Oracle needed the super fast hardware because they were never any good at horizontal scaling distributed software architectures. But as you push the hardware speed boundaries the cost goes up exponentially. And trying to solve every problem with a relational database paints you into a corner sometimes.
8 replies (most recent on top)
From a business case perspective I think the mainframe is dead and I think in the not too soon future the case for spark will also be seen. If there's not a market to support the product why continue?
Yes you can keep some few happy for awhile but eventually they need to find another solution to their IT issues.
http://www.cringely.com/2016/06/15/mainframe-dead-long-live-mainframe/
Mainframe is not dead yet. Ask the banks, they are still running it. Like SPARC, it's a niche product, not for everyone.
Biggest thing is who wants the sparc? Especially if now Fujitsu dropped it. Whats the cost basis to develop a chip no one uses but Oracle in data centers fir cloud? Or just drop the development and stick with x86 which will also struggle against ARM in short order. Sun belongs in the museum with IBM mainframes. Yes it's great product but for the cost and few buyers. Oracle is better off staying out of hardware. Maybe it should merge with ibm or cisco or intel.
What locations of hardware?
It's very quiet from hardware end. We are demoralized and under utilized. I hope Oracle just announce something about hardware. Good or bad, we cannot keep going on like this.
Intel too
That's what I heard too - hardware to be turned upside down and many layoffs. Nobody knows what's going on. Look this website: Qualcomm, Cisco and NetApp have the most messages, guess what - that's all hardware