DDR4 RDIMM memory is 3200 MT/s at 64bit (or up to 72bit w/ ECC) which 8bits/byte would give you 25.6GB/s. what you are missing is that memory is multi-channel. Intel Xeons can go up to 8-way memory, AMD Epyc can go up to 12 way. So you're talking 204GB/s or 307GB/s. DDR5 RDIMMS are even faster so throughput would be greater. What I do not know is the inherit latency of memory access (core to core, or package to package) where I suspect the mainframe would be better (lower latency). Is main memory on the z16 static ram or dynamic? Likewise PCIe I/O is up there on server platforms, a dual processor AMD Epyc has 128 PCIe lanes (Gen 4 for Milan, or Gen 5 for Genoa) for a total of 256 PCIe lanes of generation type. However should be stated that just pure lane counts does not compare well. Mainframes are set up to OFFLOAD a lot of signal & I/O processing much more than mid-range/server. (concept in the server world is just taking hold with dedicated processing units (APU/DPU) but mainframes have been doing that forever. There are a bunch of other items as well, but it's like comparing an armored personnel carrier to a car. Both can get you from A to B but one is much more robust. :)
Steve is correct! And the odds are there are other factors and subtleties that make such a comparison of questionable value anyway. But unless they let me benchmark it... :-)
The Z16 has plenty of local DRAM. That robot around 14m is inserting DDIMM (not DIMM) modules. They work with a transactional interface called OpenCAPI (aka OMI in a simplified version) that runs around 32GB/s full duplex compared to DDR4 25GB half duplex, and the Z16 chip has a lot of those channels (they are lower power and take up half the silicon on the CPU chip of a DDR4 channel). So that bandwidth is additional to the distributed "virtual L4" comprised of idle borrowed L2 capacity. As the L2 cache is SRAM it is likely significantly lower latency than DRAM even if it is in another CPU.
@@monad_tcp ...It's more about electronics than electricals. The specific _course_ would probably be computer engineering. Let's wait for the original commenter to offer further advice.
@@monad_tcpCheck out Google sometime. They can answer inane questions like this. 45 seconds: Computer Hardware Engineering majors. Oh wow, look at that.
I retired in 2020 and I refused to work on z16 because it would have kept me working to age 81! But I was a z system architect up thru z15 and had worked on the mainframe since the system 360 model 85 which contained the very first memory cache. I always knew I was on a team working on one of the most truly great products ever conceived. I loved this presentation. Thank you.
I started on an IBM 1401 in high school then an IBM S/360-91 at UCLA. Joined IBM as a Systems Engineer. I installed 3033s, 3084s and 3090s, etc. It was an amazing journey!
The 360-85 was years ahead of it time. I assisted with hosting several corporate benchmark sessions at the Kingston plant in 1969 before it was withdrawn from the market because of its speed. Years later I had a occasion to work on a 3033 and realized it was just a 360-85 with faster logic chips, solid state memory and a CRT console. Like others here, I have worked on 1401 1410 1460 360 370 and 4300 systems. Plus 20 years on PCs from about 1990 to 2011. Now running Proxmox and Truenas in my retirement lab. Lots of GREAT memories!!!
Graduated college in 1968 wirh a BS in Computer Science. It made a great career on IBM Midrange - System 36, 38, and what is known as IBMi (last I heard anyway). It was quite a ride that lasted >50 years. It would be interesting to see how it would stack against the Z
Thank you so much, Dave, for this fascinating tour! Now retired, I worked 41 years as an Application Development Software Engineer in the barbarian Microcomputer Revolution of the great unwashed out on the vast steppes. I always wondered what was still going on high up on Mount Olympus in the original Birthing Power Mainframe World! Now I have gotten a glimpse of the still evolving Sacred Temple! I absolutely LOVED this! Kudos to all of you Mainframe Engineers and Technicians commenting on this thread who worked in this World in the lofty Clouds! Homage to all of you!
This is INSANE, every time you mentioned a feature I thought there was no way they could top it. And then they're just casually like "oh yeah and these 64 5.2GHz cores can also run 2 bajillion VMs and encrypt all of main memory and correct for cosmic rays"
@@jovetj Because I got an earworm from that, I need to share it with you. I'm standing in your server room 'Cos I got kicked while playing Doom The lights are dim but I can see Your hardware is so quality I need to apply the patch So I can get back in the match But something's caught my eye A mainframe reaching to the sky
@@AllenCavedo Yes. The system I worked on was a small business server, and I restarted it weekly, but if something went wrong with it we usually wouldn't even know, IBM would just show up _out of the blue_ to fix.
People have been predicting the death of the mainframe for a long time-3 decades, at least. It's not coming anytime soon. It's just too easy for people to poke fun at or write off what they do not understand.
@@MaximZemlyanoyabout half of mainframe workload is now Linux. And much of the z/OS workload is now Java. It's time for this "it's top expensive to move" meme to go away
@@briansomething5987Linux is good, Java is 💩, Apache should use Zig and Bun, no need for a Dinosaur, Linux should be done on Zig for even more Performance
@@willd0g Assembler I expect. Operating System components were written in Assembler in my day, and IBM used PL/S a lot, and later PL/X. Both could GENERATE(inline assembler code).
@@maxpowerlive8852 Its for a machine. Empedded XP. This thing works and does one job and only one job. Why change it? Chances are, this thing doesn't even have a network connection - and IF it has, surely no Internet connection.
