Intel Meteor Lake Analysis - Core Ultra 7 155H only convinces with GPU performance

[Redaktion]

Intel drops the familiar naming scheme for the new Meteor Lake generation and the latest mobile processors are called Core Ultra from now on. Intel promises a big performance increase for the new Arc GPU as well as increased efficiency, which we tested.

https://www.notebookcheck.net/Intel-Meteor-Lake-Analysis-Core-Ultra-7-155H-only-convinces-with-GPU-performance.783320.0.html

[NikoB]

What's the problem with Zen4 Phoenix? It is nowhere to be found, not even in the USA! Sane models for 7840U! And the same will happen with the virtual Zen5, sold only on paper.

We look at newegg, for example, at the end of 2023 before Christmas (sales season, damn it!!!) - and what do we see there? Several pathetic models from Lenovo and one from Acer!
Most have unsoldered memory, a measly 16GB. Most often there is one USB40 port, instead of 2 built into Zen4 Phoenix. Compare with Intel versions - there are usually 2 ports.

There is not a single business model even with 2.5k, not to mention 18" 16:10 (you can't even dream about 4k@120Hz, even in 16" with IPS - there is only a rotten AMOLED with 4k and vile low-frequency PWM), purely for the office with a normal keyboard, 2 memory slots, 2 M.2, and a bunch of other ports, including a pair of USB40 located symmetrically on the left and right, as well as usb-a Gen2 ports on the left and right. With rear power, hdmi 2.1, rj45 2.5Gbps+

Where are these models? Without discrete adapter?

Or with lpddr5 7500 (shame on Intel, it couldn't even provide 7500 in the new series!!!) but at least 32/64GB options out of the box to accelerate 780M.

Give people mass models 14, 16, 18" with 16:10 4k@120Hz with fast response and contrast from 1500:1. With symmetrical usb-a, usb40 ports - on the left and right closer to the rear edge of the laptop and in the back there is an rj45 2.5Gbps+ port , hdmi 2.1 (48Gbps) and a power port (usb-c or round) if it is strictly backwards, I don't care what type of plug it is.

With audio outputs on the left and right also symmetrically, plus an optical spdif output at the rear.

With a normal full-fledged keyboard with a full numpad in 16.18" models with a large key travel of at least 1.8 mm, with elastic tactile feedback and which can withstand intense fast touch typing for at least 5 years.

With coolers configured by a proprietary utility according to their activation threshold and speed (within safe limits), as the owner wants.

All! And sell, sell, sell! But there is NOT a single such model on the market! Amazing!!!

[NikoB]

And a request to all manufacturers! Damn it - bring back eSATA ports to laptops!!! They are extremely important for proper full testing of new HDDs in exclusive mode under DOS in programs like HDAT2! Testing from under Windows is not suitable through, especially through a USB adapter!

[paviko]

If nothing changes with retail units (low hope), then it's the biggest disaster for Intel for decade.

[dick james]

Clearly Something wrong with the Acer Swift Go 14

[PHVM_BR]

All available reviews of the ZenBook 14 with Ultra 7 state that the power limits in performance mode are, in fact, 28/50W.

Only notebookcheck arrived at these 33/50W.

I believe it is an error that must be corrected...

[LostInSpace]

If nothing changes with retail units (low hope), then it's the biggest disaster for Intel for decade.

No, they have a 'Hindenburg Disaster' every couple of years now. For example: Running Ad's mocking Apple, they publicly saying: "We want to be Apples Fab partner."

Yeah, that is not the way to make that happen.  ;-)

[Surreptitious]

Why are you comparing a Meteor Lake with a 45W/50W PL to a Ryzen 7840U with much lower power limits enforced (~15W by my estimate of your ST results)?  Also given your power consumption charts, (showing the zenbook producing higher scores than the Acer swift go at lower power consumption) why would you draw conclusions based on what appears to be a bad laptop design?

[hs4]

The power performance curve is usually convex upward, so this 155H measurement, which is convex downward, is quite odd.

[brian seabury]

which witcher 3 did you use. are you using the new witcher 3. I wanted to test the performance of my 1165g7 versus the new gpu in the witcher 3 but I don't know which witcher 3 was mentioned. I use some optimizations on my cpu and gpu to improve performance. im curious if it gets similar fps to the new ultra. however it should be noted that they probably should have tested with xess since that gets to run in hardware on the newer processor which is a big advantage.

[Laptops are dead]

@NikoB:

Because nobody cares about laptops anymore, seriously. You want a 7840U device in stock / available everywhere? Then get a Z1 Extreme handheld. Need additional ports / expansion / peripheral connectivity? Buy a docking station. Hooking up to an external 4k 120hz OLED TV would solve your display issues as well (Unlike mobile displays, they don't suffer from same low freq. PWM) And the type of people that need > 64GB ram are mostly into tech, 95% of these people just get MBP's anyway.

Take the hint, seriously. If AMD cared about laptop users they'd not be selling all their chips to handhelds. If OEMs care more, well, they'd be releasing better devices with good pricing, availability and supply. But nobody gives a sheet, so why should you?

[brian seabury]

I just tested my 1165g7. it seems like the 1.50 ghz graphics core clock on some 13th gen processor makes a big difference. in 1080p with ultra settings with hbao+ in the witcher 3 directx12 enhanced edition I get 11 fps without xess upscaling. at 900p I get 15 fps. and turning some settings down like level of detail and shadow quality but keeping screen space reflections at high. I get 17 fps. I also optimized the process with process lasso and saw about a one fps difference but it was usually at 17 fps. that being said the 1165g7 so far is better at running directx11 games anyway except in the future there wont be many of them so this is still a good benchmark for determining the performance of the igpu. I havent tested intel xess upscaling yet in the witcher 3.

[brian seabury]

when I rebooted the game I got 20 fps. then I enabled intel xess ultra quality on the 1165g7. with optimizations with throttlestop and process lasso. I got about 24 fps. still not sure if I was using directx11 mode. last time I booted I used directx12 mode. but intel xess wouldnt be available in directx11 mode I don't think so it was a pretty good result.

[brian seabury]

on an 1165g7 with intel xess ultra quality at 900p with shadows low foliage density low water low screen space reflections high level of detail low ambient occlusion off I get 24 fps. for some reason I was only getting 16 fps during one run through it looks like some settings don't apply correctly until after restarting the game. not entirely sure with ultra settings and hbao+ ambient occlusion I was getting 11 fps at 1080p. at 900p I was getting 15-16 fps. 900p isnt really much of a sacrifice and its essentially the resolution that the memory bandwidth of the igpu maxes out at in most games. at 900p with its superior xess support I bet the new ultra 155h processor gets about 40 fps but obviously it was supposed to be an apples to apples benchmark of the capabilities of the card which is fine. still at least the new igpu should be able to run city skylines 2 since its essentially just discrete graphics in an igpu form factor.

[john1]

can you please test single core power efficiency for all three core types? (set process affinity)