News Intel Core Ultra 5 225H delivers 14% better single core and 16% improved iGPU performance than Meteor Lake per early benchmarks — the CPU still fal...

JRStern

Distinguished
Mar 20, 2017
170
64
18,660
Could you make this harder to read?
Is all "performance" now a matter of running games better?
I'd like to see some "low power performance", maybe power/heat (aka battery life) running typical web browsing in a somewhat accelerated for test purposes manner.
 
  • Like
Reactions: KyaraM and bolweval

usertests

Distinguished
Mar 8, 2013
928
839
19,760
My guess is Intel wants to get this out before AMD crushes them at CES with their new APU's.
If you're referring to Strix Halo, AMD has to convince laptop OEMs to adopt a big new design and ditch Nvidia graphics, in order to compete in gaming laptops. A tall order. Then AMD has to open up an entirely new "AI workstation" category. Nothing to do with a mid-range "Ultra 5" chip.

I guess Arrow Lake-H graphics is going to be slower than Lunar Lake, which is not ideal. But they could just sell more Lunar Lake for devices with no discrete GPU.
 
  • Like
Reactions: KyaraM

watzupken

Reputable
Mar 16, 2020
1,171
655
6,070
"Arrow Lake ships with Intel's upgraded Xe-LPG+ architecture with the Core Ultra 5 225H sporting the Arc 130T iGPU with 7 Xe cores (112 Xe Vector Engines) clocked at 2.2 GHz. Oddly enough, 7 Xe-LPG+ cores outperform 7 Xe2 cores (Core Ultra 5 228V), at least in OpenCL, a synthetic test."

My take as to why the 7 cores iGPU is faster, Intel wants to show that Lunar Lake is power efficient and made a conscious decision to hard cap the power limit on Lunar Lake. I am pretty sure the iGPU in Lunar Lake is not running near its full potential, which is also reflected in the significant clockspeed regression from Arc Alchemist.
 
Sep 3, 2024
6
1
15
Apple isn't tho? M4 chugs power compared to lunar lake (ironic lol) and isn't topping any charts.
You must be thinking of the M4 Pro ... the base M4 CPU has higher performance at lower power than Lunar Lake - heck the M3 CPU has higher performance at lower power than Lunar Lake, graphics is more equivalent. For the M4 Pro, Apple ditched the M3 Pro's design and went back to the M1/2 chopped die approach which means the CPU consumes far more power having more P-cores and access to greater memory bandwidth. And both M4 and M4 Pro are topping charts in performance and perf/W in their respective categories (thin and lights and 14" laptop chips).
 

Notton

Commendable
Dec 29, 2023
859
754
1,260
The performance looks good for a mobile i5.
If it's cheap enough, I'd love to see these in mini-PCs with less thermal and power constraints compared to laptops.
 

skiwi44

Prominent
Jun 6, 2023
23
7
515
Apple isn't tho? M4 chugs power compared to lunar lake (ironic lol) and isn't topping any charts.
Mmmm... Interesting planet you are living on...

From Notebookcheck....

On performance:

"The M3 generation already offered extremely good single-core performance, leaving the competition from AMD, Intel and Qualcomm in the dust. Only Intel's new desktop processor, the Core Ultra 9 285K, was on a comparable level to the M3 SoCs. Apple has stepped up its game even more with its new M4 processors, further widening the gap massively. The P-cores' maximum clock rate, which is about 500 Mhz higher, results in a performance boost of over 20% compared to the M3 models.

Its advantage over the next fastest mobile chip, the Snapdragon X Elite X1E-84-100—which is currently only available in the Samsung Galaxy Book4 Edge 16—is almost 40%. The current CPUs from AMD and Intel have an even greater disadvantage."

On efficiency:

"When it came to single-core efficiency, the new M4 Pro fell behind most M3 devices, although the processor's additional consumption was easily compensated for by its performance gain. However, since we are talking about very low CPU core consumption values (maximum 6.5 watts), other factors such as the additional RAM may distort the picture somewhat. At this point, we are already looking forward to what the base M4 Pro with 24 GB RAM delivers—as well as the regular M4 with 16 GB RAM. Even so, all other mobile processors (including Lunar Lake, Snapdragon X Elite/Plus, Zen5) were still clearly beaten in terms of efficiency, despite their significantly lower performance."
 

defunctup

BANNED
Feb 1, 2024
15
11
15
Mmmm... Interesting planet you are living on...

From Notebookcheck....

On performance:

"The M3 generation already offered extremely good single-core performance, leaving the competition from AMD, Intel and Qualcomm in the dust. Only Intel's new desktop processor, the Core Ultra 9 285K, was on a comparable level to the M3 SoCs. Apple has stepped up its game even more with its new M4 processors, further widening the gap massively. The P-cores' maximum clock rate, which is about 500 Mhz higher, results in a performance boost of over 20% compared to the M3 models.

Its advantage over the next fastest mobile chip, the Snapdragon X Elite X1E-84-100—which is currently only available in the Samsung Galaxy Book4 Edge 16—is almost 40%. The current CPUs from AMD and Intel have an even greater disadvantage."

On efficiency:

"When it came to single-core efficiency, the new M4 Pro fell behind most M3 devices, although the processor's additional consumption was easily compensated for by its performance gain. However, since we are talking about very low CPU core consumption values (maximum 6.5 watts), other factors such as the additional RAM may distort the picture somewhat. At this point, we are already looking forward to what the base M4 Pro with 24 GB RAM delivers—as well as the regular M4 with 16 GB RAM. Even so, all other mobile processors (including Lunar Lake, Snapdragon X Elite/Plus, Zen5) were still clearly beaten in terms of efficiency, despite their significantly lower performance."
Geek bench isn't reputable benchmark. they come out with a new version/change the benchmarking algorithm every time Apple is ousted from the #1 spot. happens every time, go look at the dates on when Geek bench releases a new version vs when Apple loses the top spot in the version before the benchmark "update".
 

skiwi44

Prominent
Jun 6, 2023
23
7
515
Geek bench isn't reputable benchmark. they come out with a new version/change the benchmarking algorithm every time Apple is ousted from the #1 spot. happens every time, go look at the dates on when Geek bench releases a new version vs when Apple loses the top spot in the version before the benchmark "update".
You're boxing shadows.

Read the reports that are available online, like the one I referenced from Notebookcheck, and you will see that they used far more then Geekbench to test.

These are facts. Get over it.