Intel Could Rename its Semiconductor Nodes to Catch Up with the Industry

[yeeeeman]

Well, given the fact that intel 10nm has the same density and characteristics, presumably, as the TSMC 7nm, then I think it is well deserved to do so. Samsung 5nm is also similar to TSMC 7nm, so everyone is trying to seem better then they actually are and funny enough, Intel is the most conservative right now with their naming. So stop being stupid and haters.

 

[las]

There are actually people out there that do not requires the latest and fastest just for the sake of having it. I still have a old Phenom 2 x6 machine that I use daily, it does exactly what it needs to do and it does it solid as a rock after all these years.

If you can live with Ryzen 1600 performance, you might as well pick up an old Intel combo with BOARD, MEM and CPU for less than a new Ryzen 1600 and get identical performance if not better in gaming.

My old 2600K ran 5 GHz and easily beat Ryzen 1600 @ 4 GHz in 99.9% og games and applications

I'd never buy or recommend Ryzen 1000 series from NEW, probably not 2000 series either, unless it's cheap as dirt on sale, it's 3000 or 5000 series

1600AF is essentially 2600, only 1-2% slower (-200MHz base, -100MHz boost, same Zen+ die). And it seems like we have a bit different understanding of "not much more" for R5 3600. It's $130 vs $260, at least in my area. Basically you save half the price for 20-25% performance penalty. At that sub-$200 price range all you have is OEM R5 3500X, R3 3100/3300X which for some reason is always out of stock, or 3200G. 3500X is a cutdown version of 3600 with HT disabled, and while single-core performance is better, multi-core is within the margin of error comparing to 1600AF. And it's still more expensive.
Just because you don't buy it doesn't mean that the rest of the world doesn't.

Once again, I work with b2b sales and we have tons of huge retailers onboard, Ryzen 1000 is NOT selling like hotcakes, far from it.

What sells like hotcakes, especially in the laptop segment, is Intel, Intel and more Intel. Why? AMD can't deliver. Intel easily can, because they make their chips themself.

I think we ship 1 AMD laptop for every 100 Intel laptop, or so.

Irrelevant. We're talking about process nodes and not microarchitectures.

Irellevant, when someone with a Glofo 12nm calls Intel 14nm inferior? Yeah Suuuuuure.

TSMC is literally the reason why 3000 and especially 5000 performs great. If AMD were stuck with Glofo still, it would be a whole different story, with subpar clockspeeds like all 1000 and 2000 series chips.

 

[8BitZ80]

Volkswagen named their diesels "BlueMotion" to sound environmentally friendly, but it didn't stop any of the pollution.

 

[RedelZaVedno]

LOL... Rebranding will not make Intel's CPUs suck any less. There's only so much marketing department can do. Get your s***t together Intel, TSMC is rolling all over you.

 

[silentbogo]

Well, to be honest, the last time that was true was I think 130nm. Since than, the "nm" is actually not the smallest feature, just a naming.

It's still the case. Only 20 years ago we measured FET size as the smallest feature, and now it's fin or gate size, because on modern CPUs the term "transistor" became kinda vague.
It gives manufacturers some leeway on what to pick as their metric and not end up being called a liar on technicalities )))

My old 2600K ran 5 GHz and easily beat Ryzen 1600 @ 4 GHz in 99.9% og games and applications

In which universe? Even if that were somehow true, it's a dead-end platform, so as anything up to Kaby Lake. Not only do you lose modern features, there's also no prospect for upgrade.

Once again, I work with b2b sales and we have tons of huge retailers onboard, Ryzen 1000 is NOT selling like hotcakes, far from it.

And I'm Lisa Su's nephew.

What sells like hotcakes, especially in the laptop segment, is Intel

No need to retract to laptops. Let's keep going with overclocked 2600K that beats Zen+ in 99.9% of games and applications.

 

[Selaya]

Node density is only half of the story ...
Logic gate size (which is what the X nm denotes after all) matters very much, because that dictates power consumption and thus cooling requirements ...

 

[RH92]

Renaming will not solve technical issues

Sure but this is not what Intel intended . Intels 10nm is very similar to TSMC 7nm physical size , same for their 7nm compared to TSMC 5nm ( if i recall correctly ) . This is what Intel intends to fix with this strategy .

