IBM z14 Announced, Support Added To LLVM Clang

[Dawn]

1) I doubt that. All articles I have read say everybody flees off from IBM Mainframes. Show us links there are new customers.

2) I read that IBM sells 100-200 Mainframes each year, that was an article in www.theregister.co.uk who covers Mainframes sometimes. I stored that number on my memory, thats it. I will google for the article again and see if I find it.

3) That is a lie. We all know that performance of Mainframe cpus are abysmal. If the Mainframe cpus were super fast - why hide it? Why not post benchmarks? "IBM does not post benchmarks that prove Mainframe cpus are superfast, because IBM does not want to hurt the feelings of competitors?". Or what? What do you think exactly when you keep defending Mainframe cpu performance? What are IBMs motives to not post benchmarks where Mainframes is 100x faster than Intel Xeon? If they were that fast, everyone would rush to Mainframe cpus. Does not IBM want new customers?

4) You are toast now. I have worked in one of the largest exchanges in the world, for many years. We know everything about the competitors. And there is NO exchange running on Mainframes. Large exchanges earns all money on trade volume. And you get volume if you are fast. Mainframes are dog slow, so no exchange would get traffic using Mainframes. And that exchange would go bankrupt. You are toast - I know the largest Exchanges very well. I want you to post links to an exchange that runs on Mainframes. You can not. This proves you are deliberately lying. So, "Dawn" is a liar and FUDer and can not be trusted. He is a paid IBM liar.

Oh, boy. Another chance to engage with Oracle lies about something I don't actually give much of a shit about.

1) http://www.zdnet.com/article/ibm-lau...ta-encryption/ "the company brings on anywhere from 5 to 15 new z customers a quarter"

2) That number is, again, way low.

3) No. "we all" don't know that. You talk yourself into believing it because you're fanatically pro-Oracle, and have been posting long streams of pro-Oracle copypasta for the last ten years. IBM, because they're IBM, is godawful at marketing; customers who ask can see z performance data to their hearts' content. I have already seen a document with z14 benchmarks, and per-core performance is not low. I've used a z13 myself, and can say the same about that one from direct personal experience.

4) Nonetheless, I know of an exchange that does, from a friend that works there. I'm a "she", by the way, and if I was paid by IBM I would be saying far nicer things about them than "z is insanely overpriced" and "IBM is godawful at marketing." I have no real fondness for IBM - I think they're incompetent a large percentage of the time, and overpriced a large percentage of the rest, but technical genius has never been a problem there.

[bws111]

GE locomotives are really underpowered, way less power than a Tesla. I know this because I read a couple of articles: one by a guy who built a model train, and one by a guy who rode a train once. They are, of course, now experts in all things train. Anyway, they both said that the power of a GE locomotive is low. As further proof of how weak they are, GE won't even publish the 0-60 times or quarter mile times for the things. Instead they make up their own benchmarks, like traction force and fuel used per ton-mile. Why don't they use the same benchmarks as everyone else uses for cars?

See how dumb that sounds?

Writing an emulator and porting an OS in no way qualifies anyone to speak on the performance of real mainframes. Contrary to your statements, they are not experts.

Benchmarks are only useful IF they reflect the way a user actually would use a thing. Nobody uses mainframes for CPU-bound stuff, not because the mainframe is incapable of doing it, but because it makes no economic sense. Why would you pay for that tremendous I/O system only to have it sit idle because you are running CPU-bound workload, no matter how fast the processors are? On the other hand, it would be useless to have that I/O system if the processors were slow and unable to keep it busy.

It is funny that you would mention religion, because YOU are clearly the one who is willing to beleive things without any proof. You are perfectly willing to believe 'someone who ported Linux' as an expert in performance, even though you have absolutely NO idea what they claim to have benchmarked. Why?

The fact is, there are NEW mainframe users who use mainframes to run thousands of instances of Linux. Places don't use mainframes because they are 'stuck', they use them because they are often the best tool for the job.

And your supposed sales figures are off by at least a factor of 10.

In short, you appear to know absolutely nothing about mainframes.

[kebabbert]

Oh, boy. Another chance to engage with Oracle lies about something I don't actually give much of a shit about.

1) http://www.zdnet.com/article/ibm-lau...ta-encryption/ "the company brings on anywhere from 5 to 15 new z customers a quarter"

2) That number is, again, way low.

3) No. "we all" don't know that. You talk yourself into believing it because you're fanatically pro-Oracle, and have been posting long streams of pro-Oracle copypasta for the last ten years. IBM, because they're IBM, is godawful at marketing; customers who ask can see z performance data to their hearts' content. I have already seen a document with z14 benchmarks, and per-core performance is not low. I've used a z13 myself, and can say the same about that one from direct personal experience.

