Purchasing servers and other hardware for local use is not necessarily the optimal solution for many companies. The cost often exceeds the need for those who only periodically need a lot of computing power or deliver services with varying loads. Here, data center giants such as Microsoft, Amazon and Google have found their niche where virtual machines can be rented out when used.
Intel has long been the mastermind of processors in the server industry, but AMD’s success on the consumer side also spills over into data centers with the Epyc family. It is now clear that AMD is squeezing additional market shares, and will be the power behind Google new range with new virtual machines called Tau.
This spring, AMD released the third generation Epyc “Milan” with the Zen 3 architecture under the hood. Google intends to take advantage of Tau, which according to the company offers the best price-performance ratio in scalable applications. This feature sets Google against ARM alternatives and explains that Tau is 42 percent sharper. Compared to the common “Cascade Lake” family from Intel, Google promises 89 percent better price-performance ratio.
Epyc usage shows that AMD is seriously in the heat again, though in conversation with Zdnet Google’s Sachin Gupta explains that this can change quickly. There is a clear race going on in the processor industry right now, and Google is not closing any doors to other manufacturers.
If an answer in the future is we need to have an Intel-based, or an ARM-based instance in the Tau family, whatever it is that’s best for our customers for scale-out, we would look to build against that and optimize for their application need. – Sachin Gupta, Google
Google is open to both ARM and x86 processors, if it is beneficial in relation to customer needs. Gupta further explains that it is traditionally an obstacle to having to convert the applications between instruction sets, but that they are willing to work for the possibility of having both Intel and ARM instances in the Tau family.
Read more about server-oriented hardware:
Source: SweClockers by www.sweclockers.com.
*The article has been translated based on the content of SweClockers by www.sweclockers.com. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!
*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.
*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!