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!
The latest development in more efficient computing by Amazon might give it such an edge it might dominate computing even more with AWS than it already has. It looks like this latest development may give it even more of an edge in computing itself as it lowers the cost to such an extent that it doesn't pay to do a significant portion of computing on your own platform. Amazon already has massive financial, data center, technical capacity and is now looking like the master of strategically adapting new technologies to become even more dominant. Is AWS trying to be the best economic and business answer to computing in general at the lowest cost? Could this be the beginning of a major new push and expansion of AWS? Will AWS become a major platform to build businesses on as the Amazon market place has? Could this be the final fundamental change in converting computing to a utility? Will AWS be able to scale to the point they can develop and achieve for all intents and purposes a monopoly in many areas of computing? If Amazon makes this move, it will give them the ability to deliver AI/ML as a mass market service, changing and leading the game over everyone else. It could even make their product the world dominant standard with the ability to add scale at a faster scale with more versatility than anyone else. This could covert AI/ML from a specialty product to a household everyday product with Amazon being the standard. Any thoughts and comments on Amazon expanding their dominance or alternative views would be appreciated.
The is SO much money to be made in the cloud and how do you competitively differentiate one offering from another? You make your own chips that's how. Just ask Apple.
The is SO much money to be made in the cloud and how do you competitively differentiate one offering from another? You make your own chips that's how. Just ask Apple.
The is SO much money to be made in the cloud and how do you competitively differentiate one offering from another? You make your own chips that's how. Just ask Apple.
The recent purchase of Mellanox by Nvidia is based on moving computing from the corporations to ultrafast data centers just indicates how far along this trend is and the importance it is viewed with.
The recent purchase of Mellanox by Nvidia is based on moving computing from the corporations to ultrafast data centers just indicates how far along this trend is and the importance it is viewed with.
It does make sense on paper, Nvidia is more diverse and they have collaborated on supercomputers. Thing is nvidia is a smaller company, 11528 according to Wikipedia buying mellanox is quite the demographic shift.
The is SO much money to be made in the cloud and how do you competitively differentiate one offering from another? You make your own chips that's how. Just ask Apple.
Companies like Microsoft, Apple, Google, Amazon, Facebook, Huawei, and Tesla all have teams to design their own special-purpose chips. But on the other hand Oracle dropped their in-house chip design effort. To me it's inconsistent to Oracle Chairman Larry Ellison's style. Do you know the reason behind it?
There are so many built in secrets in cloud infrastructure that a chip alone does not change much.
Amazon already runs customized versions of all software it runs, virtualization already uses a custom silicon to avoid the overheads and they even get their own SKUs from Intel and AMD.
