I'm curious to see what the total timings would be like to get the data in a useable form - as opposed to just fetching a record from a data store. As noted - these data stores just store and retrieve data and don't do things like joins or ordering, etc.
Could there be a comparison between these datastores and the traditional ACID compliant databases when it comes to retrieving actual data in a useful format? E.g. perhaps doing a join or an ordering of some sort? I don't expect databases (e.g. Oracle, MS SQL Server, DB2) to be faster in raw performance, but I do expect them to be faster in terms of total development time and bug fixing since the application developer wouldn't have to do the locking, page pinning/unpinning, etc. manually.
I think if you look closely at the specs, I think the Xeon has about 2.5MB of L3 per core, while P8 has about 8MB of L3/core. There are bigger differences than this. But the examples above have both processors set at SMT2 (~30 seconds vs ~10 seconds). Differences are greater for SMT4 and SMT8. Of course, if your code fits nicely into the L3 - e.g. if your hot code consumes 3MB, then Power8 will be a bigger winner since there may be more thrashing for the x86 chip - unless, of course, the prediction routines is straight forward and the cacheline can be prefetched (think sequential access vs random access).
From what I saw elsewhere, highly parallel code runs faster on P8, where as single thread perf is faster on x86. So if your HPC app is basically single threaded on one of your compute nodes, then that would be faster. But if your HPC code is highly multithreaded on the core, then P8 may surprise you.
It depends on what you mean by "demanding". I think that single threaded applications on x86 (with nothing else running) will smoke the Power8 system. But if you have lots of threads and high levels of concurrency going on, i think that the results will be very much workload dependent.
it's surprising how much cheaper these Power8 boxes are compared to the Power7 boxes. You can get a tyan 2U, 1 socket system for $2800 FOB HK: http://www.tyan.com/campaign/openpower/
Bah. There are so many apps that I purchased that are complete junk, that, IMO, paying a $0.99 for an app that I do use seems to even itself out. And what about completely fraudulent apps (e.g. search for "blacklist" in the App Store). How are they even getting in?
Given the huge amount of crap in the App store, the only real way of finding things that are decent is to go to specialized aggregators. I go to BoardGameGeek in order to find reviews of games. The detailed reviews there mean more to me than the huge stinking pile of games in the App Store. There needs to be a better way of sifting through the stuff in the App Store than using the App Store Application itself.
This is funny. I thought that the whole paradigm of startups is to first release the minimally useful featureset and work on the other items later? I mean - isn't the goal is to first get revenue and then fix up what needs working later?
Could there be a comparison between these datastores and the traditional ACID compliant databases when it comes to retrieving actual data in a useful format? E.g. perhaps doing a join or an ordering of some sort? I don't expect databases (e.g. Oracle, MS SQL Server, DB2) to be faster in raw performance, but I do expect them to be faster in terms of total development time and bug fixing since the application developer wouldn't have to do the locking, page pinning/unpinning, etc. manually.