Last year it was end of May. Will there be any this year?
64 replies (most recent on top)
@8uir+1beUeseK
How many positions do you have at AMD?
And how many people can you hire totally?
For those reading this topic in Power, AMD is hiring engineers with system and chip skills in Austin, apply before the jobs are filled
The idea that an application built on OCP/K8s can be deployed / re-deployed to Z/Power/x86 on-prem or to AWS/Azure/GCP/IBM Cloud simply and easily without making any changes is ridiculous to anyone who has ever deployed an application. This is the late 90s Java "write once run anywhere" nonsense all over again.
Power declining is not new - it was already having problems meeting shipping targets since at least early 200x and has been on a competitive decline since. This was the main reason why Apple switched from Power to Intel in 2005.
I don't understand why anybody still has the illusion on the survivability of Power architecture. The installation base of Power ISA is less than 1%. It takes time and resource commitment to revive it. Apple succeeded in the late 1990s. Do you see ibm makes any resource commitment?
On a separate note, who in the right mind will buy a hardware business with less than 1% of installation base? If the BU is sold, it will either go to a bottom feeder, or one of the WITCH. Working for either is as unpleasant as working for ibm.
I believe you are not looking at the underlying capabilities of each platforms abilities. Yes openshift and kuberneties work on Intel/LINUX platforms (commodity as you call it) but they also work on Power and mainframe platforms with a lot more functions. This is just the past repeating itself Unix was viewed as the great equalizer 20 years ago, but it turned out once you added in the proprietary extensions on unix, every flavor of unix was different (each providing an advantage vs vanilla unix). Openshift (call it enterprise) and Kurbeneties (call it vanilla) are two different solutions depending on the platform. Google likes to play with Kurbeneties/ vanilla platform where IBM likes to play with “enterprise openshift” Each has its advantages and each has its lifecycle. Intel life cycles tend to run 2-3 years where as Mainframe tends to run 10-15 years. If you are building proprietary extensions that will last 10-15 years, that is a competitive advantage.
Addressing your IBM abandons everything statement. YEP that’s true, because time marches on. I will comment that IBM got played by Globalfoundries in that GF promised a lot of things for IBM’s patents and 1.5 billion, and then walked away. Shame on IBM for not doing their due diligence, BUT that story is not done yet. Samsung has back filled after 2-3 years, but a pile of damage was done to IBM’s underlying HW plan. My guess is systems will never fully recover, and IBM will have to embrace RedHat even tighter, NOT abandon it
What does IBM do best? Answer: abandon stuff…
This IBM company can’t compete, can’t fix anything, the only thing it knows how to do is to abandon stuff.
IBM will end up a mainframe company only. The reason is because the mainframe is the only place where they have a monopoly and no competition.
OpenShift is already a commodity that IBM wants the customers to pay for when they (customers) can get Kubernetes for free. Not to mention that a lot of customers see no point to run everything in containers.
Bottom line, Red Hat will be sold off within 5 years at a huge loss.
4ncg Now you are catching on IBM wants out of the HW business That means OS/400 and AIX also go. The exit will take some time as IBM will help their top 20-30 major ISV’s (SAP, Oracle, Infor, etc etc)move to LINUX. IBM’s plan is for Power to focus on 980’s, Z & Storage chip sets, and AI optimization. IBM labs will pick up most of the design and futures. Everything else will go. Kyndryl will pick up channel sales and TSS. Adieu Power adieu
I predict based on observation in the trenches that if p11 remains most of the design will move to pok and boe with less dependency on Austin, would also explain why ibm pulled out of leasing domain tower 2 for power just prior to covid on direction of Ginny. There is a big plan that is not shared but it makes sense
IBM HW wise only wants the very very large customer installs. That is Mainframe, DSXXXX, and Power 980’s. Everything else can and will go. There is an exception to that, and it’s niche Storage SW and flash. Remember IBM gets 20% of their Power sales from the storage sector. I could see POK becoming just an “enterprise” solution plant. Austin will become all things LINUX. Rochester will need to carve out a niche or be closed. IBM is morphing back to what it was in the late 70’s to early 80’s A mainframe company with some “niche” products.
Storage can be broken up into 4 categories: Mainframe, distributed HW, Storage software and cloud. And within storage you have disk/flash and tape.
Mainframe refresh cycle drags mainframe disk & tape
Distributed storage hardware and software competes against many, many competitors, that have a larger market share. I believe this is the main area IBM storage struggles. IBM is not agile enough, marketing has been lousy for many years, they are often late to market, have to sell at deep discount, have complex pricing, are slow to execute, and cannot differentiate enough from the competition. It has turned into a commodity market and IBM continues to lose money quarter-over-quarter.
