Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Supercomputing Cloud Microsoft

Microsoft Reveals Its First Quantum Computing Chip, the Majorana 1 (cnbc.com) 31

After two decades of quantum computing research, Microsoft has unveiled its first quantum chip: the Majorana 1. CNBC reports: Microsoft's quantum chip employs eight topological qubits using indium arsenide, which is a semiconductor, and aluminum, which is a superconductor. A new paper in the journal Nature describes the chip in detail. Microsoft won't be allowing clients to use its Majorana 1 chip through the company's Azure public cloud, as it plans to do with its custom artificial intelligence chip, Maia 100. Instead, Majorana 1 is a step toward a goal of a million qubits on a chip, following extensive physics research.

Rather than rely on Taiwan Semiconductor or another company for fabrication, Microsoft is manufacturing the components of Majorana 1 itself in the U.S. That's possible because the work is unfolding at a small scale. "We want to get to a few hundred qubits before we start talking about commercial reliability," Jason Zander, a Microsoft executive vice president, told CNBC. In the meantime, the company will engage with national laboratories and universities on research using Majorana 1.

This discussion has been archived. No new comments can be posted.

Microsoft Reveals Its First Quantum Computing Chip, the Majorana 1

Comments Filter:
  • by TeknoHog ( 164938 ) on Wednesday February 19, 2025 @01:39PM (#65179515) Homepage Journal
    It's not crashing, it's just in a superposition of up and down! https://courses.cs.washington.... [washington.edu]
  • Way behind (Score:4, Funny)

    by brunes69 ( 86786 ) <slashdot@keirs[ ]d.org ['tea' in gap]> on Wednesday February 19, 2025 @02:59PM (#65179741)

    IBM has been making quantum chips commercially available for years. Their latest Heron chip is 156 QBits.... that is way, way more advanced than this Microsoft chip.

    • IBM has been making quantum chips commercially available for years. Their latest Heron chip is 156 QBits.... that is way, way more advanced than this Microsoft chip.

      But can the IBM chip be in both crashed and non-crashed state at the same time? That's a feature quantum chip users have desperately been clamoring for, and Microsoft is sure to deliver.

    • by ceoyoyo ( 59147 )

      Comparing quantum processors using only the bit count is even worse than comparing cameras by only megapixels. The Microsoft processor uses topological qubits which, if Microsoft's claims are true, are much lower noise than the qubits in other chips and potentially easier to scale up as well.

      • by TeknoHog ( 164938 ) on Wednesday February 19, 2025 @05:15PM (#65180167) Homepage Journal

        The Microsoft processor uses topological qubits which, if Microsoft's claims are true, are much lower noise than the qubits in other chips and potentially easier to scale up as well.

        The problem with topological qubits is that they can't tell the difference between a coffee cup and a donut, so I wouldn't bet on them.

    • Not really fair to compare them by qubit number. Microsoft took a higher risk approach to quantum computing. Their qubits are based on majorana qubits. IBM has been doing the older more established method of superconducting circuits (Josephson junctions.) Which type of qubit is better in the long run is still up for debate. TL;DR Microsoft thinks they have a new type of qubit working.
      • Sounds good in a PR message, but it not only is not alive, it's not proven, and not validated in any way. And while the advance *could* be quite a leap, today, it's hope and a prayer and a stock price boost.

        Can we count the number of stock price boosts that turned into dust especially in quantum computing?

        "Hey Merv, this AI stuff isn't making money unless we put a subscription on Notepad... what do we do?"

        "OK Brad, let's do this-- Quantum Leap! It works for lots of tech companies! Um, how about talking abou

        • by ceoyoyo ( 59147 )

          but it not only is not alive, it's not proven, and not validated in any way.

          I don't know WTF you mean by "not only is not alive", but Microsoft has published several papers about their topological qubits, including the one linked in TFS.

          • I've actually read those papers.

            This is not a working system. It does not do work. It does not compute, it has not been audited, it exists in some quantum states that have done no real work, have no real language support, and are fiction until it's not.

            The theory is juicy. It's an approach that might be viable. And it's years away. We'll know its viability when 8q actually work with low noise-- and this might happen by the end of the decade. Until then, it's a placemarker. Juicy placemarker, but placemarker

            • by ceoyoyo ( 59147 )

              "it exists in some quantum states that have done no real work, have no real language support"

              Sorry, I don't know WTF you're talking about.

  • Ywa, AND Microsoft will also solve inflation, house the unhoused, remove crooked traitor republians and have AI become "Actually Intelligent."

    Or

    None of that. Stupid media for repeating stupid MS PR that says outrageously impossible things.

    IBM is almost up to 160 Qubits. I'm prettty certain (I'd bet money on it if any of the online betting sites would offer i) that MS isn't going to have a million quibits let alone on one chip anytime this century.

    Microsoft has sucked since they stole CPM to make PC-DOS, s

    • IBM is almost up to 160 Qubits.

      Not all qubits are created equal.
      IBM uses transmon qubits (information encoded in the Josephson junction)
      MS is using topological qubits, which are largely immune to the noise decoherence that comes with trying to operate qubits as close to absolute zero as you can get.
      They are a better technology. One topological qubit will be worth many transmon qubits, if the technology pans out.

      Microsoft has sucked since they stole CPM to make PC-DOS

      Fiction. Did not happen.
      SCP's QDOS was written to look like CP/M, MS licensed the rights to it.

      stole PC=DOS to make MS-DOS

      Fiction. Did not happen.
      MS-D

    • Hopefully it runs Shor's algorithm effectively, at least.

      You forgot to mention that they stole Unix to produce XENIX.

  • by Ol Olsoc ( 1175323 ) on Wednesday February 19, 2025 @03:20PM (#65179795)
    And the crowd goes mild.
  • by 4wdloop ( 1031398 ) on Wednesday February 19, 2025 @05:13PM (#65180159)

    .. windows?

  • Stupid story is stupid.

  • Be another two decades to get it working in a product.....hope it would be brown !

The key elements in human thinking are not numbers but labels of fuzzy sets. -- L. Zadeh

Working...