https://www.thedailyupside.com/technology/semiconductors/ibm-and-amd-team-to-develop-quantum-centric-supercomputers/
14 replies (most recent on top)
64TB of RAM ?!?
But I heard that 640K should be all that anyone would ever need!
The investment by Intel for big memory or multiple threads just isn’t worth it to them. Their product is good enough for the mass market, and they will concede the “Niche” marketplace where power shines. NOTE Intel is eating the low end of Power at an accelerating pace mostly due to ISV’s embracing Intel Linux. IBM realizes this, and has offered Power cloud as the alternative for customers and ISV’s still supporting legacy code stacks. This is not a bad strategy given the very low margins on small Power servers. I expect IBM will consolidate around one cloud scaleout offering as time moves forward due to economics. This in no way means Power isn’t a great offering, but it does mean IBM can’t make it economically viable as a general on prem offering. Instead as time moves on, IBM will offer “enterprise” on prem solutions and everything else will be a virtual solution.
The only thing keeping Power alive is SAP and the only reason SAP is keeping it alive is the amount of memory it supports (HANA is an in memory DB). Power11 goes up to 64TB. Historically Intel Xeon capped out at 4TB per socket. You could go higher but required a custom setup. If Intel offered 64TB support OOTB, SAP would drop Power like a hot rock tomorrow.
I suspect it’s IBM acknowledging they backed the wrong horse when it comes to Linux.
Linux on Power has its “niche” for particular ISV’s. Yep we are looking at you SAP (big memory and multiple threads play well for enterprise customers) For everyone else, Intel is good enough.
When it comes to Linux on Intel, the market has spoken, and power reflects that with their market share losses (aix and OS/400 have both seen 50% + drops in legacy ISV support)
It was time to change horses, and Power was in the bullseye
There is an under the radar observation to be made which is interesting. Look at the AMD content that has wormed its way into IBM products sets. This is especially true when it comes to storage.
This playbook has been used ever since AK came into leadership. ( and probably before ) It’s really just gaslighting. That’s all it is. Sleight-of-hand and misdirection. Make up some hype that will be the next great thing. Then when it comes time to deliver. , you forget about this and come up with another great hype to be the next great thing . Use as many buzz words and cliché as you can. See how many investors you can fool. Really is amazing the corruption level. The greed of the senior leadership at the cost of the company is really sad. It was once a great company. I worked there for 18 years. I saw the spiraling first hand. Mostly stunningly bad over the last handful of years. Every year spiraling down has accelerated a little more. At some point people will stop wanting to play the shell game., and the house of cards comes crumbling down .
Let’s prop up the stock price anyway we can so I can sell my share before the next earnings call.
Nothing wrong about lying… let’s have more of that.
Does anyone believe McKinsey other than senior manglement?
I read a comment about C racker B arrel and their debacle.
"C racker B arrel is not a broken brand but it has a broken board."
Now replace CB with IBM.
An investor warned CB MULTIPLE TIMES... "don't do it" was his message, but the CB CEO and the board didn't listen, ignored him completely, and went with it anyway. Burned tons of cash, stock value, etc.
IBM needs a courageous investor that has the c oo j oo n ee s to speak up, sooner rather than later. But who?
It should come as no surprise that Alvind is a con man and scam artist (the low calibre type).
He's trying to salvage the shreds of his bonus for 2026 buy making false declarations all over the news. He would sell his mother and grandmother to the Mexican cartels if they were around.
Yes, we are all saying the same thing, with different words, nuances - but we are in agreement that this is the same AK B S attempt to H Y P E for survival. "Fake it long enough for as long as he can"
In the meantime, time goes by, money is drained and American jobs are gone down the toilet. But "who cares"...
1) This has absolutely nothing to do with AI.
2) Quantum computers (meaning truly fault tolerant qubits or noisy qubits with good enough error correction that can run actual quantum algorithms like Schor's or Grover's) DO NOT EXIST TODAY, so announcing a partnership to build them is nonsense.
3) This is just a silly press release whose purpose is to get IBM back in the news and arrest the stock price decline.
I started a post about the "AK and Gang Quantum Hype" (due to the AMD so called "partnership"), but the moderators (whoever runs this website) deleted it. I guess it's because it was also discussing s t o c k sss.
But I said there the same thing. So DITTO.
I'm still having a hard time (yeah, I know, I'm that si_lly) with the fact that nobody is seeing AK and Gang erratic behavior, which is a clear signal that they have no eh fing idea what they are doing. Throwing s h IT to the wall to see what sticks has more intelligence than these people.
Quantum is the next AI hype inflation bubble that produces no working products but sends billions of dollars upwards that will never come back to our economy, better jump in now
This is not changing horses (authentically), but AK being desperate to get some traction and notoriety like Intel did.
He would love to be invited to the WH and have his photo op, some more money from us tax payers to spend and 5 minutes of fame.
He's afraid they will kick his a-s soon, because good ol' IBM is not being invited to any parties.
Ego is something, isn't it?