Intel Previews Its Alder Lake Chip, Promises Hybrid CPUs for Desktops and Laptops

Intel has spent much of 2021 announcing plans for its future: a new IDM 2.0 strategy, new naming schemes for its process nodes, and new desktop GPUs. At Intel's Architecture Day 2021, we finally got a preview of how some of those changes are coming together in new chips, starting with the upcoming Alder Lake lineup later this year. From a report: As the company has been teasing since last year's Architecture Day, Alder Lake will feature Intel's latest hybrid architecture: instead of simply offering the next generation of powerful Intel CPU cores, it'll offer a mix of both performance and efficiency x86 cores, both of which Intel previewed as part of its announcements. Additionally, Alder Lake will be the first chip released on Intel's newly renamed Intel 7 technology node (not to be confused with Intel 4, which was previously known as Intel's delayed 7nm node, and will be available to consumers sometime in 2023 under the codename "Meteor Lake"). Intel 7 still uses similar technology to the company's current 10nm tech, instead of the bigger leap in manufacturing processes planned for Intel 4.

The new x86 performance core -- codenamed "Golden Cove" -- is the successor to the Willow Cove cores that are currently found in the company's 11th Gen Tiger Lake processors. Intel claims that it's the most powerful CPU core its ever built, but the company only offered a comparison to its Cypress Cove cores (the version of its 10nm architecture that Intel ported to its 14nm process), not the more advanced Willow Cove cores. Meanwhile, the company's new x86 Efficient core (codenamed "Gracemont") aims to be "the world's most efficient x86 CPU core" while still offering higher IPC than the company's Skylake chips. Intel claims that for single-thread cases, one of its new efficient cores hits 40 percent more performance at the same power (or similar performance while using 40 percent of the power) of a Skylake core, improvements that double when comparing four Efficient cores running four threads to two Skylake cores running four threads.

Have they announced how many security holes yet?

[nagora]

Or are they keeping that as a surprise again?

Re:

[iggymanz]

All architectures have those, not just x86-64 and certainly not just Intel

Re:

[BAReFO0t]

Not to that degree.
Not that blatantly.
Nice try.

Re:

[iggymanz]

You mean you're hoping, because security holes are of two types, known and unknown.

Nice hope and optimism there

Re:

[ArchieBunker]

So in other words, yes.

Re:

[gweihir]

Oh, but in Intel they are _easier_ to use and Intel has a few that nobody else has! You know, really distinguishing features only Intel offers!

Re:

[iggymanz]

Nonsense, AMD has unique ones too. I guess you guys follow hoopla and not CVE

Re:

[gweihir]

Every one has unique ones. The question is how easy are they to exploit. AMD is much, much superior to Intel in that regard, because they were careful. For example, only Intel has Meltdown and it shows nicely how badly they have messed up. Because they were warned years before about that risk.

Also, CVE? That does not give you good information. Too many people file CVEs just to make their discoveries look more important. What about reading the actual research papers?

Re:

[iggymanz]

No, research papers are where people are trying to make a name for themselves. CVE get evaluated and ranked for risk by expert.

You're wrong about Meltdown, various ARM and Power chips also have it.

Not even known at present time how badly each chip by AMD and Intel is affected by all the Spectre variants. And just last April we had new vulnerability that can break through mitigations for Spectre on various AMD and Intel.

Re:

[anonymous scaredycat]

No but the holes will perform better(quicker to exploit) or be more efficient(easier to exploit).

Re:

[gweihir]

They will have different security holes on the small and the large cores! You, know, everything to make it easier for the attackers!

Re:

[DarkRookie2]

They added the 64 part back in 2000. What else more do you want.

Re:

[andydread]

Well it was AMD that added that part. Intel simply used the AMD innovation and renamed it. Just like they are renaming 10nm to give the impression that it's 7nm(intel 7)

Re:

[gweihir]

No. AMD added the 64 bit part, that is why it is called the AMD64 bit architecture. Intel tried and failed. Intel licenses it from AMD. (Assuming "they" meant Intel. Otherwise, my apologies.)

Re:

[CaptainLugnuts]

There's been lots of innovation, just the big stuff all failed.

You've got all the SIMD stuff that has mostly worked out, then Itanium and Larrabee which were both huge failures.

Re:

[lamer01]

If you read Jim's interview, AMD-64 was brand new, the tossed out a lot of compatibility modes. They have since added some crud but his point was that when an architecture is brand new, it is free of crud but with each iteration more and more crud accumulates as compatibility becomes a significant variable in any redesign.

Re:

[The New Guy 2.0]

Well, x64 was rather quickly accepted by Microsoft... anybody having consumer problems with AMD?

Re:

[Tough Love]

There hasn't been a major advance since 1986... that's what x86 means.

Like to listen to yourself spout do you? The "6" in 8086 stands for "16", it was the 16 bit variant of the 8088. The rest of the number stands for 8080, which doesn't really mean anything except for being a variant of 8008, which was the 8 bit successor to the 4004, Intel's 4 bit processor, the world's first mass production microprocessor. No year numbers anywhere to be seen.

Need all that power

[DarkRookie2]

For all the tabs in Chrome people do not close.

Hybrid CPU's

[rossdee]

Run on electricity and gasoline.

TL;DR: "Intel claims ..."

[BAReFO0t]

Also note what Intel didn't claim.

Re:

[UnknowingFool]

I found this interesting: "Intel claims that for single-thread cases, one of its new efficient cores hits 40 percent more performance at the same power (or similar performance while using 40 percent of the power) of a Skylake core". That is an odd comparison considering Skylake was launched in 2015 (6th gen?) Why not compare it to something newer like Comet Lake or Tiger Lake (11th Gen)

Re:

[Calinous]

Because - usually in the industry - you replace a computer that's three years old, so any comparison would be made against the two-three generations ago.
There are people for which a new generation offers enough extra performance to make upgrading worth (financially) - they could be either performance-bound (and a new processor could bring extra cores, IPC, memory support, ...) or "battery-life bound" - when a new laptop offers enough battery life to move to the next "battery life tier" - like, for example,

a heavenly name: "Alder Lake"

[swell]

I did an informal poll at our house and discovered that most of me, in face 100% of the people here today, would rather buy a computer with an 'Alder Lake' chip than a 'Lucifer Excrement' chip. We agree that Intel's marketing department is right on with that catchy name.

But isn't this cute naming ritual a bit tired? For about a century we've had automobiles, soft drinks, breakfast cereals and sports teams with supposedly sexy names. A product like a chip, whose life expectancy on the market is 9 months doesn't need that. It needs a name that clearly shows it's relationship with predecessors and siblings and indicates its relative potential. Probably some awkward to pronounce combination of letters and numbers, but a name that has meaning for the people who buy it.

Re: a heavenly name: "Alder Lake"

[Osgeld]

One sounds cold and clammy the other sounds like hot shit...

Re:

[ahoffer0]

Photo of Intel engineers fishing for higher performance in Alder Lake

https://www.eugenecascadescoas... [eugenecascadescoast.org]

PS: Note the ironic lack of Alders

PPS: Unless it is supposed to be _this_ Alder Lake
https://www.bestfishinginameri... [bestfishinginamerica.com]

Re:

[dryeo]

Photo of Intel engineers fishing for higher performance in Alder Lake

https://www.eugenecascadescoas... [eugenecascadescoast.org]

PS: Note the ironic lack of Alders

Looks like an Alder behind the picnic table, though hard to say based on the bark

Re:

[thegarbz]

A product like a chip, whose life expectancy on the market is 9 months doesn't need that.

If you were talking about a product with a 9 month life expectancy I'd agree. But in reality the product generation runs for about 2-3 years and continue to sell for long after that as well.

Question

[Artem S. Tashkinov]

Can anyone explain why Intel needs big and small (heterogeneous) cores for its future desktop and mobiles CPUs?

It makes little sense to me as their current mobile CPUs can draw less than 1W when being relatively idle and it's trivial to manage their power consumption by limiting their frequency.

With different cores, specially if they support different instruction sets, it all becomes quite difficult to manage.

Re:

[ChunderDownunder]

They're not different instruction sets, it's still x86. They're just mixing and matching their two chip streams; high performance Core with energy efficient Atom.

ARM chips in phones already include both big and little designs to prolong battery life.

https://en.wikipedia.org/wiki/... [wikipedia.org]

Re:

[Agripa]

Can anyone explain why Intel needs big and small (heterogeneous) cores for its future desktop and mobiles CPUs?

One core is optimized for single thread performance and the other is optimized for throughput performance.

Put a "intel atom inside" sticker...

[unique_parrot]

...on your notebook, for protection against stealing. Won't get stolen, unless somebody needs a fancy doorstop.

Software?

[The Wily Coyote]

It's all nice and techy to have different kinds of cores in your box. But I have a question (partly born from my ignorance on the subject) -- how good will general purpose software be at using these different kinds of cores? Is this gonna be like vector instructions, where some special purpose code benefited a lot, but the rest of us who use normal software didn't benefit much at all?