This is fine, because a NPU is memory bandwidth and compute (compute performance is still based on memory bandwidth) in disguise:

Quote from: 48 and 64 GB RAM nice on January 23, 2026, 11:19:25The Ryzen 350:

Quote from: www.amd.com/en/products/processors/laptop/ryzen/ai-300-series/amd-ryzen-ai-7-350.htmlOverall TOPS
    Up to 66 TOPS (I think it's 8-bit / INT8)
NPU TOPS
    Up to 50 TOPS (same)

Quote from: nvidia.com/en-us/geforce/laptops/compareGeForce RTX 5050 Laptop GPU: 440 AI TOPS (4-bit, scammy NGREEDIA, so it's half that -- 220 -- in 8-bit)

GeForce RTX 4050-Laptop-GPU: 194 AI TOPS (8-bit)

194/66 = ~3 times, so it's 3 times slower.

3dmark.com/search:
4050 (notebook): Average score: 8288
Ryzen AI 350' 860M iGPU: Average score: 2885

8288/2885 = ~3 times, which is the same 3 times.

-> Looks like Ryzen AI 350' NPU has to be mainly understood as its iGPU, really. An iGPU is still an ASIC, the most power efficient way. Maybe a NPU is just marketing, instead of just saying it the way NVIDIA says it ("AI TOPS", no mention of a NPU).

Which tells us that it has been all along about what I said in my previous comment ("it's all about memory size, memory bandwidth and the usually, out of it, resulting GPU performance") ;)

And for running various AI LLM models locally and privately (see article: notebookchat.com/index.php?topic=315954.0 ("Your own ChatGPT, offline: AI without the cloud on your laptop")), a NPU does not need to be used anyway.

PS: It's almost like the comment from January either predicted or gave AMD an idea to release non-AI Ryzen. Tho, binning chips is not new and why throw these perfectly fine chips away, just because of NPU PR reasons? Certainly not.

PPS: Let me predict another thing: "Ryzen AI" will remove its "AI" in its product name later. It will still be AI capable, but as proven in the quote, its -- and generally speaking as well -- AI capabilities come from its iGPU's compute and that, in turn, comes from it's memory bandwidth. Memory size is the other one.

I've noticed that bonus programs tend to change quite a lot as platforms develop, and when I was comparing different options, aus-jokaroomvip.com caught my attention because it helped me understand how the current offers are presented and what has actually changed over time. In my view, JokaRoom VIP is a good example of how online platforms can adapt their bonus approach while keeping the overall experience easy to navigate. Older promotions often seemed more straightforward, while newer programs may include different conditions, loyalty elements, or more flexible rewards. I also think mobile access has influenced this evolution, since users now expect to check offers quickly from different devices. What matters to me is not simply the size of a bonus, but how clearly the terms are explained and whether the platform keeps its information updated.

Quote from: heffeque on Today at 00:19:10It should use 10-20 MB of RAM at most.

What the hell is it doing with the other 1000 MB?

Probably collecting data. Like Google apps. Google clock and calculator have alot of background use on some devices sometimes. And for what? Apps like that don't need any form of background use or high ram usage unless the company has some evil hidden motive.

Windows does not need a built in weather app. Or 95% of the other bloat it ships with. If you want to check the weather do it on an Internet browser.