3 ms·
I find that claim pretty hard to believe tbh. Any actual benches published? It’s like Oracle claiming they are the shit but then actually not allowing any perfo
by dilyevsky 4y ago
I find that claim pretty hard to believe tbh. Any actual benches published? It’s like Oracle claiming they are the shit but then actually not allowing any performance benching whatsoever. The 14nm process that they use to build those kinda leads me to believe it’s all marketing
- kitd 4y agoWhy should it be so hard to believe? A mainframe box is about the size of a rack of servers, and all the components are much more tightly integrated. Plus the utilisation on m/fs is much higher than on normal servers. I don't have benchmarks, but googling "mainframe energy efficiency" brings up a load of results, eg: https://www.bmc.com/blogs/mainframe-sustainability-green-it/ https://www.bmc.com/blogs/mainframe-sustainability-green-it/ https://www.serverwatch.com/guides/for-energy-savings-consider-a-mainframe/ https://www.serverwatch.com/guides/for-energy-savings-consid... https://www.suse.com/c/mainframe-versus-server-farm-comparison/ https://www.suse.com/c/mainframe-versus-server-farm-comparis...
- dilyevsky 4y agoThese all seems like pure marketing. I would love to see actual technical benchmark (this would also make prime HN material even today). Usually companies that actually think they have a performance edge publish those ;)