Nvidia's stake in Intel could have terrible consequences. First, it is in Nvidia's interest to kill Intel's Arc graphics, and that would be very bad because it is the only thing brighing GPU prices down for consumers. Second, the death of Intel graphics / Arc would be extremely bad for Linux, because Intel's approach to GPU drivers is the best for compatibility, wheras Nvidia is actively hostile to drivers on Linux.…
Conclusion: Buy AMD. Excellent Linux support with in-tree drivers. For 15 years! A bug is something which will be fixed. Nvidias GPUs are theoretically fast on initial benchmarks. But that’s mostly optimization by others for Nvidia? That’s it. Everything Nvidia has done is a pain. Closed-source drivers (old pain), out of tree-drivers (new pain), ignoring (or actively harming) Wayland (everyone handles implicit sync w…
Nvidia buys $5B in Intel
471–480 of 638 posts
Re: Nvidia buys $5B in Intel
#472Earlier quoted context omitted.
Where exactly did he say anything remotely violent
If you watch the clip and know what's going on in the US right now, saying such vile disinformation now nothing but aims to up the temperature. That's the deciding reason he got shut down too. Absolute inability to read the "room", even though what he said would be ugly at any point.
You can watch the monologue here https://www.youtube.com/watch?v=-j3YdxNSzTk&t=122s
Re: Nvidia buys $5B in Intel
#473The reason why Nvidia is buying now does not have to do anything with Arc or GPU competition. There are mainly two reasons. 1) This year, Intel, TSMC, and Samsung announced their latest factories' yields. Intel was the earliest, with 18A, while Samsung was the most recent. TSMC yieled above 60%, Intel below 60%, and Samsung around 50% (but Samsung's tech is basically a generation ahead and technically more precise),…
Are you comparing Samsung against Intel here specifically, or also TSMC?
Re: Nvidia buys $5B in Intel
#474Re: Nvidia buys $5B in Intel
#475Earlier quoted context omitted.
Hold on, you're only getting 45 tokens/sec with Mistral 7B on a 5090 of all things? That gets ~240 tokens/sec with Llama 7B quantized to 4 bits on llama.cpp [1] and those models should be pretty similar architecturally. I don't know exactly how the scaling works here but considering how LLM inference is memory bandwidth limited you should go beyond 100 tokens/sec with the same model and a 8 bit quantization. 1. https…
My understanding is that quantizing lowers memory usage but increases compute usage because it still needs to convert the weights to fp16 on the fly at inference time. Clearly I'm doing something wrong if it's a net loss in performance for me. I might have to look more into this.
If you're using llama.cpp run the benchmark in the link I posted earlier and see what you get; I think there's something like it for vllm as well.
Re: Nvidia buys $5B in Intel
#476Remember when Microsoft invested in Apple when Apple was down in the dumps? This is giving similar vibes. That deal was arguably what saved Apple near its nadir. I’m not a fan of Intel’s past monopolistic practices, but for the sake of sustaining competition in the CPU/GPU market, I hope this deal works out for them even half as well as the MS deal did for Apple.
>Remember when Microsoft invested in Apple when Apple was down in the dumps? This is giving similar vibes. Doesn't feel the same because the 1997 investment was arranged by Apple co-founder Steve Jobs. He had a long personal relationship with Bill Gates so could just call him to drop the outstanding lawsuits and get a commitment for future Office versions on the Mac. Basically, Steve Jobs at relatively young age of 4…
Investing in Apple and Borland were an counter-anti-trust legal move, keeping the competitors alive, but on life support. This way they could say to the government "yes there is competition".
Google does the same these days by keeping Firefox alive.
Re: Nvidia buys $5B in Intel
#477Earlier quoted context omitted.
Apparently it is 5% ownership. Does that give them enough leverage to tank Intel’s iGPUs? That would seem weird to be. Intel’s iGPUs are an incredibly good solution for their (non-glamorous) niche. Intel’s dGPUs might be in a risky spot, though. (So… what’s new?) Messing up Intel’s iGPUs would be a huge practical loss for, like, everyday desktop Linux folks. Tossing out their dGPUs, I don’t know if it is such a huge…
Intels gpus are a better solution for almost all computing outside high end gaming, ai, and a few other tasks. For most things a better gpu is overkill and wastes energy
Re: Nvidia buys $5B in Intel
#478Nvidia's stake in Intel could have terrible consequences. First, it is in Nvidia's interest to kill Intel's Arc graphics, and that would be very bad because it is the only thing brighing GPU prices down for consumers. Second, the death of Intel graphics / Arc would be extremely bad for Linux, because Intel's approach to GPU drivers is the best for compatibility, wheras Nvidia is actively hostile to drivers on Linux.…
Conclusion: Buy AMD. Excellent Linux support with in-tree drivers. For 15 years! A bug is something which will be fixed. Nvidias GPUs are theoretically fast on initial benchmarks. But that’s mostly optimization by others for Nvidia? That’s it. Everything Nvidia has done is a pain. Closed-source drivers (old pain), out of tree-drivers (new pain), ignoring (or actively harming) Wayland (everyone handles implicit sync w…
On the other hand, my Steam Deck has been exceedingly stable.
So I guess I would say: Buy AMD but understand that they don't have the resources to truly support all of their hardware on any platform, so they have to prioritize.
Re: Nvidia buys $5B in Intel
#479Nvidia's stake in Intel could have terrible consequences. First, it is in Nvidia's interest to kill Intel's Arc graphics, and that would be very bad because it is the only thing brighing GPU prices down for consumers. Second, the death of Intel graphics / Arc would be extremely bad for Linux, because Intel's approach to GPU drivers is the best for compatibility, wheras Nvidia is actively hostile to drivers on Linux.…
This misses the forest from the trees IMO: - The datacenter GPU market is 10x larger than the consumer GPU market for Nvidia (and it's still growing). Winning an extra few percentage points in consumer is not a priority anymore. - Nvidia doesn't have a CPU offering for the datacenter market and they were blocked from acquiring ARM. It's in their interest to have a friend on the CPU side. - Nvidia is fabless and has c…
> Nvidia is fabless and has concentrated supplier and geopolitical risk with TSMC. Intel is one of the only other leading fabs onshor[e]
TSMC is building state of the art fabs in Arizona, USA. Samsung in Texas, USA. I assume these are being done to reduce geopolitical risk on all sides.Something that I never read about: Why can't NVidia use Samsung fabs? They are very close to TSMC state of the art.
Re: Nvidia buys $5B in Intel
#480Nvidia's stake in Intel could have terrible consequences. First, it is in Nvidia's interest to kill Intel's Arc graphics, and that would be very bad because it is the only thing brighing GPU prices down for consumers. Second, the death of Intel graphics / Arc would be extremely bad for Linux, because Intel's approach to GPU drivers is the best for compatibility, wheras Nvidia is actively hostile to drivers on Linux.…
I agree. As a Linux user who favors Intel hardware due to their Linux support, I gotta say the future looks bleak.
> As a Linux user who favors Intel hardware due to their Linux support
I'm confused here. Are you talking about Intel CPUs or GPUs? And does AMD not have excellent Linux support for their own CPUs and GPUs?