I've found some specs:<p>- up to 141 configurable processors<p>- new 22nm 8-core processor chip@5GHZ<p>- 110GIPS<p>- Single Instruction Multiple Data (SIMD) support<p>- On chip cryptographic and compression coprocessors<p>- up to 10 TB of memory (configured as 'memory raid', RAIM)<p>PDF is here: <a href="http://public.dhe.ibm.com/common/ssi/ecm/zs/en/zsd03035usen/ZSD03035USEN.PDF" rel="nofollow">http://public.dhe.ibm.com/common/ssi/ecm/zs/en/zsd03035usen/...</a>
While it's an undeniable technological achievement, I don't think it's a good strategical one.<p>Nowadays you don't create a single monolithic system, really expensive too, which is capable of processing all of your transactions, as it says, like 100 cybermondays every day.<p>Instead, you have a distributed system, with many cheap, geographically distributed servers, each one capable of a much lower number of transactions, but still quite high today...
and also you can spin up new servers as needed, or destroy servers as they are no longer needed, so you control your costs when you don't need as much instead of having a very expensive mainframe 90% underutilized 90% of the time.
That's a fairly content free release as these come. Nothing about the actual architecture, clearly aimed at managers rather than techies, IBM does enterprise sales very well. But of course I'm more interested in the nitty gritty details (though I'll likely never be in a position to pull the trigger on a purchase like this). Funny how Linux is mentioned in the press release, and other platforms don't even rate a mention.
I have been working in the mainframe industry for over 5 years now and i know tons of company (government, banks, insurance and others) are still running mainframe. I picked this field because i saw the opportunity of how alot of the old folks are retiring and companies are willing to pay alot of money to get young folks to join and learn the mainframe. Honestly its not even that hard as long as you are willing to learn....
what are your thoughts about it?
If someone was willing to train people, would people be interested in it?
The part about the in-memory analytics reminds me a lot of the SAP HANA spiel:<p><pre><code> Transaction Database --> copy --> analytics db = slow
In-memory transactions with built-in analytics = fast
</code></pre>
I totally understand the value of analytics and BI applications, but does it all have to be realtime? And what "mobile analytics" are they going to compute exactly? Forget analytics, I can <i>tell</i> you what mobile users are doing right now---they're all playing candy crush.
It's fascinating to see the rhetoric focussing on mobile transactions being the main driver of demand. While it's certainly huge, there are still a lot of other uses.
> z13 is the first system to make practical real-time encryption of all mobile transactions at any scale<p>Or you know, any kind of transaction because there is no difference.<p>Fuck I hate marketing.
The "encryption" buzzword is worthless without details. And I wonder if they'll stand behind their product if their crypto flavor of choice is broken and needs to be changed. Otherwise that's an expensive brick of swiss cheese.