I believe it is too late for IBM to turn this around.
Tape is somewhat unique in that IBM is the sole manufacturer of LTO - resold by Spectra Logic and Quantum (among others) who compete quite successfully against IBM in various markets.
IBM also sells virtual tape into the mainframe space and have very few competitors.
We all know their struggle with cloud. The push for hybrid cloud solutions also drags the same issues as above, and is thus in my opinion a flawed approach.
This discuss really needs to be broken into three groups when talking Systems
MF, Power and Storage
MF is fine, its not going anywhere, its also based up in Pok and Boe, they are fine and Pok has little competition to take talent away, technically the have issues where most of the unique MF skilled engineeers/programmers are at or near retirement age.
Storage, no idea, not as familiar with that area so won't comment on what I am not very familiar with
Power is in big trouble. Based primarily out of Aus with some in Roch and a few in Pok. The leaders of Power were pushed out in a power struggle and are now at Nvidia and AMD and pulling over talent for their growing businesses. Other Austin companies are also pulling talent away. Why ? Because of lack of direction, lack of leadership, shrinking revenue, treating eng/prog as commodities, low pay compared to the Austin Market... Power is now at the tipping point where it can not sustain because of the talent loss not to mention Z has priority over P for good reasons but that causes P to lose resources to Z when in need. P is also used to pipeclean for Z chips
Anyway, back to this thread, I would like to see other responses based on the three groups and what others see as far as each's future at ibm
"Having a hardware solution is not exclusive to having a cloud solution.
Why don't you guys get this? "
^^^ This... to all the silly small platform fan bois... LOL!
In the end... it's still all one big computer mess.
The platform doesn't really matter, especially if you think in computing and computer science terms. Just different implementations of "the wheel".
" This is again cr-p. IBM marketing BS.
How much can you think it generates revenue? 4-6 billion max? IBM results."
Is 4-6 billion something to throw away?
Agreed... marketing is, by definition, BS.
I work on many of the accounts that use MF... I can confirm... the financial backbone of the world is still largely mainframe.
"Now compare it with any cloud or datacenter companies. AWS alone has more than 30 billion+ quarterly revenue. Haven't even compared with MS, Google, datastax, Apple."
Not a correct comparison... IBM is not in the consumer market.
"How much do you think world depends on it? World's way bigger than IBM mainframes. "
Yes it is... but the backbone is the backbone... rest of the body is in bad shape without it...
I agree they can't abandon MF. But they can move it to Kyndrl. Not only MF, move cognos, DB2, liberty, every other old hardware/software junk to kyndrl. Focus on a new IBM, consisting of open source products, most used Redhat tools, again jump into public cloud with hybrid option open. Improve Watson services.
Director again after few days. Happy to see thread is still active and many people are supporting my view.
It's like IBM is divided same as US election, supporting Biden (cloud) vs Trump (MF).
Having a hardware solution is not exclusive to having a cloud solution.
Why don't you guys get this?
If the concern is image then ibm can split off separate divisions like Alphabet. Ohh wait! It's already a separate division.
Infact all companies including F,G,MS,Amzn etc are developing their own hardware to vertically integrte their cloud. (Ibm already have this)
MF is self sustaining unit, it can't give ibm the growth but can be a helpful platform because IBM HAS NO OTHER PLATFORM.
How much revenue does IBMs public cloud generate? Ibm bought softlayer for that and it has completely failed.
All the "hybrid " and "AI" is basically some software running on cloud or other platform which Amzn,G,MS will give for free to their cloud customers until they run down the competition.
I agree that ibm shouldn't have abandoned it's public cloud, but because they already did they can't afford to abandon MF as well.
"Because of the percentage of the world's financial backbone that runs on it and the revenue it generates? Not to mention it's reliability?"
This is again cr-p. IBM marketing BS.
How much can you think it generates revenue? 4-6 billion max? IBM results.
Now compare it with any cloud or datacenter companies. AWS alone has more than 30 billion+ quarterly revenue. Haven't even compared with MS, Google, datastax, Apple.
How much do you think world depends on it? World's way bigger than IBM mainframes.
This is an interesting discussion. Can someone ask Arvind in AMA?
What do you want to focus MF vs Cloud? In case you have to pick one.
"But why aren't you accepting MF is no more valuable to IBM also? "
Because of the percentage of the world's financial backbone that runs on it and the revenue it generates? Not to mention it's reliability?
