Hacker Newsnew | past | comments | ask | show | jobs | submit | jokowueu's commentslogin

Metabolic dysfunction is the root of many diseases which addiction is one of them .

Where were you two weeks ago! Gonna try it


Rather, where were they three years ago?


Iraq flash backs , they were sure very happy to greet their liberators , it's amazing to see propaganda's effects working in action


How much are NPUs more efficient than GPUs ? What are the limitations , it seems it will have support for deepseek R1 soon


A decent chunk of AI computation is the ability to do matrix multiplication fast. Part of that is reducing the amount of data transferred to and from the matrix multiplication hardware on the NPU and GPU; memory bandwidth is a significant bottleneck. The article is highlighting 4-bit format use.

GPUs are an evolving target. New GPUs have tensor cores and support all kinds of interesting numeric formats, older GPUs don't support any of the formats that AI workloads are using today (e.g. BF16, int4, all the various smaller FP types).

NPU will be more efficient because it is much less general an GPU and doesn't have any gates for graphics. However, it is also fairly restricted. Cloud hardware is orders of magnitude faster (due to much higher compute resources I/O bandwidth), e.g. https://cloud.google.com/tpu/docs/v6e.


NPU also has no more memory bandwidth than CPU, but then the GPU on these machines doesnt either.


Agree on NPU vs CPU memory bandwidth, but not sure about characterizing the GPU that way. GDDR is usually faster than DDR of the same generation, and on higher end graphics cards has a width bus width. A few GPUs have HBM and pretty much all datacenter ML accelerators (NVidia B200 / H100 / A100, Google TPU, etc). The PCIe bus between the host memory and GPU memory is a bottleneck for intensive workloads.

To perform a multiplication on CPU, even SIMD, that values have to fetched and converted to a form the CPU has multipliers for. This means smaller numeric types penalised. For a 128-bit memory bus, an NPU can fetch 32 4-bit values per transfer; the best case for a CPU is 16 8-bit values.

Details are scant on Microsoft's NPU, but it probably has many parallel multipliers; either in the form of tensor cores or a systolic array. The effective number of matmul's per second (or per memory operation) is higher.


Yeah standalone GPUs do indeed have more bandwidth, but most of these Copilot PCs that have NPUs just have shared memory for everything I think.

fetching 16 8 bit values vs 32 4 bit values is the same, this is the form they are stored in memory. Doing some unpacking into more registers and back is more or less free anyway, if you are memory bandwidth bound. Largely on these lower end machines everything is memory bound not compute bound, although the CPUs cant often use the full memory bandwidth in some systems (eg the Macs) but the GPU can.


Yes, agree. Probably the main thing is the NPU is just a dedicated unit without the generality / complexity of a CPU and so able to crunch matmuls more efficiently.


Fantastic , thanks


What browser are you using ? Its laggy mess on mine


Firefox with NoScript. Javascript is slow indeed.


Much less of an issue with XR and after a few months you won't feel any discomfort

Most annoying thing for me was the metal taste in my mouth for a few months

Had to take it for my reactive hypoglycemia


Standardizing ink cartridges would be amazing


Unlike smart watches where integrating the watch and phone probably has lots of opportunity for innovation left, I agree there isn't much value in proprietary printer inks and standardization probably has more consumer benefit.


Openvpn is trivial to see , what you need to do is wrap it in stunnel


Yeah I suspect they know they can never block everything but if they can block 98% of "casual users" they've probably reached their goal. They will just put out propaganda that the other 2% technically apt people who get around it are conspiracy nuts, western civ sympathizers, traitors to mother russia, etc.


It's only good for anime where the fps is extremely low


Counterpoint: this YouTube rant by an animation person called Noodle is a pretty good overview of why frame interpolation sucks. https://www.youtube.com/watch?v=_KRb_qV9P4g

Basically, low FPS can be a stylistic choice, and making up new frames at playback time often completely butchers some of the nuances present in good animation.


Low FPS can be a stylistic choice, but as a member of the audience, I tend to disagree with that choice.

Perhaps it depends on the quality of the execution, but there are shows where I wished I had frame interpolation.


Personally I'm dubious. There may be times when the low framerate is a stylistic choice, but the vast majority of the time it's purely a budget thing.


Which the style is then centered around.

If it had looked better with frame interpolation the studio could have baked it in, before release.


If you are a good director you can make the most of that low budget. Look at the first episodes of Scum's wish (https://www.imdb.com/title/tt6197170/) if you want a good example.


Animation is the worst use case for motion interpolation because the frames are individually drawn and timed by the animators to achieve a particular look and feel.


*Used to be. Now they're rendered


A lot of animation is still drawn. Some CG anime is also combined with 2d drawn elements (like in SpiderMan into the Spiderverse).


The luddite hacker in me wants to try interpolation on early Ghibli movies.


Do it! I personally love the old Looney Toons, it makes them so smooth it's almost unreal (but in a good way).


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: