News Meteor Lake laptop chip performs like Ryzen 7 desktop PC chip in leaked benchmark - next-gen 16-core Ultra 9 185H flaunts 5.1 GHz at 45W

[Admin]

Images of Samsung's upcoming Galaxy Book 4 show off its Meteor Lake processor and some technical specs.

Meteor Lake laptop chip performs like Ryzen 7 desktop PC chip in leaked benchmark - next-gen 16-core Ultra 9 185H flaunts 5.1 GHz at 45W : Read more

 

[-Fran-]

Funny, because my 5900HX also performs like a 5700G on the desktop.

Regards.

 

[Роман]

What I see on image on top of article?
Why so much thermal?
And what actually covered chip?

 

[ingtar33]

that sounds like a bugged result, because other leaks directly contradict these results badly, with a much wider group of tests.

 

[Deleted member 2731765]

Steppings are organized by letter (a distinct version) and divided by number. If you're familiar with software updates for games or GPU drivers, C0 means version 3.0.

A stepping of C0 implies a somewhat longer development process than usual. For comparison, Ryzen 7000 Raphael chips use the B2 stepping, Ryzen 5000 Vermeer CPUs initially used the B0 stepping but later moved to B2, and Intel's Raptor Lake processors are all on B0. Alder Lake CPUs had a stepping of C0, much like Meteor Lake, and there are parallels between the two chips.

You might need to check some of the data on the CPU’s “stepping” and the revision of the silicon. Apart from being in a longer development process, a letter change could also signify a change to both the base layer mask revision and metal layers, whereas a change in the number could indicate a change in the metal layer mask revision only, in Intel’s terminology.

But for consumers and end-users the stepping/revision number is not THAT much of an issue while purchasing new silicon.

But in any case, not all Ryzen 7000 Raphael chips use the B2 stepping. Check the database and spec sheets again. Also, the statement, “Intel's Raptor Lake/RPL processors are all on B0” is also not accurate.

Some of the SKUs in the final RPL lineup sport the previous Alder Lake C0 silicon. Few processors also use the H0 die.

The same is true for Alder Lake-S/ADL as well. Alder Lake CPUs didn’t just have a stepping of C0 either. For example, the final Retail version of the Intel Alder Lake-S 6-core i5-12400 CPU uses the C0 silicon, even though there is one more variants in the market, sporting both the H0 silicon die.

There are more SKUs as well.

Basically for the ADL lineup, two silicon stepping variations were used, C0 and H0, which showcased both the 8+8 full core implementation (C0), with the slimmer variation lacking the smaller Efficient cores (H0 die).

Btw, the 10-core Alder Lake-K SKU sports the C0 silicon stepping (i5-12600K), and one using a lesser 6-core non-K SKU is based on the H0 silicon stepping (i5-12600).

 

[DavidC1]

You might need to check some of the data on the CPU’s “stepping” and the revision of the silicon.

The author's assumptions are correct in this case.

Raptorlake had a very fast development time with little issues. Meteorlake went through a lot of troubles. So is Sapphire Rapids. Granite Rapids and Sierra Forest(to a lesser degree) has been in development for quite some time too, including move from Intel 4 to 3.

Emerald Rapids is coming earlier despite the fact that it's been developed quite a bit later than either server chips.

Sapphire Rapids development was at a time when Intel fired the entire server validation team at one point.

 

[JayNor]

Nope, all indications are that Intel isn't repeating anything like the 10nm problems in Intel-4.

Alder lake was Intel's 4th consumer chip on 10nm, btw.

If the process were to blame, then we'd be having a Cannon Lake moment and we wouldn't be seeing stories on Samsung Galaxy Book 4.

 

[cyrusfox]

Multithread beats most 8 core/16 thread cpuz results, but the real win is how well single core is performing
Its score of 767 ties the 7950x, and trounces any AMD mobile chip. Only a 4.5% uplift over the 12900H, so meteor still bringing Gen on Gen improvements, although a bit underwhelming.

Need to see real world battery usage + constrained performance test to see how it stacks against Apple M1/M2/M3