"BUT mainframes continue to run the world, and their underlying code is pretty rock solid. "
The pile of APARs and PTFs in RETAIN would argue that the code base is still buggy. No different than any other code stack. It just has a 20-30 year headstart.
What's the saying, "Any program with more than 3 lines of codes, has a bug"? LOL!
Dude this is total bu-----t. I know systems People are more vulnerable and they visit this site too often. It can be seen from this most active thread.
But appointing a CEO from cloud is totally BS. Anyone deserving can be selected. Arvind had his vision of hybrid cloud which IBM has bet on. That's the reason he is appointed. Tom is SVP, not VP. He also seems to be happy in his role. So cut the cr-p and go back to your school.
Also, MF or Cloud debate leave on execs. It's above our pay grade.
I agree with you all. IBM cloud and AI is bad. But why aren't you accepting MF is no more valuable to IBM also?
Just looks at IBM.com, there's no mention of mainframe on main page. They also bet on Redhat for hybrid cloud concept. Currently, Redhat and it's open-source tools are more important to IBM than traditional mainframe.
Our CEO is Arvind (who is from cloud & cognitive), where as Tom (systems VP) is very knowledgeable and deserving, but he couldn't make it.
I am an intern, still I know some of internal politics. But you experienced guys are trying not to accept it or very d-mb to understand what's going on.
Dude you don't know how bad ibm "cloud" and "ai" is.
Ibm is a MF only company.
Do i like it? No.
Would i be stupid to just sell it off? No way
Why? These are platform wars, even G is losing money deploying their cloud but they recognize that all the services and software will be bought from the company whose platform customer is on.
MF is IBM's only platform left. Whether it's good or bad doesn't matter.
Your point is throw away the only platform left without 1st gaining any momentum elsewhere.
That's just stupid but if you're an intern it's fine to be stupid, just lose the arrogance.
Also i get plenty of interviews so maybe you need to make a good resume and present yourself well.
You guys are arguing the age old question of Good, Fast, and cheap (you get to pick two) These are very large generalities, but mostly hold true.
Good and Fast (mainframe and Power) = pricey but quite stable and reliable
Good and cheap (Intel) = low quality (they have been trying to patch holes for 30 years in the OS and design) My guess is ransomware is 3 steps ahead and LINUX is only opening the door further
Fast and Cheap (single function ) = slow (not always true about speed, but single function maximizes design)
Good luck with your mainframe argument, BUT mainframes continue to run the world, and their underlying code is pretty rock solid. You can’t say the same for Intel, Windows, or Linux
If people took the time to understand the concept of the mainframe, you would realize how many of those concepts have been broken out into today's technology: virtualization, hierarchical storage, geographically dispersed compute/storage, parameter driven options. As others have already stated, today's technology is often a (poor) regurgitation of what the mainframe was doing many years ago (and in a much more structured, manageable fashion).
I grew up with the mainframe, it was my computing foundation. I still apply those concepts I learned then, in today's world, and it serves me well. The biggest problem with the mainframe today is that it isn't seen as being cool, more of a dinosaur. I contest it is still a very capable platform, but thwarted by a lack of education offered to newer generations, with more focus on the multitude of "sexier" platforms and their associated coding requirements.
Sorry for my poor English and laziness bro, brah.
I'm very sad to see people judge others based on their English skills. If you were skilled based only on English, you would have been a CEO. But you are still a lazy/outdated mainframes.
IBM MF is trash and will always be. Once it was a great product but not presently. Accept it and move to cloud or other tech companies.
@2cex+1beUeseK
"Its possible they won't have one because of the number of people leaving the Austin site at a fast clip. "
How high is the attrition rate?
If you still work at IBM, I would suggest you learn the mainframe because at the end of the day IBM will end up being just a mainframe company. This BS of OpenShift and CloudPaks will get sold off in time once IBM figures out it can’t compete with the big boys.
It’s not worth getting spun up by whomever is posting this garbage about mainframes being trash. I replied a couple of times myself because what was being said pis**ed me off, but the poster can’t even speak proper English, much less understand what’s best for the company
IBM will not pull the plug, they will just move it to where the burden rate is less. Kyndryl anyone?
IBM will soon be Powerless no matter what, so it's best to just pull the plug on Power in a controlled fashion, and take whatever lawsuits and customer loss that ensues from the fallout.
I am wondering how Power will do their 2021 layoff (they have had one every year for the last five years and revenue is falling off a cliff)
Its possible they won't have one because of the number of people leaving the Austin site at a fast clip. They are now trying to keep specific people by tossing money at them but all the ones I know that were given that option turned it down. For those not on a list, they are doing nothing for them.
There are a lot of older workers on the chip side and they will most likely not stay for a P11 if there is one. Also people on power chip side are getting pulled off to help Z on a regularly basis which will delay or cancel P11 plus P10 is struggling to get out.
So they may not need one because of how fast the Austin site is losing people. There are way too many jobs in Austin and places like AMD and Nvidia are handpicking the best engineers out of IBM.
AWS and Apple too.
There are also IBMers in Austin that are "WFH", basically gave up, burned out, just going thru motions and others using the moment to get IBM to pay them a boatload more to prevent them from leaving too by first getting an offer and then getting their management to try and match.
" thinking maniframers doesn't know design patterns, CI/cd, cloud and distributed storage techniques, API development"
Are you serious??? All those things were on mainframes long before small systems even existed. The small system script kiddies just reinvented the wheel and made new computer messes because they were too lazy to learn what already existed... and tripped over all the same issues along the way.
"To save this company, I support director and other few intelligent people who asks to scrap MF. "
Again, are you serious? There's nothing wrong with MF hardware? It's the most powerful commonly used general computing architecture out there...
Geezus the stupidity that the media has perpertrated is unbelievable.
I don't have a time to answer your questions again and again. I am still targeting reputed tech companies. Based on this exp, I am currently giving many interviews for full time with way good pay. Hope would get it into a good company unlike you IBM loosers.
As soon as I talk I work in mainframes, they just rejects me thinking maniframers doesn't know design patterns, CI/cd, cloud and distributed storage techniques, API development. So I have to cut the cr-p and tell them I am working for IBM cloud and Redhat tools, which is false. But still I am managed to go around based on my school skills. Mainframes are total trash in 21st century. And no one would value ur skills.
To save this company, I support director and other few intelligent people who asks to scrap MF. At least, hide that work from the customers. They are interested in cloud and AI. Once IBM was a great hardware company, now everyone needs to isolate hardware and sell best software on top of that, which IBM is missing since last many years.
If IBM just want to stay a mainframe company, it's very near to be 10-15 billion$ company in next 5-10 years. After kyndryl, already 50-60 billion revenue. Competitors are making 200+ with less number of employees and exec.
Accept the truth and move on.
@jpt+1beUeseK
Can you detail "Power is finished."?
#internforpresident 2024
"Intern here. Currently working on z project. We are testing CI/cd for z os. It's totally cr-p, can't work productively. "
Oh ge-z, this makes me laugh.
Apparently you don't even understand basic computing...
Your supposed "productivity" is based on layer upon layer of bloatware and cr-p to get anything done.
Mainframes are easy to get things done on.
If you need an IDE to do it... that is to say, can't do it with a simple text editor... you are walking on complicated crutches.
Reminds me of Enzo Ferarri... when Mitsubishi came out with the 4 wheel drive 3000... the motoring press asked him... "Enzo... now that the Japanese are making a 4wd sports car, will you?"
Enzo responded, "People who know how to drive don't need that cr-p"... LOL!
Intern! Please tell us how to save the company!!! It is clear that you know how to do and work everything so of course we would love your input after being here for a few weeks.
Cut the cr-p, guys!!! Mainframe is not worth a dime. Believe it or not, you older guys just see the market.
Intern here. Currently working on z project. We are testing CI/cd for z os. It's totally cr-p, can't work productively. I better enjoy my school project. I highly doubt how would customer be happy with such cr---y services on MF.
It's time to say goodbye to IBM and you old re--rded folks. Mainframe sucks and so as IBM or hybrid cloud strategy.
"In my opinion, cloud was invented to resolve the flaws by mainframes and legacy systems. "
I believe cloud was invented to solve the hardware market saturation problem.
The idea was to migrate folks to a subscription based payment scheme rather than having to continuously sell new hardware.
IBM's earlier attempt at it was called "On Demand Computing"...
Cloud was not about flaws in mainframe computing, but about the critical flaw in selling something that is a durable good... how to get repeat sales in a saturated market.
Heck, mainframe VM/360 goes back to the 60's. IBM was part of Darpanet. IIRC, a 360 was one of the original nodes...
There's no problems with mainframes. Mainframes can be "cloud"... the key element of cloud is the provisioning layer...
The problem is one of perception and marketing. The small/distributed systems folks out marketed mainframes... and did it with less advanced hardware...
FWIW I have yet to see an "more modern" or "open source" or "cloud" anything have the same quality of documentation the early mainframe stuff had.
The new stuff isn't better, except in perhaps one way... GUIs.
"I think there is a world market for maybe five computers." said Thomas Watson - he was talking about what would someday become the mainframe.
I think he may turn out to be correct.