsinger fan. The only thing he's done since he got appointed that I agree with is the aggressive IFS strategy. And he needs to stop his strategy of "Speak loudly and carry a wet noodle."
siliconangle.com