You’re unstoppable, Dave! Hearing an engineer like you talk about mainframes provides invaluable content. Thank you for sharing your knowledge with the world.
You overlooked a key reliability feature .. the IBM Z series includes extensive internal self consistency checking of the logic inside the CPU so it can detect when the logic fails and retry a transaction, shutting down the core if necessar. "Commodity" CPUs are tested after manufacture but after passing test most later errors are silent. Recent testing for silent errors in servers at Google and AWS have found that silent errors are quite significant. IBM has taken them seriously for decades.
That is one of the most important features, and also one of the most important differences between mainframes and micros. The US government has a huge set of operating guidelines that must be met by the bank's IT departments, and I'm out of date on that stuff, but those generic CPUSs probably don't pass those guidelines for certain transaction processing.
The first machine that I was responsible for as a system programmer was an IBM 4341 back in the early 1980s. It was used for software development and testing running VM/CMS and supported 6 concurrent users. They’ve come a long way.
We had some baby 4300's--41, 61, 81. A 4341 could easily support 30 or 40 users in VM/CMS 370 without a sweat. We even bonded two 81's together with one as a "real system," and the other as a "test dummy," running loads and having the 2nd frame testing responses. Basically debugging one mainframe from another! LOL Our "big iron" was 3033s, 3081s, 3084s (replaced by 3090s). Cool toys! And we even have a baby 9370. I miss those days and the cold DC room.
Six sounds pretty few, we had 70 or so on a System/370 Model 168. Under VS2 Release one, all TSO users shared a single address space and swapping was painful (3330, 3350 DASD). When we got OS/VS2 (MVS), everyone had their own and things were fine. In Australia, the 43XX computers had special firmware available to allow them to run Fujitsu operating systems, OSIV/F4 and OSIV/X8.
Company I worked at the 4381 was the local branch machine, anything really needing processing went to the remote 3090/600S. The VAX/780 was used for DTF and ascii/ebcdic conversions. Now I'm all nostalgic about loading tapes/carts. lmao
WOW! Quite a system. I started working at IBM in 1967 and retired in 2005. My first IBM system was the 360 model 50. It had a single 32 bit CPU and weighted 3 tons. The max RAM was 512KB and four I/O channels. The CPU cycle time was 500nS. RAM speed was 2uS. It managed about 200k instructions per second. We’ve come a long way baby.
I started on a 360-30. IBM used to supply a Systems Engineer along with the machine back then. Lots has changed since then. It was quite an interesting field to be in.
This is fascinating. I was a programmer on IBM systems for many years before commoditized systems took over most processing tasks in many businesses. It's nice to see that mainframe systems are still relevant and that deep-dive security and fault-tolerance are pursued with such meticulous zeal. It's also cool that this is still being done in the US. Thanks so much, Dave, for producing this.
well, you look at 13:14 and now you know why. some things need to be really really reliable. what happens to a normal PC data-center when there's such an event ? the entire data-center needs to fail-over to another entire data-center. You can't just keep working because of annoyances like earth-quakes.
One trend I'm noticing in these comments is a lot of "was" and "now retired" people. What are the prospects of these systems going to look like in the near future when the next generation of techies coming up won't know F all about languages like COBOL, etc.? Who will maintain these applications in the future?
Thank you Dave for the in-depth tour. As a software dev for over 25 years on the mainframe, it truly amazes me at the vast raw horsepower that the mainframe has. Job well done on the BTS of the Z series.
As a software engineer for z/VM , you really DO get impacted by ALL of these things. The way you handle I/O failures, System failures etc.. in the code, is crazy. Came from being a web developer. From React/NextJS/insert JS framework here to PL/X and Assembly. I went from high up in the stack to relatively low on the stack.
Being a retired guy who started out as a customer engineer on RCA IBM 360 compatible mainframes your video brought back many memories. As years went by I supervised an IBM customer system programmer group whose responsibility, amount other things, was to evaluate vendor hardware, mainframe and peripherals. As part of the evaluation we made many trips to IBM and other vendors around the country. We did an anual trip to Poughkeepsie even stayed at the IBM homestead a few times. I was even part of an IBM customer council looking at proposed new technology. Anyway, your RU-vid series is a great source of current and past IT technology. This video was a surprise in that it covered technology very few are exposed to.
If you have access to the source code of Assembler H and scan for my initials, JCS, you might find evidence I was there. I also did work on PL/I and LE/370, but that was mostly testing and I didn't get to make my mark.
Yes. Mainframes are awesome, and they have been for 53+ years. IBM hasn't always been that awesome, but they have sure kept their eyes on the ball as far as their core product.
@@marred2277 Yeah I interviewed a while back and I remember feeling it wasn't really up to snuff with the more modern companies, perhaps "progressive" is the word. In any event I wasn't impressed.
Your channel gets better and better! Thank you for making such interesting content and sharing your own life and experiences. I appreciate it very much.
@@knerduno5942 Because after Dave started making more money on his RU-vid channel than his wife explaining these things we lie awake at night wishing we understood or knew someone in the business to keep us old farts up to date he is not no longer confined to the garage.
I work at the Poughkeepsie facility you visited. It's very cool to see you here as you were one of the first channels that I ever watched when I started building computers many years ago. And now I'm working at IBM to put myself through school. Thanks for inspiring Dave!
