Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Businesses The Courts

Nvidia Hit With DOJ Subpoena In Escalating Antitrust Probe (reuters.com) 13

According to Bloomberg (paywalled), Nvidia has received a subpoena from the U.S. Department of Justice as the regulator seeks evidence that the AI computing company violated antitrust laws. "The antitrust watchdog had previously delivered questionnaires to companies, and is now sending legally binding requests," notes Reuters. "Officials are concerned that the chipmaker is making it harder to switch to other suppliers and penalizes buyers that do not exclusively use its artificial intelligence chips."

The development follows a push by progressive groups last month, who criticized Nvidia's bundling of software and hardware, claiming it stifles innovation and locks in customers. In July, French antitrust regulators announced plans to charge the company for alleged anti-competitive practices.

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This discussion has been archived.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

Nvidia Hit With DOJ Subpoena In Escalating Antitrust Probe

Comments Filter:
  • Say what? (Score:4, Insightful)

    by quonset ( 4839537 ) on Tuesday September 03, 2024 @07:11PM (#64760060)

    who criticized Nvidia's bundling of software and hardware, claiming it stifles innovation and locks in customers.,

    If Nvidia is creating everything in house, how would these people like them to go? Shoudl Nvidia go to AMD and hand over what they've done so AMD can get their act together? I fail to see what innovation is being stifled if everyone is racing to catch up to Nvidia. These other companies are having to find new and inventive ways to compete, the very opposite of stifling.

    If and when these companies find a better, less expensive way to do what Nvidia is doing, guess what? People will leave Nvidia and go with someone else.

    This is the opposite of what Microsoft did by deliberately sabotaging their own software so competitors couldn't get a foothold, or bullying any competitor [zdnet.com]. Let us know when Nvidia does the same.

    • Re: (Score:2, Interesting)

      by Tailhook ( 98486 )

      Kid, it's election season. Campaigns need money. Shaking down the giant money piñata that is Nvidia is par for the course. The incumbent party that controls DOJ et al. gets to tap that a bit of coin. Nvidia will make some DNC contributions, fund the right non-profits and this will go away after the election. That's all this is.

    • Re:Say what? (Score:4, Insightful)

      by drinkypoo ( 153816 ) <drink@hyperlogos.org> on Tuesday September 03, 2024 @09:57PM (#64760322) Homepage Journal

      If Nvidia is creating everything in house, how would these people like them to go? Shoudl Nvidia go to AMD and hand over what they've done so AMD can get their act together?

      Nvidia's single biggest competitive advantage isn't their hardware, there appear to be plenty of people who don't mind spending a little more power on some other solution. It's CUDA, and they seem to be trying to prevent anyone from making any CUDA-compatible APIs. Though, not very aggressively... yet? I'm not super doom and gloom about it but even as a committed Nvidia customer for years (every time I've tried to go AMD I've regretted it, more on that another time if anyone wants to hear it again) I really absolutely don't want them to try to prevent anyone else from making their own compatible interfaces. Openness is of significant importance to me, and while I would also prefer open drivers and firmware and documentation I am absolutely adamant about open APIs.

      I still don't have any experience with significantly 3D accelerated UNIX(tm) machines (my Indigo and Indy both had entry graphics) but I do go back to and have experience with almost all of the first consumer-grade PC 3d graphics accelerators except perhaps Kyro. I had PowerVR, Mystique, Permedia 2, Voodoo 1-3 (and later a mac with a 5) riva 128, tnt, tnt2, ati rage and rage pro and I know I'm leaving some things out here but what really made all of those which weren't a bad joke relevant was support for OpenGL and Direct3D. All of the also-rans which went away had to have specific application support and they left only the ones with a GL implementation (if we're willing to count MiniGL.) And GLIDE died out before 3dfx did (And without it, we probably never would have been cursed with Direct3D.) Nvidia still has excellent OpenGL and Vulkan support today.

      Nvidia put into their license agreement (and then I think took it out? that might be wishful thinking, I'm lazying out here soz) that you weren't allowed to come up with a competitor by reverse engineering. But reverse engineering for the purpose of interoperability is explicitly protected by the DMCA, so pfffffft*

      * Translation, I am quite perturbed and I guess I will start having to learn about AMD GPUs again just in case.

    • by tlhIngan ( 30335 )

      It's basically CUDA. You cannot implement a CUDA like layer without potentially incurring the wrath of nVidia, so even if AMD releases a great GPGPU, if it doesn't run CUDA, it's useless to most people.

      It's basically vendor lock-in.

      You might be wanting to look at Apple who is basically in a similar position - they created the hardware, software and app ecosystem for their device all with their own money and effort. Yet they can still be guilty of many things, as seeing EU laws and even the DoJ lawsuit attes

      • by etash ( 1907284 )
        yes but why should nvidia be forced to open up its protocol/library/api? it's not like they are forcing anyone to use theirs. What you're suggesting is it should be okay to reverse engineer cuda and have an opencuda to be used to other hardware too, just so developers don't have to rewrite their cuda code? It did take them a lot of time and resources to write that code, billions probably. It's like saying they should be forced to opensource it. None should be forced to write open source code
        • why should nvidia be forced to open up its protocol/library/api?

          APIs are uncopyrightable open standards, that's why.

          Open Standards are what make computing work, that's why.

          Reverse engineering for the purpose of interoperability is explicitly protected by the DMCA, that's why.

          TL;DR: It's the law, anything else holds everyone back, and again, it's the law.

    • The problem is that everything that they release is primarily a lock-in, anti consumer tech.

      It has been their MO since they purchased Aegia and tried to lock everyone using PhysX.

      The other problem is, we are no longer smart consumers, but literally sheep that blindly follow these corporations.

      Perfect example, DLSS. Back then the media would criticize it because of its anticonsumer nature.
      Now? Its a praised as a must have feature.

      But since we no longer have real reviewers, instead we have bribed influencers,

  • 1 - Threaten company with DOJ charges
    2 - Congressional hearings ensue
    3 - Politicians from both parties buy the dip
    4 - Company cleared, nothing to see here, LOL!!!
    5 - Politicians dump at new ATH
  • they're regressive in every way. this is stuffy propaganda: remap words

Computer programmers never die, they just get lost in the processing.

Working...