 

[R0H1T]

intel 10nm has the same density and characteristics, presumably, as the TSMC 7nm

Density isn't everything, in fact to get to such high clocks Intel had to sacrifice density on 14nm (with their ++ nodes) & IIRC even for ICL & TGL on 10nm. If you know about Si & how modern x86 designs have evolved you'll also know that "density" is in fact a huge issue for clocking the chips high (enough) & for desktops or HEDT it's not actually better!

Ice Lake Rebadged to 10nm: Why?

So now we have Cannon Lake on ‘10nm’, and Ice Lake originally on 10+ but now rebadged to 10nm, but a different 10nm, with no real explanation as to why. In discussing with a number of peers and analysts in private conversations, the apparent conclusion they have come to is that Intel did not want to admit that its first generation of 10nm product had failed. Ever since then, Intel has attempted to quietly and discreetly shift Cannon Lake under the rug, as if it didn’t exist (it does exist, we did a big review on it, and Crimson Canyon is still for sale today at some of Intel’s biggest partners and major retailers).

Without Intel needing to admit that the first generation had failed, Ice Lake was the true ‘native’ 10nm product that was destined for life in the fast lane for consumers. If that was the case, then the low key presentation at CES 2018 stating it was shipping in 2017 was simply to meet investor targets. Intel never promoted Ice Lake as its first 10nm product, but the fact that the Cannon Lake product wasn't great meant that the company had to try and remove it from people's minds.

On Ice Lake, we studied the Ice Lake design, and we’ve seen lots of notebooks built on it. The fact that Intel called it ‘10th Gen’, and then also released the Comet Lake 14++++ product also called ‘10th Gen’, really ended up confusing the company even more, even in presentations to the press. It was the first time Intel had two products within the same generation of marketing name under different process nodes. It even confused OEM partner marketing teams as well as sales staff.

The problem with calling Ice Lake the new 10nm, is that internally the engineers still called it 10+. As Intel also announced other new products, such as Snow Ridge, or Lakefield, despite these meant to be called under the new 10nm, they would often be cited as ‘10+’ depending on which department of Intel you spoke to.

10NM CHANGES DIRECTION, TWICE

However, I now have an official decoder ring for you, to act as a reference for both users and Intel’s own engineers alike.

AnandTech's Decoder Ring for Intel's 10nm
Product	2020+	First Update	Original
Cannon Lake	-	-	10nm
Ice Lake Ice Lake-SP Lakefield (compute) Snow Ridge Elkhart Lake	10nm	10nm	10+
Tiger Lake SG1 DG1	10nm Superfin	10+	10++
Alder Lake First Xe-HP GPU Sapphire Rapids	10nm Enhanced SuperFin	10++	10+++

For clarity, 10nm Superfin is often abbreviated to 10SF, and 10nm Enhanced Superfin to 10ESF.

Just to be clear, this is also for everyone who thinks that ICL is on the same 10nm as intended for CNL. It is not, ICL is the first 10nm+ product & probably wouldn't make it to the market if it was on the original 10nm node!

 

[Bwaze]

Zen2: 3.8 billion transistors on 74 mm² - 52 MTr/mm²
Zen3: 4.15 billion transistors on 80.6 mm² - 51.4 MTr/mm²
Renoir: 9.8 billion transistors on 156 mm² - 62.8 MTr/mm²

Intel transistors counts are really hard to come by. There was a throwaway comment in some financial call right after Ice Lake release saying it has over 7B transistors.
Ice Lake die that was the only one out at that point is 122.5 mm² which puts density at about 57 MTr/mm².
Same ballpark. Amount of different elements in the die (cache, GPU etc) probably play a bigger role in density of a specific chip at this point than manufacturing process.

Thank you, I haven't found the numbers I knew I read somewhere (estimations, there are no released specs).

Compared to the Intel promo material for 10nm it's quite a bit less dense:

 

[Turmania]

It will only in turn make TSMC change it as well. That is why their 7nm process name was actually inferior to Intel's 10nm. What they did not count on was Intel not getting it right for ages.

 

[mechtech]

Maybe they should. Because alot of fabs are too optimistic about their processnames.

GloFo 12nm was/is far worse than Intel 14nm for example.

Samsung 8nm is not great, but decent - Its probably worse than Intel 10nm

Intels 10nm, is more like TSMC 7nm.
I expect Intel 7nm will be more like TSMC 5nm or even 4nm.