I grew up in a mainframe family. Dad was 40 years at Burroughs/Unisys. I was a co-op electronics student at IBM, then worked summers at Burroughs/SDC/Unisys. My now employer of 30 years relies on IBM big iron, and the young folks coming in have no idea. All they see are the apps that still run green screens, not the heavy lifting happening behind the scenes. Thank you for this episode.
the reason your videos are SO good to me is that the parts i dont fully understand are still entertaining, i i think thats due to your delivery. no clickbait , no artificial suspense , just solid presenting. love it!
Their fill station still runs Windows XP. I guess when you absolutely positively need your fill station to work every time, you only rely on the very best. The way you put this Mr. Dave. Brilliant!
I remember a 1990s news story buried by most news services, US Navy outfitted a ship entirely with Windows XP, in control of everything. Maiden voyage and all computers blue screened leaving the ship dead in the water, they had to tow it back to port. Fortunately it wasn't in war time.
@@jrstf XP was released in 2H 2001. If you're referring to Windows 98 (first edition), that would be believable because it crashed with a BSOD during the full press demo of plug-and-play.
@@jrstf If I remember correctly this fiasco was caused by some early Windows NT. An incorrect manual data entry on a keyboard caused the complete ship to shut down and it had to be towed back to port. It was not possible to restart the engines.
If my deteriorating memory is right then a university professor found the XP system would crash at intervals because a 32 bit counter overflowed(?) I think it was about 7 days.
@@jrstf"Buried by most news services..." or, as an intelligent person would say, "not reported because it's not exactly newsworthy to the vast majority of people." But sure, it was "buried".
Worked originally as a COBOL programmer on IBM MVS mainframes. Many career changes I finally left the Mainframe behind in about 2000. Found this interesting. I watched loads of organisations move away. Great to see they are still around!
Dave; thank you for bringing us along on your journey through IBM's new mainframe. I joined IBM in 1958 as a Field Engineer on a computer system IBM bult for the US Air Force used for air defense called SAGE (ANFS-Q7). The computers (each SAGE building had 2 mainframes for reliability) were housed in a 3 story building (with major air conditioning) and each computer had 68,000 vacuum tubes. Main Core Memory (RAM) was 64K of a 33 bit word, and the system ran on a 6 microsec memory cycle. Following my time in SAGE I returned the commercial business word and I worked on 1410/7010's in Boston. Then 360 models 30, 50, and 65. Then I was transferred to engineering in Poughkeepsie and worked in the 705 building supporting reliability on the 303x systems.. So what a journey when I compare the Z16 to my life with mainframes and PCs (I still build PC's once in awhile). Thank you - Fascinating (and I wish I could start over again with a Z16). How much maintenance is required for one of these?
> How much maintenance is required for one of these? Not a lot, they basically very rarely go wrong, and when something does it can be replaced in flight generally.
My company is the largest financial institution by assets and I joined in 2007... the funny saying that management says every year during planning... The mainframes are on their way out. Then every few years we do a massive mainframe upgrade lol... Those things are beasts!!! They chunk through data faster then anything else. You have to do a ton of custom integration, but damn, are they massive processing powerhouseS. When the micro second counts to make money in trades and bank transactions, nothing beats a mainframe. That single drawer is just one part. The mainframe is the full cabinet. You customize it by each drawer. In each of our data centers, we have MANY of the full Z frame setups. They are enormous and the redundancy is insane.
@@mephInc nope, black rock owns damn near everything. But we work with black rock and at our annual RIA conference that my company puts on (and I owned and engineered the technology for the event for 7 years) over 800 desktops, 400 laptops, 2000 monitors, two racks of full mobile servers, blades, SAN, networking in a rack. Where in the rack we would run around 30,000 virtual desktops that would build on demand and deploy for items and destroy after. At any time, 3,000 vdi desktops running, about 30 virtual servers running the back end systems and network capacity of 4, 10Gbps fiber links fully redundant running from 4 different isps, and setup a funny NOC/SOC onsite, and lots more. Now I'm out and moved on from that stress, but they run a lot in cloud now that we've started some implementation into that area. But we are the largest financial institution by assets, as in customer money in accounts. With our combined Acquisition we made 2 years ago and just on the end of finishing integration, nearly 11 trillion under our management. Official is 8 trill because of they are not finished moved over yet and the old name we purchased to go bye bye in history. Been with this company since 2007 when I was 23. Started as a contractor doing 21" CRT to dial 15" LCD monitors in two buildings here of around 9000 employees. And moved to support, then took on engineering and built and designed our VDI deployment of 12000 virtual desktops for offshore and retired our old HP blade computers (like a rack mounted laptop where a chassis had like 25 of them. And moved on to desktop engineering and then to security when we formed our new security org, then to various parts where I owned the products and built security policies and procedures and deployed a full rdp/ssh/admin (secondary) account vaulting and session management, recording, keystroke logging, monitoring and review, etc... and then to lead architect. Literally started as the lowest on the tottem pole and in my 16 years there, moved up to sr manager and lead solutions architect of some of our more sensitive security infrastructure. Lol, and since day one... mainframes will be out in the next 2 -3 years, we are retiring them! But nope. There is just things that with active customer trading, world wide bank transactions, batch processing, and much much more, that a mainframe can do in a micro second with 100% accuracy that a computer or server can't. And when a milisecond is the difference of tens of thousands of dollars, they are worth their ungodly price tag. (Millions spent yearly on them for support and maintenance and IBM has staff onsite that their job is to sit and if something breaks, for it right away. And a new unit is multi millions when you factor in power, cooling, data center floor space, setup, configuration, migrations, fibers, interconnects, etc... it's mind boggelinging
Same at my financial comoany. MF is a dirty word. I think its cultural. The new sexy buzzwords get the attention, but all trades still clearing via the MF
I worked on an AS/400 back in the 90's. One memory was coming in on a Monday morning to greet an IBM engineer in reception. He said that our AS/400 had suffered a cache problem over the weekend and had contacted them for help. He opened it up, replaced the CPU cache board and closed it. The system basically said "thanks" and it proceeded to speed up to full speed. Absolutely astonishing. And that was over 30 years ago.
