this post was submitted on 29 Sep 2025
164 points (85.7% liked)

Technology

75606 readers
2238 users here now

This is a most excellent place for technology news and articles.

Our Rules

  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots

founded 2 years ago
MODERATORS
L3s@lemmy.world
enu@lemmy.world
technopagan@lemmy.world
L4s@lemmy.world
L3s@hackingne.ws
L4s@hackingne.ws
164
Snapdragon X2 Elite Extreme crushes Apple M4, Intel, and AMD in new benchmarks (www.windowscentral.com)
submitted 1 day ago by Delta_V@lemmy.world to c/technology@lemmy.world
64 comments fedilink hide all child comments
 

...In Geekbench 6.5 single-core, the X2 Elite Extreme posts a score of 4,080, edging out Apple’s M4 (3,872) and leaving AMD’s Ryzen AI 9 HX 370 (2,881) and Intel’s Core Ultra 9 288V (2,919) far behind...

...The multi-core story is even more dramatic. With a Geekbench 6.5 multi-core score of 23,491, the X2 Elite Extreme nearly doubles the Intel Core Ultra 9 185H (11,386) and comfortably outpaces Apple’s M4 (15,146) and AMD’s Ryzen AI 9 370 (15,443)...

...This isn’t just a speed play — Qualcomm is betting that its ARM-based design can deliver desktop-class performance at mobile-class power draw, enabling thin, fanless designs or ultra-light laptops with battery life measured in days, not hours.

One of the more intriguing aspects of the Snapdragon X2 Elite Extreme is its memory‑in‑package design, a departure from the off‑package RAM used in other X2 Elite variants. Qualcomm is using a System‑in‑Package (SiP) approach here, integrating the RAM directly alongside the CPU, GPU, and NPU on the same substrate.

This proximity slashes latency and boosts bandwidth — up to 228 GB/s compared to 152 GB/s on the off‑package models — while also enabling a unified memory architecture similar in concept to Apple’s M‑series chips, where CPU and GPU share the same pool for faster, more efficient data access...

... the company notes the "first half" of 2026 for the new Snapdragon X2 Elite and Snapdragon X2 Elite Extreme...

you are viewing a single comment's thread
view the rest of the comments
[–] Alphane_Moon@lemmy.world 192 points 1 day ago* (last edited 1 day ago) (5 children)

Keep in mind the original X Elite benchmarks were never replicated in real world devices (not even close).

They used a desktop style device (with intense cooling that is not possible with laptops) and "developed solely for benchmarking" version of Linux (to this day X Elite runs like shit in Linux).

This is almost certainly a premeditated attempt at "legal false advertising".

Mark my words, you'll never see 4,000 points in GB6 ST on any real products.

[–] boonhet@sopuli.xyz 65 points 23 hours ago* (last edited 23 hours ago) (3 children)

They also used the base M4, not M4 Pro or Max

[–] Reverendender@sh.itjust.works 3 points 8 hours ago

Now this all makes sense

[–] Ugurcan@lemmy.world 6 points 10 hours ago (1 children)
[–] boonhet@sopuli.xyz 3 points 10 hours ago

lol that’s just the cherry on the whole apple pie.

[–] CmdrShepard49@sh.itjust.works 36 points 22 hours ago (1 children)

Seems like they're also using two different Intel chips in their testing for some reason.

[–] circuitfarmer@lemmy.sdf.org 24 points 18 hours ago

I'll take cherrypicking for $500, Alex

[–] itztalal@lemmings.world 5 points 15 hours ago

desktop-class performance at mobile-class power draw

This made my bullshit detector go haywire.

[–] Zak@lemmy.world 23 points 22 hours ago (1 children)

I imagine things would be much closer if they put a giant heatsink that Ryzen 370 they're comparing and ran it at its 54W configurable TDP instead of the default 28W.

[–] pycorax@sh.itjust.works 6 points 16 hours ago

Shouldn't they also be comparing it to Strix Halo instead?

[–] tal@olio.cafe 12 points 22 hours ago

Ah. Thanks for the context.

Well, after they have product out, third parties will benchmark them, and we'll see how they actually stack up.

[–] SharkAttak@kbin.melroy.org 7 points 22 hours ago

I saw someone liquid cool an Arduino to push it to the max, but you couldn't declare it to be a regular benchmark...