Recent content by StoleMyOwnCar

Let me know how that is, vs DSV4F. I've seen people talking about running GLM, wondering about what the quality is like vs the best 2 spark solution. Obviously 4 nodes you're talking like 16k though so I hope you have some ideas for ROI on that. That's what I meant when I thought that the Spark...

If they come with dual boot out of the box, then that's good. I had a lot of issues with trying to run Mint in any stable fashion on the HP laptop. Note that Windows tries to randomly search and destroy any Linux partitions it finds sometimes IME, so yeah. Careful with that.

The one I actually tried out at some point was not the Flow but rather an HP variant that was a traditional laptop (nice sturdy metal frame, too) with an OLED display. It was about the same price or a little cheaper at the time. Unfortunately my screen had some issues and by that point I was...

It has a lot of active development because the idea is that you can use spark cluster(s) as a development platform before you can deploy the exact same thing straight to enterprise level Cuda hardware (6000 Pro, etc). For sort of close-to-frontier model running, it takes about 2 sparks connected...

If you look at the Microcenter open box section, there are some 128GB laptops with the same chip in them for less than 3k USD. Obvious caveat is you need a Microcenter near you, but it's a nice way to get a travel-capable PC that you can also game on while running these models. On the other...

I mean, it isn't surprising. The current state of the world is that getting an entry level position just fucking sucks thanks to AI. Get rich quick and gambling schemes are kind of getting popular because there's a tiny chance that they will work, but hey if they do they sure as fuck beat all of...

I have a DGX Spark and I primarily run Qwen 3.5 122B int 4 autoround on it (well it's a hybrid int4/q8 quant). Gets about 50 tok/s gen on average at 260k context size. Have had some issues with looping, but otherwise it's a pretty smart model, and image interpretation capable. I used to use a...

You know what? I guess I actually support this. I always support more work and research into space. I didn't think it would be due to a bunch of idiots trying to build datacenters on up there to escape regulations, but hey if it gets us a moon base... yeah, I guess I don't give a shit.

Doesn't help. All of my LG TVs don't have any internet access, but it doesn't matter because this is via MS, not the monitor. Yeah and unfortunately you also miss out on firmware updates, which can either be a big deal or not a big deal. I'm pretty sure that my TV has a few things that could be...

To be fair, that's a real concern with Amazon. Sometimes they really do not take care of items that should be given WAY more careful packing and handling, especially if it's shipped by them. If you go over to the DGX Spark variants (just search "GB10"), a lot of the bad reviews are from the way...

Technically you can run DSV4F off of Antirez's repository on 128GB. It's just if you have 128GB, there are things that are technically faster to run anyway. 128GB can be had for about 4k right now, as a DGX Spark. Some AMD options are a bit cheaper. 256GB is about 8k, because that's 2x DGX...

Oh no doubt. Big Orange and co are about the only ones that really come out ahead in his own administration. A bunch of fucking institutionalized bribery on a scale hitherto unseen. Shutting down access to open weights is literally just trying to build a motherfucking monopoly on it. Greatly...

So I used that Linux based Hotspot application. For my HP card, the delta from the hottest core on the card to the hotspot, is about 10C. So if the card was at 78C (which it gets to doing long Stable Diffusion workloads with relatively low fan settings), the hot spot will be about 88C. The...

My 5070 Tis have had absolutely no issues driving 4k 120 even in a game like Indiana Jones, with DLSS. Game looks amazing. 24 is nice to have, but considering how stupid prices are right now, is it worth the inevitable cost uptick? This is going to be the shittiest Super refresh in history.

Is there any Linux app which can do the same? Edit: https://forums.developer.nvidia.com/t/new-tool-to-get-more-gpu-sensor-information-memory-temp-voltages-blackwell-hotspot/377308 Found this, might try it.