I worked for AMEX and then IBM and fully understand a) how and why a mainframe fits particular market segments in a way no other computing environment can and b) just how committed and dedicated the IBM team is. Thanks for this review. It is such a mind boggling presentation and you handled it well. I definitely don't understand a lot of it, but I do understand the why of it.
In 1969 I landed a programmer trainee job at a department store in Cincinnati. We had a 32K 360/30 running 360 DOS and Assembler was the first programming language I learned. After being away from mainframes for 20 years, I applied for a position at a shop that had 6 System 390's. The hiring manage asked if I thought that I could still do assembler - I said if "D2" was still the op code for the MVC instruction - then yes. He was surprised that I knew most of the op codes and could even read a core dump.
This was so much fun! I cut my programming teeth on System/360 in the days of punched cards. We threw a little party in the computing center when we added the second megabyte of core memory. Next came Control Data mainframes (two 6600's and a 7600) at a national accelerator lab. The modules on these machines made electrical connections to a backplane, but also freon connections to the refrigeration unit. Things have come a very long way since then, and it's good to see that IBM is still in the game.
In the late 70s I had access to a System/370 model 168MP. MP meant two full CPUs configured as a single computer. One evening I dropped in, and after a short conversation it was agreed I can have my own computer for a few hours. Without interrupting the workload, I showed an operator how to separate out a CPU, some RAM, a channel, some DASD, a 3270 controller so I could have a console for doing stuff. Then got out my 3330 disk pack and ran OS/VS1 for a few hours, doing a SYSGEN for our computers. When it was done, I showed an operator how to put it all back together. Back then a CPU alone occupied a whole box, and there was another to connect the two CPUs, and each channel also had an equally large box. I had to read the numbers on the sides to see what was what.
One thing Dave did not really cover (maybe because he didn't work on them himself) is that mainframes are a whole other world of computing. Sometimes you just have to forget what you think you know. Many of the same terms are used, but in different ways than you might expect. Partitioning is a key aspect of mainframe architecture. The bare-metal partitioning mentioned is called an LPAR. I'm sure you can find more details about that by searching for that term. A mainframe operating system can then soft-partition available memory to different tasks. This essentially contains each process. Each program running at any given time runs in its own [memory] partition. In the frame cabinet, behind the front door, there is one or two laptops that are affixed to the frame to be opened up and used. This is called the Support Element and is the lowest-level access to the complete machine's operation and status. Satellite computer(s) outside of the frame called the Hardware Management Console can do many of the things the SE can, but over the years the distinction has gotten less and less. If you wanted to turn on or reboot or partition the mainframe, you would use the HMC or SE. As mentioned, when failures happen, a hot backup instantly kicks into place, and IBM is automatically notified so a repair will be scheduled. Edit to add to my comment: The zSeries and z- mainframes are a complete 64-bit architecture. They also started out being fully backward-compatible even for old software written in the 60s. Some of that very-older comparability has been removed or minimized in newer z- models, but backward comparability is a huge assurance to customers.
Well i have worked for a long time on the software implementation end for IBMs os and other servers and know that they configure storage devices and use mount points directly to deploy oses, they are tricky to configure, I have not configured personally but I have worked on these mount points or something they called luns to setup our softwares, they even have replication directly from the storage itself
Get the Hercules (softlabs) emulator and you have a S/370, S/390 and Z-ish mainframe enough to run even z/OS and z/VM. Then go on a quest and hunt down the dasd images of whatever OS/390 or z/OS you can find. Then hunt down the documentation. Start reading the MVS and JCL book from Ranade publishing. Be prepared to read the IBM manual "Principles of Operation". This manual started in the sixties so start with the manual for S/360, that most amazing family of machines. You can also get started right away with a public MVS 3.8j distribution that is actively maintained by the community. Now you have enough to get started, you could start learning about JCL in about an hour and submit your first assembly job to MVS $HASP today.
The redundancy around the Storage is quite impressive. You can lose a disk array in one data centre and its replicating counterpart in another data centre can take over the workload without it impacting on normal IO on the mainframe... Just an increase in IO read write time due to the change in distance when going cross site.
When I went into the USAF in 1993 (AFSC 2E231), I spent almost a year at Keesler AFB learning component-level troubleshooting and maintenance of the old Sperry mainframe that still had spindle hard drives, punch card readers, and large reel-to-reel tape drives plugged into it. Boy, how times have changed. Of course, by the time I got my first PCS to McChord AFB, the 62nd Communications Squadron was actively pulling the main frame out of service and replacing it with Novell 4.11 servers and Windows for Workgroup 3.11 on Zenith Z 386 workstations.
