Quote from: generic hardware on Today at 16:14:57AI requires these things:

• Memory size to fit a decently capable LLM.

Everything else you have said is ok but this part should be:

Memory size to fit one's used AIs (which need not be LLMs).
[/list]

2200 USD + taxes on top still? for only 32 GB RAM and it's unfortunately soldered, too; a slow iGPU and no dGPU, either.

Does not deserve the "AI" in its product name: Only 32 GB RAM (nothing special) and slow iGPU. AI requires these things:

• Memory size to fit a decently capable LLM.
• Prompt processing: The larger the input, the faster the GPU you'd need, especially for agentic workflows.

Here, the iGPU scores 3614 Points in 2560x1440 Time Spy Graphics. This is ok for a few sentences of input, if you want an instant-ish replay.

• Token generation: The speed of the output generation depends on memory speed (aka memory bandwidth).

The memory score 99980 MB/s, this is not too bad for a 128-bit (99% of all PCs/laptops are 128-bit). But this is nothing special. A Strix Halo system is 256-bit wide and, as such, is twice as fast.

• (The number of CPU threads doesn't matter for running AI (aka inferencing) (4 threads pretty much top out a dual-channel PC)).)

32 GB RAM may not enough for the new SOTA LLM, again, especially in agentic workflows: Qwen3.6-35B-A3B-UD-Q4_K_M:

Quote from: reddit.com/r/LocalLLaMA/comments/1sq94qx/is_anyone_getting_real_coding_work_done_with.. I've come to the conclusion that (1) 32768 is the biggest context I can get away with in an adequately smart model, and (2) it just ain't enough.

And it's not like Windows requires less RAM than MacOS.

64 GB RAM would have to be the bare minimum for this price (and AI-in-product-name), but even 64 GB RAM can't fit the very popular Gpt-Oss-120B.
And why is the RAM running at 8533 MT/s and not 9600 MT/s. Don't get me wrong, 8533 MT/s is better than the usual 5600 MT/s and comes closer to AI's requirements (faster memory -> faster token generation, and since this has no dGPU, also faster prompt processing).

This doesn't replaces a carryable laptop, but: For 2200 you could build a desktop PC that is much more capable: While the RAM would be slower, because a desktop PC's RAM usually runs at 5600 MT/s to 6200 MT/s, the RAM size would be 96 to 128 GB RAM (2x64 GB per stick is possible), you'd have a dedicated GPU for much faster prompt processing, and also faster token generation, because parts of the LLM are offloaded to the GPU's much faster VRAM, and the desktop PC will be repairable, upgradable and run quieter.

After a long and tiring day, everyone looks for a way to unwind. Kolkata Call Girls Service provides a smooth and comfortable experience for those who want to relax. The process feels simple and easy to manage, so you don't have to worry about anything. Whether it's a short meet or a longer plan, everything feels natural. It's about enjoying moments that feel calm and stress-free.

I got this notebook on sale and my first impression is that the build quality is very solid.  It feels very good to hold in hands, like a really nice all-aluminium laptop.  There is a lot of crapware installed by HP, but it can be easily removed or disabled.  Fan noise was minimal while installing the gazillion of Windows updates and driver updates.  I'm quite satisfied with it so far.

I honestly don't understand how you measure power consumption.

From your test method article:
"Power consumption
Idle: power consumption during operation in idle mode.
Minimum: All additional radios deactivated (WLAN, Bluetooth, etc.), minimum display brightness, and activated power-saving measures (Windows power profile: Best power efficiency)"

This here is 6.5W.

Then:
"Battery life:
WLAN operation: Achievable runtime in WLAN operation with adjusted display brightness (~150 cd/m²) and activated power-saving measures (power profile Balanced or similar). For the run time test, an automatic script (updated on 05.03.2015 to v1.3 with HTML5, javascript, and no flash) is executed, which calls up a mix of Internet pages that change constantly every 30s."

This here is 14h12m, which on a 74.5Wh battery is 5.2W, 20% less than your power consumption idle test just a paragraph earlier!
The Battery Idle test, with supposedly the same settings as the power consumption part above, gives 3.5W, nearly half of power consumption data!

I understand there are losses between wall power and the actual components' power draw, but this is exceptionally different. Or the laptop is very much throttled under battery, but under idle it shouldn't matter that much, if at all.

2500$ and they could not give this with basic 64GB ram!!!

Great CPU, but I can't wait till they release one with the "AMD Ryzen AI 9 HX PRO 375 X3D Max+ XTX" CPU

Wow Hp almost copied Macbook 14 one to one... keyboard, chassis, and the screen it's nearly identical

@Russel: Elitebooks have almost nothing in common with HP's regular consumer line when it comes to structural rigidity of build quality.

That's like me commenting under a high end ThinkPad review about how I was once traumatized by a budget ideapad chromebook hinges breaking.

HP laptops have one major issue, the screen just comes off i.e detaches off the hinges....
You can just google 'hp laptop screen coming off the hinges' to see how many results you'd get.
And this is something they're notorious for.
There are other issues like increased flex and screen wobble.
So. I hope we see strix halo in a Thinkpad.