Flops and GigaFlops
By RickRamsey-Oracle on Jun 28, 2010
This blog is about the benefits of running Solaris on SPARC.
Although I was born in Santiago, Chile, I became a teenager in Lima, Peru in time to watch the beautiful Peruvian National Team advance all the way to the quarter finals of the 1970 World Cup. For a small country like Peru, this was a gigantic achievement.
If you've watched any World Cup matches lately, you've probably noticed two things:
- The best players get fouled incessantly.
- Everybody flops.
I'm not a fan of either. (And don't even get me started on the NBA.)
In Peru we believed you were a good defender only if you could strip the ball away without touching your opponent. A foul might save a goal, but it wouldn't save your honor. As a result of this appreciation for technique, the1970 team won the FIFA Fair Play Trophy. With flopping and strategic fouling so prevalent in the World Cup, it's a miracle we made it to the quarter finals!
By now you've probably heard how all the products in the Oracle stack have been optimized for the best possible scalability, security, and reliability. This is particularly true for Solaris, our Virtual Machine, and Sun SPARC enterprise servers.
This white paper takes the discussion one step further.
Written by Mike Mulkey with the help from Solaris engineers, this paper discusses the benefits of Oracle Solaris running on Sun SPARC Enterprise M-Series and T-Series servers, but drills down on the specific optimizations for reliability, scalability, security, and virtualization. It describes the superior results of taking a comprehensive, integrated architectural approach to designing the operating system with the hardware, such as:
- Solaris multi-threading capabilities, when combined with the SPARC multi-core chips, provide the capacity to run 64 threads per chip, which can make your data center more flexible, quicker, and far more reliable.
- Solaris running on the SPARC64 VII chips of Sun M-Series servers provide mainframe-class performance and reliability at lower cost. Not to mention vertical scalability. Plus hot-swapability for major components.
- Predictive self-healing in Solaris works with the highly reliable memory subsystems of the SPARC Enterprise server to stop faults from bringing down your system.
- In addition to its scalability and reliability features, ZFS lets you add, change or remove storage devices on the run.
After you finish reading Mulkey's white paper you might come to the same conclusion as I did:
The combination of Solaris and SPARC Enterprise servers lands squarely in the world of American Football, a game that leaves little room for flopping, whining, or making excuses.
(Read this forum discussion if you want to find out how many floating point operations a SPARC chip supports).