I started my IT journey on the Burroughs mainframe. The "Large" flavor (or "A Series") was a marvel in its day. Based on the first virtual memory, multi-processor architecture the CPU was a non-Veneman design that was optimized to run the ALGOL programming language. Like the IBM machines, these were reliable, and had massive I/O bandwidth.
I/O Transfer rates was always the differentiator, the amount of data they could move was huge compared to Mini's and Micro's.. Add to that the focus on reliability and the culture that runs through most of the Mainframe folks on high availability it remains a substantial weapon. Never worked on them but working with a major bank had to back-end our Internet banking to Hogan on the mainframe as you say its always has to be there. Love your reference to the cosmic ray memory flip, I have used that for years to explain to random reboots. Dave you keep reaching new levels in your content please keep it up.
I love the pure brute force approach of these chips and hardware. I do see Epyc using some more advanced technology though lately - TSMC N5 and stacked 3D cache for >1GB of cache per socket. But that’s still not as complete of a caching solution as what these mainframe chips can do. Thanks for the tour Dave!
Think about the absurdly low yield IBM has on those chips and how expensive they must cost. Low yield means a lot of refused chips when manufacturing because of defects, it must be absolutely perfect, which is very hard for a chip that big. They probably don't even do binning.
I was a MVS systems programmer back (way back) in the 80’s. It would be really interesting to me to know something about the OS environment(s) running on these things. I get it about the virtualization capabilities. But there is still something managing them. Thanks Dave!
Microcode, the MCL (on the SE) loads microcode (in POR) that creates the environment for the Image (LPAR). Your Image then has an environment with channel access to storage to load an OS from (IPL (Initial Program Load)). That's called IML (Initial Microcode Load), and until you've done that the Image environments don't exist. Once it's been done IPL for the specific Image environment becomes available. MVS has been 'rebadged' as is the current trend, to z/OS, it still looks and feels like MVS, so you'd recognise it. It's also had a whole lot of modification to make it more Java friendly and cloud capable. z/VM is the micro partitioning OS Dave mentioned, and is still quite common.
Current z/OS systems programmer. For the most part, the base MVS part of the system is largely quite similar (except for of course an expansion in instruction set to accommodate things like 64-bit instructions and addressing). There's lots of new stuff too. Most POSIX compliant Unix programs can run natively now. So lots of modern subsystems will have an interface running under Apache for instance rather than VTAM (although the old SNA protocols still exist, and ISPF is still the most common way for systems programmers to work).
It is fun to see how much everything in that place still looks totally recognizable as an IBM site, eventho I haven't been inside one since the 1990s. Oh, and the entire build of those machines also looks very much like the IBM I used to work for. Cool stuff.
Thanks for another great video Dave. Very interesting. I started my working career learning COBOL on a Univac 90/70 for a UK insurance company, later they moved up to a Univac 1100/60. It's great to see a video that brings back a load of nostalgia for me! (Sadly I may not be able to watch your channel for much longer. Recently everyone on RU-vid has been suffering much more harrassment from the company due to the use of ad blockers. If this continues, RU-vid have said they will block me. But I will hold out as long as I can! Best Wishes from the UK.)
There is a browser available that can get around it. I can't name it directly as the comment will be auto deleted. However, its name is the opposite of "coward". It will fix your problem. I've been using it for over 5 years .
Wow. Dave is awesome. Best IBM salesman whose pitch I have heard in years. IBM salesmen were fascinating - targeting their focus on $ or technical stats depending on audience to sell mainframes or to move to the next level system.
Wow! I started playing with electronics in 1969 (Heathkit), built my first radio in 1970 (Tandy), bought my first computer in 1982(C-64 I still have) and started EE school in 1989 after playing around. Watching this still amazes me. Where we have come in just one lifetime leaves me speechless. My son is at the Naval academy in Annapolis doing EE. What is he going to see in his lifetime? (okay, I started him in computers on the C64 :)). Thanks for the video. Keep up the good work.
This is a great video! As someone who works at AWS in one of their data centers around their next gen servers, this was so cool to see what IBM is up to :)
This was Fantastic Dave! If you ever get the chance to be invited to anything like this again, please go for it! This was so informative, and with your own style of storytelling that we love so much. To anyone else thinking about offering to invite Dave, please just do it!
I was there in the summer of 1992 with my father Albert . It was quite impressive. We stayed two weeks in a lodge Motel in Fishkill, NY. Father started out as a programmer for IBM Germany, he then became programming instructor in the Training facility in La Hulpe near Bruxelles,moved back to Germany where he planned and designed high speed Networks .
You mentioned virtualization, fortunately I was a IBM Systems programmer at the dawn of it's introduction. Worked DOS, DOS/VSE (skipped SVS & MFT), VS1, MVS and ZOS. The early days technology and software were much less mature/sophisticated and more like the wild, wild west. While up-time was relatively good it was nowhere near current reliability standards. Enjoyed your presentation and since I'm retired no longer cringe when the phone rang at 2am from the computer operations center. :-)
This was seriously cool! Thanks for doing this Dave and thanks to IBM for letting you guys in. Would love to hear the question and answers you previewed at the start of the video.
Interesting video, I do appreciate the varied content! With all the focus on NVIDIAs supercomputers, it was interesting to learn about what IBM focuses on and how they differ.
I remember my first IT job in 1999 working with NT40 and wondering what was up with the ancient mainframe guys at my company. Now I'm the old man. Respect.
Such a great and thoughtful overview. I hadn’t thought about mainframes in years, and now I get it. The uptime and local performance is unmatched. Thanks Dave!