4) Nonetheless, I know of an exchange that does, from a friend that works there. I'm a "she", by the way, and if I was paid by IBM I would be saying far nicer things about them than "z is insanely overpriced" and "IBM is godawful at marketing." I have no real fondness for IBM - I think they're incompetent a large percentage of the time, and overpriced a large percentage of the rest, but technical genius has never been a problem there.

1) Ok, I see that. I really doubt that number as it comes from IBM who is known to inflate numbers. On the other hand, there are many customer stories leaving Mainframes. The net effect is that IBM is loosing Mainframes customers big time. x86 is much faster for a fraction of the price.

2) Might be, but I read it on theregister. And I stored that number into memory, because I was shocked about the low number. I read that article again to double check I did not read wrong.

3) Ok, so there are many people saying that Mainframe cpus are dog slow, and also, there is no proof of their superiority. No benhcmarks. Why is that? IBM does not want to hurt their competitors? IBM does not want Mainframe market to increase, by posting benchmarks where Mainframes demolish all the competition?

4) That is bullshit. I work in that business. You are lying. Or ignorant. Let me educate you. Look, Mainframes are used a lot in banking, but never in finance. Banking is about "increase all these account salaries with 5% and update the values, and send out notifications". That is pure administration and boring. But that forms the back bone of much of the money flow in the world. Finance is about investing, not accounting. Finance is about high frequency trading, ultra low latency, machine learning, FPGA, etc. Finance is trying to get an edge and earn money by trading or investing or scanning company reports for undervalued companies etc. Investing is different from banking.

As you dont even know the difference, I promise you that your "friend's" mainframe is not driving an exchange. Either you have no clue what your friend does, or you are deliberately lying and FUDing. As you are claiming that Mainframes have fast cpus, which is a lie, probably you are deliberately FUDing and lying.

[kebabbert]

GE locomotives are really underpowered, way less power than a Tesla. I know this because I read a couple of articles: one by a guy who built a model train, and one by a guy who rode a train once. They are, of course, now experts in all things train. Anyway, they both said that the power of a GE locomotive is low. As further proof of how weak they are, GE won't even publish the 0-60 times or quarter mile times for the things. Instead they make up their own benchmarks, like traction force and fuel used per ton-mile. Why don't they use the same benchmarks as everyone else uses for cars?

See how dumb that sounds?

Yes that is dumb. That is a really dumb analogy that is not valid.

For Mainframes and cpus, there are well known benchmarks that all cpu vendors use to bench cpus against other cpus to be able to assess the performance. For instance SPEC2006 benchmarks. Why is IBM not using any of all the standard benchmarks that are specifically designed to compare cpus?

Writing an emulator and porting an OS in no way qualifies anyone to speak on the performance of real mainframes. Contrary to your statements, they are not experts.

I suggest you go and learn something about IT and development. You are totally lost. So you believe that it is easy to write an emulator? BTW, the guy who ported Linux to Mainframes and could compare performance for x86 and Mainframes - is a Mainframe optimizing expert. That is why he was tasked with porting Linux to Mainframes - because he is an Mainframe optimizing expert.

This could be the dumbest thing I heard in a long while.

It is funny that you would mention religion, because YOU are clearly the one who is willing to beleive things without any proof. You are perfectly willing to believe 'someone who ported Linux' as an expert in performance, even though you have absolutely NO idea what they claim to have benchmarked. Why?

He is an well known Mainframe optimizing expert. There are several experts saying that Mainframes are slow. And there are no proof that Mainframes are fast. No benchmarks. Who are religous? All you who believe anything IBM says? I demand proof. Any benchmark. Just a small one.

The fact is, there are NEW mainframe users who use mainframes to run thousands of instances of Linux. Places don't use mainframes because they are 'stuck', they use them because they are often the best tool for the job.

They use Mainframes because they are stuck, it is called "lock-in effect". This is partly why everybody use Oracle database - because it is a big undertaking to port all Oracle SQL code to another database. That is part of the reason they continue use Oracle database even though it is very expensive. The other reason is because Oracle database is the only choice for extremely high workloads. No other database can match Oracle in that area. Anyway, lock in effects is a big problem and that is why people continue to use a piece of hardware/software. To port all billions of codes of Cobol to Java would take many many years. There are many billions of COBOL code running on Mainframes. This is a fact.

You know nothing about development, which you proved when you claimed that writing an emulator is an easy task. I have written emulators, and emulators require intimate understanding of the cpu and all weird stuff that might happen. Have you ever written an emulator? Which one? How did you do it? Tell me about the architecture

And your supposed sales figures are off by at least a factor of 10.

Say that to www.theregister.co.uk

In short, you appear to know absolutely nothing about mainframes.

I suggest you dont talk about things you dont know nothing about. That just makes you look stupid.