First serious 14-inch workstation laptop thanks to Strix Halo: HP ZBook Ultra G1a review

[tigolebitties]

Notebookcheck's comment section is slowly getting more ghetto now that gsmarena refugees are coming...

[Psyduck]

Accept once and forever that ARM64 is the superior platform

Hard to get excited when Nvidia is blatantly refusing to serve the consumer market. People are practically begging for a true non apple alternative yet they're too busy milking AI.

[YrpUQQmk]

@ Benjamin Herzig

I had to return two HP ZBook Ultra 14 G1a units (390/64GB/OLED) due to repeated BSOD or restarts without warning.

Did you experience any of these issues during testing?

I am considering buying another unit – wondered if the issues were fixed.

[Benjamin Herzig]

BSODs no, but as detailed in the review, the ZBook restarted during the battery testing without warning. This was with the latest BIOS offered by HP installed.

[It's not ready]

I had random restarts or points where it would freeze also too. In addition to this, certain large downloads for me just kept ending up corrupted. For reference, this was on 385/32GB/FHD-IPS variant.

[Gabrielle]

However, with some newer titles, we had problems, as the games would not start due to insufficient VRAM.

wait..what? "insufficient VRAM"? its supposed to pull from 32GB system RAM to fit whatever memory need the process demand is my understanding. is hp implementation wrong or the game requires more memory than what the system can provide with 32GB pool?

[32 GB of 128 GB RAM? DOA]

Using the expensive Strix Halo, 256-bit, chip, but giving it only 32 GB RAM is major trolling. DOA.

AMD specifically advertises the Strix Halo, a 256-bit chip (4 * 64-bit, also known as quad-channel) for LLM self-hosting (see AMD's slides about it), but with only 32 GB RAM, instead of the up-to 128 GB RAM, that this 8000MT/s, quad-channel, chip supports, it's literally major wasting of silicon and hardware. At least 64 GB RAM would have been ok. Unbelievable.

[M4M]

Now learn how to talk politely on forums (maybe with the help of pills?)

Ignore the iSheep. When Cinebech shows Apple M chips to be the fastest then it's an absolute perfection of an ultimately relevant benchmark to them. As soon as Apple M is not the fastest there anymore the whole benchmark immediately becomes an absolute junk and suddenly only relevant ones are those where Apple M is still leading.

Your m o t h er is a fu ck ing sheep you c o ck su ck er x86shill.
C-Ray2.0 5K full SIMD FMLA load, 285K 300w++ =145s to render, M4pro 63.5W under full SIMD = 143.5s
M4Max absolutely annihilates 285K and 9950X in pure raw performance

[Worgarthe]

Now learn how to talk politely on forums (maybe with the help of pills?)

Ignore the iSheep. When Cinebech shows Apple M chips to be the fastest then it's an absolute perfection of an ultimately relevant benchmark to them. As soon as Apple M is not the fastest there anymore the whole benchmark immediately becomes an absolute junk and suddenly only relevant ones are those where Apple M is still leading.

Your m o t h er is a fu ck ing sheep you c o ck su ck er x86shill.
C-Ray2.0 5K full SIMD FMLA load, 285K 300w++ =145s to render, M4pro 63.5W under full SIMD = 143.5s
M4Max absolutely annihilates 285K and 9950X in pure raw performance

Skill issue. Stay mad.

[M4M]

Now learn how to talk politely on forums (maybe with the help of pills?)

Ignore the iSheep. When Cinebech shows Apple M chips to be the fastest then it's an absolute perfection of an ultimately relevant benchmark to them. As soon as Apple M is not the fastest there anymore the whole benchmark immediately becomes an absolute junk and suddenly only relevant ones are those where Apple M is still leading.

Your m o t h er is a fu ck ing sheep you c o ck su ck er x86shill.
C-Ray2.0 5K full SIMD FMLA load, 285K 300w++ =145s to render, M4pro 63.5W under full SIMD = 143.5s
M4Max absolutely annihilates 285K and 9950X in pure raw performance

Skill issue. Stay mad.

you stay mad f u c k i n g x86shill C-Ray 2.0 5k M4pro is faster than 285K and 9950X. M4pro has more raw power than desktop x86 garbage. Stay mad  r e t a r d CISCshill

[Worgarthe]

you stay mad f u c k i n g x86shill C-Ray 2.0 5k M4pro is faster than 285K and 9950X. M4pro has more raw power than desktop x86 garbage. Stay mad  r e t a r d CISCshill

And exactly this is why you don't do drugs as a child 👆

[M4M]

you stay mad f u c k i n g x86shill C-Ray 2.0 5k M4pro is faster than 285K and 9950X. M4pro has more raw power than desktop x86 garbage. Stay mad  r e t a r d CISCshill

And exactly this is why you don't do drugs as a child 👆

You dont talk to me like that l i tt l e  boy i promise that IRL you'd scream in pain like  the rest of them painfully average arrogant c l o w n s. I am a 1x pro nats winner athlete, you a r e weak p l e b F U CK !!!!

[Foxxii]

I didn't expect it to be this complicated, honestly. There are tons of sites promising to boost your IG game, and yet, most of them feel... off. Either they overpromise or the delivery is pure junk. So I went down a rabbit hole—reviews, Reddit threads, even YouTube rants—and eventually found what felt like a needle in a haystack. If you're serious about growing without burning cash or compromising your account, I highly recommend checking out the best site to buy Instagram followers cheap . It wasn't just the pricing—it was the transparency, the delivery speed, and yeah, the fact that they didn't bombard me with upsells. You want followers that actually look like they exist? That's what I got. Just don't expect magic overnight; you still need to post like a human being. But if you're stuck at 800 followers after months of content, this gives you a real push without feeling sketchy.

The ZBook Ultra G1a 14 really stands out, not just for its price but for packing AMD's Strix Halo and Ryzen AI Max Pro 390 into a sleek 14-inch form—finally, a real contender in the mobile workstation space with serious AI chops and thermal efficiency.
The ZBook Ultra G1a 14 really stands out, not just for its price but for packing AMD's Strix Halo and Ryzen AI Max Pro 390 into a sleek 14-inch form—finally, a real contender in the mobile workstation space with serious AI chops and thermal efficiency.