Recent content by Laurent06

I failed to notice one was CPU and the other was DirectML. I'll blame the lack of coffee :o It's strange that the CPU is faster than the supposedly accelerated DirectML. This make me think it isn't using the NPU.

I'm not familiar at all with that benchmark. But I'd say that Oryon does more poorly on I8 than on F16/F32: https://browser.geekbench.com/ml/v0/inference/compare/367614?baseline=369007 It gets destroyed on Text but wins on Image Segmentation. Edit: I had picked the one with 1900 as base...

Read here for instance: https://www.androidauthority.com/snapdragon-x-elite-benchmarks-3380426/ Anyway as usual I pick that with a grain of salt and will wait for consumer devices.

If I understood correctly, 80W is the max power of a full device when the max allowed frequency is 4 GHz. EDIT: more information about the various powers here: https://www.anandtech.com/show/21112/qualcomm-snapdragon-x-elite-performance-preview-a-first-look-at-whats-to-come

Elite X supports DX12 according to Qualcomm.

Never forget that Nordic people tend to follow advises they are given. French people do the contrary of what they're advised to unless you put a law and a fine in place.

Yes. But for all the rest, including Windows libraries re-implementation, that's Wine. At least that's how I understand it :) I guess they also had to make changes in Rosetta2 so that it can cooperate with this porting toolkit.

GPT is powered by Wine: https://www.codeweavers.com/blog/mjohnson/2023/6/6/wine-comes-to-macos-apple-s-game-porting-toolkit-powered-by-crossover-source-code

And that 1.50 euro ticket can be reused for 74 minutes on busses and tramways. On my side, I can’t live without a car: I’m a few km away from the closest town and 25 km from work. In fact I have two cars: one for work and other duties; and the other one to take rides on mountain roads. So I...

That's a pizza gate! QAnon mobs never disappoint.

That'd be the best way to train Artificial Idiots.

You missed this it seems:

Here is an interesting article about branch prediction (2018, perhaps a bit dated): https://hal.inria.fr/hal-01799442/document Back then the guy who was the most vocal about "pure" Perceptron was Jimenez (IIRC from Univ of Austin), not a Finn guy. Now it seems Perceptron and TAGE are used in...

I have a P1 as work laptop. I don't care about battery life as it's plugged all day long. But that thing is horribly noisy; fans are always on and start spinning fast as soon as one or two cores are used. I regret my decision to pick that Thinkpad rather than waiting for Apple Silicon based Mac...

I don't buy it. TAGE is the best (it won multiple times the Championship Branch Prediction) and it's French :) And the inventor spent some time at Intel.