I am frequently a target of derision for my love of IBM Z series. Mostly because the love of IBM's big iron somehow contradicts my 50something OG 90's hacker quasi anarchist vibe. My detractors are rubes and I feel about them the say way that Christopher Walken feels about people that don't like hot dogs. I haven't finished the video yet but given that you are doing it and given the subject matter I a preemptively liking it. :)
30 seconds into this episode you already earned my thumb up. You're dead on target with this episode's core question (my creative alter ego rewrote your theme question) "Why a z16 instead of a PC with a 4090 graphics card and the perfect gamer mouse?"
Dave this is terrific for any number of reasons. First, it's useful and informative. You share information efficiently and effectively with your (admittedly esoteric) audience. I am old, so I have to stop and replay frequently to keep up with your rapid-fire speech, but eventually I get it all! This video was especially fun for me as a retired early 1980s mainframe junkie and professional capacity planner/performance analyst. It was really interesting for you to update facts on the age-old distributed vs. centralized argument. I've long held that for most production workloads, the key limiting resources are I/O capacity and the speed of connection(s) between components. The mainframe has yet to become irrelevant. IBM has clearly not lost sight of its old "Reliability, Availability, Serviceability" mantra. Finally it was especially amusing to see you in full fanboy mode. Did they let you keep the IBM sweatsuit? Well done all the way around. Thanks to you and to IBM!!
Thanks Dave. Whenever I watch your videos, you confirm a) it's worth to be a part of the computer history b) it's worth to learn. I started with an IBM 4361, learned all about Microsoft since DOS 2.x, Windows 2, 3.11, NT 3.5, SCO Xenix, Linux, RASPIs up to todays Cloud Systems.
Yay! That was fun! Can you do a follow-up episode on Z16 system software? I bet that's just as interesting as the metal (and glass and water and silicon and...)
I remember Microsoft claiming five nines. I made my reasoned rebuttal to the Trade Practices COmmission (Australia). The TPC accepted the MS explanation, but those ads ceased. My rebuttal was simple, take away minimal maintenance time for Windows (NT), there just wasn't enough time left. It's not better with Windows 11. Linux, I reboot to change kernels. Everything else I upgrade, but choose whether to restart, individually.
This video is giving me some hope that IBM marketing/communication efforts on the midrange and mainframe technology will improve and will be more focused on the way information should be presented today. The effort I need to get IBM technology financed by management is way higher than the one needed for x86 computing, networking or security, because when people hear the mainframe word, they think to old dusty things that runs legacy code written for archaic programming languages that you access trough a text based console. Yes, mainframes can also run that kind of things, but most fo the time you need a mainframe to run very specialized tasks at state of the art level.
Thanks for sharing this information Dave! Myself, I like technology and I can be quite active on the computer but I'm realizing lately that there are so many details I don't know about how it all works together, and having someone like yourself give us an overview and explain how we interact with some of those things (like mainframes) in our daily lives on the internet is really helpful. I also found the shots of the computers bouncing and shaking hilarious. 😄 Great video!
Fascinating video. It's quite amazing to see how far mainframes have come. I just retired from a 40+ year career in IT. I started my career supporting MVS and VM on IBM 308x series mainframes. My primary role was to support VM/SP, VM/HPO and VM/XA for a decade along with secondary support on MVS and SNA networking. I left the mainframe world and moved to Unix and TCP/IP support in the early 90's. Seeing what a mainframe has become is astounding given I know the technology you described very well. Great work, Dave!!!
IBM is doing all they can , in the last couple of years to attract new minds to the mainframe. I wonder if it’s too little too late. Something that this video did not have time to touch is software discipline. It’s one of the keys if the mainframe success story .
I have been in an IBM Z team for a little over a year now as a sysprog. It was pretty foreign to me as I am only 23 and had only done a little bit of mainframes at university. I wish there was more young talent coming into mainframes but we're trying!
IBM continues to develop and build mainframes because there are customers who will pay for mainframe capabilities. If there were less-expensive solutions that worked for customers, customers would buy them instead.
One of the major advantages of the mainframe… is that the same size support staff for the OS is all of its software does not change from the smallest machine to the largest machine …. For example, a SSA, two systems programmer handle all of the zOS instances… all of your SSA interactions, including the Web interface is on Two machine geolocation diverse … 50% split use and dynamically back each other up … Two CPUs.
So what you are describing is a SYSPLEX, a cluster of mainframe images (LPARs) running as a single system (the 50% split gave it away). The geolocation diverse line might make yours a GEOPLEX, so somewhere on your mainframe complex there will be a CF (Coupling Facility) that handles the Image to Image communications between the two physical systems (it might just be in an LPAR on one of the processors).
Fantastic presentation. I was in engineering at DEC for a number of years working on some of the larger systems we made. The dimensions of the problem are the same (speed, connectivity, redundancy, maintainability, etc), but the numbers on these machines are truly mind boggling. There are just some parts of the computing world where you need all the features these machines bring. And fun to see that the passion for building machines like this is still around.