If just performance and watt-usage is good, they can call it whatever they want, could not care less

The next few years are going to be interesting. Can't wait for next gen CPU platforms in 2022+ with DDR5 and PCIe 5.0, not going to get anything before DDR5 has matured (high clocks with decent timings - not going from top-end DDR4 to mediocre DDR5 on launch, thats for sure).

I can wait. I had a teacher long ago who use to say.............."don't wish your life away, it's already too short"

 

[ymbaja]

When all else fails... lie.

 

[HD64G]

There are actually people out there that do not requires the latest and fastest just for the sake of having it. I still have a old Phenom 2 x6 machine that I use daily, it does exactly what it needs to do and it does it solid as a rock after all these years.

Your CPU isn't a Phenom but a Thuban as all 6-core AMD CPUs back then, even if it was sold as a 4-core and you unlocked it.

 

[Solid State Soul ( SSS )]

i5 13400f, powered by Intels advanced ultra micro fine nanometer rising revengeance

 

[AusWolf]

It must be really hard to figure out how many "+"-es to put at the end of node names. May I suggest firing all R&D staff, and employing more PR people to sort this out? They must really get it right for 12th gen.

 

[Vya Domus]

May I suggest firing all R&D staff, and employing more PR people to sort this out?

I am pretty sure that's what they've been doing all along.

 

[thesmokingman]

Haha! That's brilliant! Just lie and call it whatever you want. Then again this is the company that uses waterchillers hidden under tables to lie about their great new cpus.

 

[samum]

This is how they are going to ship "7nm" in 2023.

 

[TheoneandonlyMrK]

To be fair Intel did try to introduce common sense metrics to quantify nodes but no one else bit, clearly, if you can't beat em join em.
Hopefully Alderlake or nextlake shake's things up.

 

[R-T-B]

To be fair Intel did try to introduce common sense metrics to quantify nodes but no one else bit, clearly, if you can't beat em join em.
Hopefully Alderlake or nextlake shake's things up.

This. Basically intel was the only one using common sense node metrics, and it made them sound way more behind than they were (dont get me wrong, they ARE behind, but everyone else is making up numbers that really don't match reality so it's pretty much only fair intel finally can do the same).

 

[Exyvia]

Well there was a calculation done on Intel's 10nm logic to be 100Mtr/mm^2 where the wikichips is based off, obviously thats only the densest part and not the entirety.

Intel 10nm density is 2.7X improved over its 14nm node

New report says that Intel can pack 100.8 million transistors per square mm.

hexus.net

 

[ViperXTR]

Did you post this article a day early?

 

[Mistral]

I guess that's one way to "catch up"...

 

[watzupken]

To be honest, they can name their nodes whatever they want. At the end of the day, people will just look at the density and efficiency of the node. For example, you can have a supposed 5nm, but actually fare no better than a 7nm, that will become apparent in product reviews. Right now, TSMC is leading and thus, not much to compare. If Intel joins the party late, it will be easier to scrutinize.

 

[londiste]

Thank you, I haven't found the numbers I knew I read somewhere (estimations, there are no released specs).

Compared to the Intel promo material for 10nm it's quite a bit less dense:

The marketable numbers are always with the densest variant of the process node. Practically every manufacturing process has separate variants targeted towards density and performance. Performance variant usually has about 1/3 less transistors on the same area. The basic idea is more empty space between cells/transistors to allow higher voltages with lower leakage.

High Density variant is canonically used for mobile and other dies aiming at low power (and lower frequencies), High Performance variant are used - as the name says - high performance dies running at higher voltages, higher frequencies and higher power consumption.

Anandtech's 10nm Cannon Lake article laid out the Intel 10nm spec, scroll down to 'There's more to it' section - 100 MTr/mm² for High Density, 67 MTr/mm² for Ultra High Performance:

Intel's 10nm Cannon Lake and Core i3-8121U Deep Dive Review

www.anandtech.com

This is not Intel-specific. The same applies to TSMC, with N7 official numbers at 96 MTr/mm² and high performance variant in the 65 MTr/mm² range. N7+/N7P/N6 should be 15-20% more dense.

All the official numbers are standard cells - or something more dense like SRAM if manufacturer wants to show even larger density numbers. The actual dies (like CPUs or GPUs) are usually less dense than that due to number of reasons. Logic circuitry does not lend well to being tightly packed together, there are areas like IO that do not scale down that well etc.