back to article IBM warns mainframe shops of 2013 price hike

When your customers pay millions of dollars for your machines and then spend many tens to hundreds of thousands of dollars per month – and sometimes much more – on monthly license fees for your software, you can't just spring a price increase on them one day. You have to give fair warning so they can budget for the price …

COMMENTS

This topic is closed for new posts.
Silver badge

Meaningful comparison metrics anyone ?

MIPs is not objective, it depends on the instruction set. Something like the SPEC benchmark is what we want, but somehow IBM does not seem to publish the results for mainframes:

http://www.spec.org/cgi-bin/osgresults

Will someone with one of these beasts run the benchmark and let us know, please.

2
0

Re: Meaningful comparison metrics anyone ?

I would beg to disagree. The zPCR benchmarks used by IBM to determine MIPS are reflective of what customers use mainframes for. zPCR provides comparisons of the different IBM mainframe servers, for which MIPS is a valid comparison. To quote SPEC, "SPEC designed CPU2006 to provide a comparative measure of compute-intensive performance across the widest practical range of hardware using workloads developed from real user applications." While the current mainframes would probably perform well (the zEC12 is a 5.5GHz processor) how they perform doing compute intensive work is not relevant to mainframe customers workloads. The SPEC Java benchmarks may be more relevant, and it would be interesting to see how WebSphere 8.5 Liberty Profile on z/OS or Linux on System z performs there.

3
1
Anonymous Coward

Re: Meaningful comparison metrics anyone ?

It's all about throughput and having extensive IO.

These are batch processing machines at the end of the day.

1
0
Anonymous Coward

Re: Meaningful comparison metrics anyone ?

Also a high performance version of a mainframe is generally referred to as a supercomputer.

0
0
Silver badge
Devil

Re: Meaningful comparison metrics anyone ?

> high performance version of a mainframe is generally referred to as a supercomputer

LOLNO

Do you really want to run the <strikethrough>state extortion</strikethrough>tax dunning letter batches on a "supercomputer"? No. You may want to run them on a "high performance version of a mainframe" because you will be leaving at 15:30 on friday afternoon though.

1
0
Anonymous Coward

Re: Meaningful comparison metrics anyone ?

Agree, there is no meaningful way to compare System z to something like x86. System z runs at 100%, or very near, utilization with extremely high I/O. It uses a variety of specialty and off-load procs to get there. x86 runs at maybe 30% with little I/O. The theoretical benchmark numbers are going to be realistic for z and absolutely wacky for x86.

It would also ignore the reasons people use z, like EAL5 grade security. It would be like comparing an armored truck vs. motorcycle based upon maximum speed. If you are using an armored truck, speed is pretty far down the list of importance.

1
1
Anonymous Coward

Re: Meaningful comparison metrics anyone ? (@AC November 2012 17:01 GMT)

That's a load of BS right there. Mainframes have their uses, but there is no magic - as always it depends on particular use cases and there are many "buts" and "ifs".

You are gravely mistaken if you think that x86 tops out at 30% with little IO. Such a system would be wholly unsuitable for the tasks it is actually used for (in the real world) - supercomputing, rendering farms etc., usually requiring very fast and heavily used IO connectivity. Speaking of which - what do you mean by "extremely high I/O"? Hundred 8Gb FICON ports going flat-out? Or more sedate 8 or 16, which is far more usual? Because your home PC with SLI does almost the same IO as 8 8Gb FICON ports combined when you play a game. And if you are concerned with blocking IO, then you can use 160 threads on a budget with 8-socket x86.

That is not to say that x86 is on par with mainframe (it isn't), but that the comparison is not so straightforward performance-wise when smaller or midsized mainframes are concerned.

As for EAL5 - if you use zVM (which I suppose quite a lot of people do), you have EAL4+ only. The same as Windows or Red Hat Linux on x86.

One of two real reasons to use mainframe over x86 (which sadly you have completely failed to mention) is RAS. With regard to RAS, mainframe is far, far ahead of machines built on RISC/x86.

Majority of mainframe use cases however have to do with existing software, which is the second reason. Porting applications from mainframe is usually such a great undertaking that it is often seen as not worth the risk.

1
0
Silver badge

Re: Meaningful comparison metrics anyone ?

and reliability+uptime over everything else.

1
0

The FWLC products almost all pre-date 2001 and very few of these products are still being marketed by IBM. Most customers are on current hardware will not be affected by these changes, or only minimally.

0
0
Silver badge
Pirate

Squeeze that goose!

IBM need to extract as many golden eggs from the mainframe base whilst the still can. Bend over and hold your ankles, mainframes, IBM needs to fund another shares buy back.

2
3
Anonymous Coward

Re: Squeeze that goose!

Whatever, System z is still the gold standard.... I am sure any users that HP tricked into moving to an Itanic platforms wishes they were still on z. A 10% uplift on certain software tools is a trifling sum as compared to the capital and operational cost of migrating off of Itanic platforms.

2
1
Anonymous Coward

Re: Squeeze that goose!

Depends. If they moved from z to VMS on Itanium, then I wholeheartedly agree. They are in a world of pain. But if they migrated to HPUX and some standardized platform (java, oracle/db2), it may not be such a big deal.

1
0

Mainframe

Use of IBM 'mainframe' systems is till in the realm of -keeping the company business going no matter what- with the best ROI.

Quite boring to a large percentage of computing device users across the globe, but keeps IBM in positive revenue streams yearly.

1
0

MSUs = Millions of Service Units

- no meters involved.

1 MSU <> 7 MIPS depending on processor type

0
0
This topic is closed for new posts.

Forums