Dave's Excellent Video Production Services strikes again! Dave, amazing video! I'd (almost) PAY for that video loop of the drive up there with the changing leaves! As someone at a very large multinational company with IBM mainframes, and soon to have the Z16s to replace the 15s, I can say how cool, sexy, and awesome they are. Z16s are far, far, far past any mainframes we had in the 80's and 90's. I do run mainframe code at home on my "homemade mainframe," ( LOL ) but it's not the same of course. Although I only need 300W of power on my Dell PowerEdge T320 to run VM/CMS and MVS/370. Please THANK IBM for allowing you such access to their new toy. I'd love a view inside there myself. But I geek out on stuff. Oh and congratulations on marrying your "childhood sweetheart" and still being together. As guys in our generation, it's rare, but so very awesome.
Dave, cool video! I remember reading that financial tractions would be processed by 3 mainframes in lockstep with comparators on the input and output. That would catch errors made by one mainframe and use the output from the agreeing 2, flagging the error. If all 3 disagree, it would halt the system. You could take one mainframe down and not lose redundancy. Thanks for the video!
Dave this was a great video to help understand the Mainframe more. I work for a company that does Mainframe hosting (I work on the opensystems (VMware) side) and truly didnt and still dont really understand them. But I am always fascinated when they bring a new one in and online. Love your videos and keep up the good work sir.
I worked at IBM in 2001 & 2002, all I remember was white hallways, black door trim, and the occasional poster of sailboats, all lit by the cheapest fluorescent lights money could buy. One of the departments I worked in managed the office spaces in each building, we had one room in the basement we kept off the lists where we "stored" a ping pong table and mini fridge. At the Santa Teresa location (San Jose, CA) there was a twisty roadway that led up to a golf course that was popular with street racers, IBM had an aerial photo done of their buildings and you could see all the burnout marks on the street from the racers. The golf course clubhouse was a popular location for lunch.
The construction of these machines is quite impressive. It’s interesting to see related processors crunching so much data. I work mostly with the Power architecture, but the small version of it - a three-core e200z4 with a eTPU timing coprocessor. The heatsinking here is very impressive! In my world (automotive ECU design) we run them at 150MHz to keep the power dissipation down.
Shhhh, don't tell people that their cars are likely running small versions of these CPUs. Well actually the eTPU is running their engine, the CPU is just watching and adjusting.
Last place I worked had a Z series running a medical records system. Mainframe itself was not in a standard rack just a huge monolithic monster but very stylish. Redundant everything including 2 racks filled with as many 1tb drives as they could hold. I think the whole thing required 6 50 amp circuits. Redundant power too. It was an amazing beast and thank you for helping me understand it better.
It's important to understand the difference between a mainframe and an I/O device. What you are describing with this: > Redundant everything including 2 racks filled with as many 1tb drives as they could hold. Is an Enterprise Storage Server (likely a DS8000 series machine). Mainframes don't have hard drives (except for the management elements) They outsource I/O operations to control units (CUs) in I/O devices. So a mainframe doesn't store data, it gives it to a CU and says store that. The storage controller handles storing it and where and data protection measures. Latest Enterprise Storage Servers do sub 1ms response times. Imagine every card transaction that takes place being credit checked in real time as you do it (because it is). The mainframe (processor) software application requests the record from an Enterpise Storage Array and gets an answer back in less than a millisecond. So when you swipe your card, and it says approved, that went across the communications network to the mainframe, which processed it and requested the relevant records from a DB on the storage, the application checked the numbers and answers the credit check (approved/denied) and the application then responds back to the retailer, over the network. Whos card machine says approved/denied. That's happening realtime on every transaction for every card right across the world. 24*7*365. Mainframes for the processing, Enterprise Storage Servers for the data storage. It is such a volume of traffic that a £15 million upgrade in that system can pay for itself in a day of processing those kinds of transactions, because of the increased volume of transactions that the upgrade facilitated.
@@davetdowell The mainframe itself was it's own beast. There were 2 other racks next to it full of IBM branded storage top to bottom. Fiberchannel stuff if memory serves it's been a while. I don't know what the full specs of the mainframe was but it was supporting about 20,000 total clients + printers.
Great video Dave!!! I lived in Cold Spring, NY, 18 miles south of Poughkeepsie. My older brother worked at IBM in Poughkeepsie and East Fishkill, NY and retired from there. My father worked there part-time for a few years. A few families in Cold Spring worked at the IBM complexes in Poughkeepsie and East Fishkill.
I used to work repairing and switching terminals of IBM in the early 2000's. An doing things as installing special network equipment, but just to a point, I made sure the Cisco Internet connection racks, huge machines 4U most of the times, connecting them to the outside via secure encrypted lines for BT and then the engineers of IBM took over to connect them to their encryption racks, for banks, the robots of car manufacturing and other high availability dependant systems. I always was fascinated at the tech IBM had deployed and only saw strictly what was required to have signed my worksheet by BT and IBM. That was at the peak of my carreer and I learned to respect Big Blue. Great video Dave, thanks!
Really cool seeing the end product I work on IBM's processors. We just submit our stuff and don't hear much more as we start on the next chip It's also really cool to see all the test that are done. We don't get to see these from the IBM plant I work at. Great video!
Cool video Dave! Mainframes will be around for quite a while due to the quantity of data they can move. I wonder how the Cray supercomputer I/O back in the day compares to today's numbers. Maybe in a future video you can stop by HP and get a look at their NonStop systems?
@@mikestewart4752 I saw a CDC Cyber, designed by the same bloke. Unclocked logic, tuned with not-quite-random length pieces of copper wire. I think it was octagonal, with one side open to provide space for the CE to work. This was at CSIRO in Canberra.
