Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

phoronix.com

Post reply on HN
Print view View original

561–570 of 722 posts

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#561

Post by pclmulqdq »

ryao wrote: post #534
pclmulqdq wrote: post #533

Earlier quoted context omitted.

You are both correct. AI inference is a comparatively easy problem from the perspective of parallelization when compared to most HPC problems.

Facebook did a technical paper where they described their training cluster and the sheer amount of complexity is staggering. That said, the original poster was interested in inferencing, not training. https://arxiv.org/pdf/2407.21783

Training is close to traditional HPC in many ways. Inference is far simpler since it's a simple forward-going pipeline of a relatively small working set.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#562

Post by LMYahooTFY »

wing-_-nuts wrote: post #185
philistine wrote: post #158

Earlier quoted context omitted.

I'm baffled that PC gamers have decided that 1440p is the endgame for graphics. When I look at a 27-inch 1440p display, I see pixel edges everywhere. It's right at the edge of losing the visibility of individual pixels, since I can't perceive them at 27-inch 2160p, but not quite there yet for desktop distances. Time marches on, and I become ever more separated from gaming PC enthusiasts.

I used to be in the '4k or bust' camp, but then I realized that I needed 1.5x scaling on a 27" display to have my UI at a comfy size. That put me right back at 1440p screen real estate and you had to deal with fractional scaling issues. Instead, I bought a good 27" 1440p monitor, and you know what? I am not the discerning connoisseur of pixels that I thought I was. Honestly, it's fine . I will hold out with this setu…

It's true but I don't run into this issue often since most games and Windows will offer UI/Menu scaling without changing individual windows or the game itself.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#563

Post by LMYahooTFY »

knowitnone wrote: post #461
laweijfmvo wrote: post #122

Earlier quoted context omitted.

Can it compete with the massive used GPU market though? Why buy a new Intel card when I can get a used Nvidia card that I know will work well?

warranty. Plus, do you want a 2nd hand GPU that was used for cryptomining 24/7?

Throw a new cooler on it and then yeah sure.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#564

Post by my123 »

michaelmrose wrote: post #528
barkingcat wrote: post #517

Earlier quoted context omitted.

No matter how few GPU cores the M4 has, it is still an extremely potent product on the whole. Better than most of the pc's out there.

Do you mean subjectively because you find Mac pleasant to use? The $600 mini certainly isn't more performant than 85% of desktops.

> The $600 mini certainly isn't more performant than 85% of desktops.

The average desktop isn't exactly a gaming machine. But a corporate box or a low end home desktop with a Core i5 and using the iGPU.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#565

Post by ein0p »

treprinum wrote: post #26

Why don't they just release a basic GPU with 128GB RAM and eat NVidia's local generative AI lunch? The networking effect of all devs porting their LLMs etc. to that card would instantly put them as a major CUDA threat. But beancounters running the company would never get such an idea...

Pretty obvious stuff, right? I mean, you don't even need HBM for that, you just need a TON of memory channels. Sure, that kind of setup would only be efficient for highly coalesced reads/writes, but that's what you need these days for inference - highly coalesced reads and writes. You could even get by with 64GB of DDR5. DDR5-4800 (rather modest) is 38.4GB/s per channel. To get 1TB/s you'd only need 26 channels. With the more expensive DDR5-6400 you'd only need 20. That doesn't at all sound insurmountable for a company of Intel's caliber. Heck, break up the dies (and the channels) across several chiplets even, if the interconnect is decent it'll still run really well.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#566

Post by qludes »

philistine wrote: post #158
giobox wrote: post #97

Earlier quoted context omitted.

For sure its been a sweet spot for a very long time for budget conscious gamers looking for best balance of price and frame rates, but 1440p optimized parts are nothing new. Both NVidia and AMD make parts that target 1440p display users too, and have done for years. Even previous Intel parts you can argue were tailored for 1080p/1440p use, given their comparative performance deficit at 4k etc. Assuming they retail at…

I'm baffled that PC gamers have decided that 1440p is the endgame for graphics. When I look at a 27-inch 1440p display, I see pixel edges everywhere. It's right at the edge of losing the visibility of individual pixels, since I can't perceive them at 27-inch 2160p, but not quite there yet for desktop distances. Time marches on, and I become ever more separated from gaming PC enthusiasts.

That's more of a function of high end Nvidia gaming card prices and power consumption. PC gaming at large isn't about chasing high end graphics anyway, steam deck falls under that umbrella and so does a vast amount of multiplayer gaming that might have other priorities such as affordability or low latency/very high fps.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#567

Post by slavik81 »

phkahler wrote: post #354
bayindirh wrote: post #307

Earlier quoted context omitted.

You use at least half of this stack for desktop setups. You need copying daemons, the ecosystem support (docker-nvidia, etc.), some of the libraries, etc. even when you're on a single system. If you're doing inference on a server; MIG comes into play. If you're doing inference on a larger cloud, GPU-direct storage comes into play. It's all modular.

No. I've been reading up. I'm planning to run Flux 12b on my AMD 5700G with 64GB RAM. CPU will take 5-10minutes per image which will be fine for me tinkering while writing code. Maybe I'll be able to get the GPU going on it too. Point of the OP is this is entirely possible with even an iGPU if only we have the RAM. nVidia should be irrelevant for local inference.

The Ryzen 5700G is one of the APUs tested on the Debian ROCm CI [1]. It works quite well with the Debian / Ubuntu system packages.

[1]: http://ci.rocm.debian.net/

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#568

Post by wtallis »

amaranth wrote: post #482
hulitu wrote: post #207

Earlier quoted context omitted.

> These are $200 low end cards Hm, i wouldn't consider 200$ low end.

There isn't a cheaper card that's worth buying over using the iGPU you already have so yeah, that's the low end.

NVIDIA and AMD don't even make new GPU silicon this low-end. Their smallest current-gen GPUs all debuted at higher price points, though the Radeon RX 7600 is now available at the same price that the B580 is launching at.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#569

Post by IceHegel »

treprinum wrote: post #26

Why don't they just release a basic GPU with 128GB RAM and eat NVidia's local generative AI lunch? The networking effect of all devs porting their LLMs etc. to that card would instantly put them as a major CUDA threat. But beancounters running the company would never get such an idea...

After carefully reviewing all of the other comments explaining the many technical and organizational reasons why they should not “just do that,” I have come to the conclusion that it was a big missed opportunity by intel.

Re: Intel announces Arc B-series "Battlemage" discrete graphics with Linux support

#570

Post by ryao »

jsheard wrote: post #248
p1esk wrote: post #228

Earlier quoted context omitted.

Apple could do it. Why can’t Intel?

Because Apple isn't playing the same game as everyone else. They have the money and clout to buy out TSMCs bleeding-edge processes and leave everyone else with the scraps, and their silicon is only sold in machines with extremely fat margins that can easily absorb the BOM cost of making huge chips on the most expensive processes money can buy.

Transistor IO logic scaling died a while ago, which is what prompted AMD to go with a chiplet architecture. Being on a more advanced process does not make implementing an 512-bit memory bus any easier for Apple. If anything, it makes it more expensive for Apple than it would be for Intel.
Post reply on HN