Cray was bought by SGI. What made SGI great was openGL and a separate bus for reliable uninterrupted high bandwidth communication between boards. The separate bus idea was a precursor to PCI. Eventually new SGI management came in and wrecked it’s huge advantage. (Draining reserves by buying Cray, creating a joint effort called Fahrenheit with Microsoft for an alternate to OpenGL which evolved into DirectX, and settled cross litigation with a startup NVidea by sharing their IP plus allowing them to cherry pick SGI’s best employees).
The one thing IBM needs to do, but has always been vehemently reluctant to do is provide a hobbyist version of zOS... Eventually they're going to run out of people that have the knowledge of how to run these machines. One of the reasons that OpenVMS still exists is due to DEC and those who bought them maintaining the hobbyists that became employees running those systems. Hercules can run zOS (or at least the older versions of it), and I imagine there would be a significant number of fellow geeks that would enjoy being able to learn on it.
@gt2847c - Long ago, I thought about a hobbyist version of VM/CMS but what programs would I run on it? Same question about zOS. These operating systems are not designed for games or other programs that are typically run on PCs. . Also, I am sure that these operating systems are HEAVILY optimized for the IBM hardware and this hardware is RADICALLY different from a PC. It would be like the difference between a motor cycle with four-speed left-foot shifting and a semi which has a shift lever with two auxiliary thumb controls (a five-speed main transmission with two high-low auxiliary transmissions, giving 13 speeds). I think that it would take a huge effort to get zOS to run on a PC (or some other hardware that a hobbyist could afford to buy.
We have these in our data centers today. Not my primary hardware, but I still have to sign into it every once in a while. Pretty cool to see how the sausage is made!
I worked for 13 years in an HP wafer fab / both research and production.... I love these insights into other production facilities... Really gets the enginerd in me going.
I used to work with mainframes, would be nice to get back to them. Back around 1988 - 512 Meg main storage (RAM), 4 CPUs totalling maybe 80 MIPS and a few hundred Gig worth of storage. Supposedly our first government department to exceed 1 TB of connected DASD did that in 1993.
I started 20 years earlier. I recently spent a day scanning the latest equivalents to the manuals I used in the 70s. The assembler code I wrote then would still work today, i was going to say except for the loss of remote 3270 terminals, but I suspect those would still work. Provided I could cook up the appropriate (emulated) 270x controller. I had a 3704 back then.
My Burroughs B5500 mainframe (circa 1968) now runs as an emulator in my browser. It supports 8 tape drives, multiple card readers and printers, card punches, datacomms and head-per-track disk. It supports full virtualization of memory and can easily run half-a-dozen ALGOL and COBOL jobs simultaneously in 32K words.
Your memory calculation for memory bandwith (for Xeon or any other x86 with 3200MT DDR4) is ~25GB/s per single channel. Most desktop CPUs are dual-channel, however modern Xeons are 8-channel and AMD Epycs are 12-channel, with DDR5 (4800MT/s) so depending on system you could be talking (Xeon DDR4 3200MT 6-channel) 150GB/s up to (AMD Epyc DDR5 4800MT 12-channel) 450GB/s per socket.
So cool. I worked as an operator in the Toronto Computing Centre at 245 Consumers Rd from 1975-81. I started on the S360/65 and /75 and I do remember virtualized S370/158VM systems that were on the floor by the late '70s, although I went the S370/168 route with its MVS/Jes3. Phew, can't believe I remember those model numbers!
I worked a semester at IBM RTP in '85, working on a PC app. All our developer builds were performed by a 3090 (mainframe) in the basement. Compared to compiling on the PC, it was wicked fast. Since then, I worked at a bunch of places. Only one of them had a better developer-build solution than what I had back in '85... and that place definitely wasn't Microsoft.
It has been at least 20 years since I took a real look at what mainframes actually do and what's under the hood. Thanks for the great content. Interesting to see the principle of relieving software performance bottlenecks with hardware design innovations is still fundamental in design. I got to work for a small software company doing basic tech support in the mid 90's. They worked with HP mid range mini computers, and some of them used fiber technology for storage device IO cables and optic data lines on portions of some of the system boards.
I was an IBM Field Engineer, worked on System 360 and 370 mainframes. Found this to be very interesting. Much changed since my time in the field. Thanks for the video.
I did consulting for HP's at their Florham Park NJ facility back in 2000, doing HP-UX kernel functional tests, and later, stress tests, for the [doomed] Itanium. HP already had solid 64-bit machine experience in their PA-RISC processors. The facility had a raised floor room similar to what was shown here. Lots of iSCSI-over-fibre-channel. Main trunks were *several* T-3 lines. Was back in the day where the backing stores were all hard drives, and all except boot drives were SCSI-68/SCA-II. On the floor below us was where Veritas and ISO-9600 file systems were developed. Actual customer failover service was provided on their large 64-core machine for several key clients: US Air Force, Nasdaq stock exchange, and several airlines (though I got the client names 2nd-hand; never verified myself because I never used those machines. We were exclusively using the [new] Itanium boxes, which, for the record, first booted a LINUX O/S before bringing up our own HP-UX port!)
Just fricken awesome. I received my Comp Sci BS back in '83 when it was crucial to conserve and maximize every byte possible within the confines of a PDP11. 40TB of RAM? I sit here grinning.
It might be a licensing issue (like you pay for activated RAM and can upgrade later) or for redundancy. That way if a DIMM fails it can just swap in some of the spare 10TB.
