Discussion Intel current and future Lakes & Rapids thread

[A///]

Just for fun, based on the discussion back over at the Zen 5 thread, I was interested in how GLC effected clocks in relation to WLC.

that reminds me. when you lot were debating whether peter or paul, whatever fatty's name is was making up claims of the 20-28% lead over zen 4 all I could think about was some of the claims sound like something I guessed a while back about zen 4, I remember the weekend I made the claims of 25-35% IPC increase but I was also drinking all weekend long. it isn't the first time we're seeing websites or youtube "personalities" aka deranged people who make up lies, take estimates or guesses from here and spin them as fact. in fact wasn't one of your posts lifted by hasan's team of daredevil morons months back?

 

[Hitman928]

RPL took 2.5 years of dev time. ADL took 3 years.
MTL taking 2-3 years of dev time is essentially the same time it took for RPL from development to launch. Using RPL timeline as a template for this debate is perfectly acceptable.
Also, MTL has been confirmed to launch to Q3 this year. MTL's tape in, tape out, and power on dates have all been announced by Intel themselves. Its timeline fits. I'm not making this stuff up, literally search up the individual events and dates if you want to yourself. Sure, chiplets in theory might shrink down design timelines, but with MTL, it appears to have taken a similar amount of time as RPL/ADL.
MTL literally was not design ready for a late 2022/early 2023 launch date. Idk what to tell you. Pat also mentioned that MTL will PRQ in Q3, so that would mean the time between PRQ and power on would be still ~15 months. That means MTL spent around the same amount of time in post-si validation as RPL did (slightly more actually). Intel has been announcing engineering milestones, it all lines up.
Plus Exist50 (god bless his banned account lol) has been complaining about MTL's design woes for a while on this thread. It apparently wasn't smooth sailing.

Gelsinger and the CFO both said in the recent earnings call that MTL was launching/PRQ in Q4. Then, like 4 days later Gelsinger said it was Q3. Seems like they made a lot of progress in those 4 or so days. I think it’s likely we see essentially a paper launch in Q3 with actual ramp starting Q4 based upon this but who knows, we’ll see soon enough.

 

[A///]

a paper launch in Q3 with actual ramp starting Q4 based upon this but who knows, we’ll see soon enough.

q3 ends at end of september. anything should come online in the last 2 weeks of august. odms will likely get some laptops in stock shipping via orders and as you said as things ramp up they should appear in retail. I think this is a smash win for intel because zen 4 laptops are rare to see in stores at least anything that isn't basic.

 

[lightisgood]

MTL tape in in mid 2021 would mean ~27 months would have passed between tape in and launch. RPL took <24 months, and while ADL isn't known specifically, even if we assume that the full extra 6 months of dev time ADL took went between tape in and launch, ADL would have taken <30 months.
So ye, MTL's schedule fits very similarly with ADL/RPL, which should be pretty disappointing considering Intel was claiming that by disaggregating their tiles in chiplets they would be able to shrink their dev time by ~1 year. GG go next lol.
It might have been possible for MTL to launch early 2023 if their design teams had it in them too, but they couldn't. I wouldn't blame the node guys on this one.

My understandings are very simple.

1. MTL dev team had enough time.
2. I have no evidence that MTL design got tangled up.
3. Fab 34 was't manufacturing ready for I4 by end of 22, at least.

I wouldn't blame MTL dev team, you too. lol.

 

[dullard]

Gelsinger and the CFO both said in the recent earnings call that MTL was launching/PRQ in Q4. Then, like 4 days later Gelsinger said it was Q3. Seems like they made a lot of progress in those 4 or so days.

So far all signs are pointing to an announcement in their innovation event mid-Sept (Q3) and launch in Oct (Q4). So which quarter should they refer to? It wasn't a lot of progress, it is just where the quarter end happens to be. With many Intel launches in October historically, this happens somewhat frequently.

Innovation Event, Sept 19 and 20: https://www.intel.com/content/www/us/en/events/on-event-series/innovation.html "In this track, learn how we are accelerating edge-to-cloud silicon, delivering on future processor technology such as Meteor Lake" More specifically, this is in their Next Generation Systems & Platforms track, Intel Client Hardware Roadmap and the Rise of AI section: "In this session, you will learn about our latest client hardware platforms, including the highly anticipated Intel® Core™ Ultra processors (codename Meteor Lake), what you can expect from our future roadmap" https://reg.rainfocus.com/flow/inte...talog?search.trackanchor=1686744345550002foNk

 

[Hitman928]

So far all signs are pointing to an announcement in Sept (Q3) and launch in Oct (Q4). So which quarter should they refer to?

I would say Q4 in this case. Companies pre announce products all the time, look at console launch history where you might get an announcement 2 quarters ahead of the launch. A company would be rightly criticized if they said a product was coming in May but it was just announced in May and would be launched until November. No one really cares when you announce something, they care about when you can get the product into the market.

 

[dullard]

I would say Q4 in this case. Companies pre announce products all the time, look at console launch history where you might get an announcement 2 quarters ahead of the launch. A company would be rightly criticized if they said a product was coming in May but it was just announced in May and would be launched until November. No one really cares when you announce something, they care about when you can get the product into the market.

Well there you have it, it is Q4. There is no "made a lot of progress in those 4 or so days".

Their specific fact sheet to go off of for their earnings call says: "Building on strong demand for our 13th Gen Intel® Core™ processor family, Meteor Lake is ramping well in anticipation of a Q3 PRQ (production release qualification)". If you are going to nitpick over a few days, then that statement is not quite the same as claiming it will launch in Q3. https://download.intel.com/newsroom/2023/corporate/2Q2023-Earnings-Call-Comments.pdf That is the last really key step before release. The way I read your comments is as if they claimed it would launch then.

Edit: reading more into this, I think you are speaking of Emerald Rapids, not Meteor Lake.

 

[Hitman928]

Well there you have it, it is Q4. There is no "made a lot of progress in those 4 or so days".

Their specific fact sheet to go off of for their earnings call says: "Building on strong demand for our 13th Gen Intel® Core™ processor family, Meteor Lake is ramping well in anticipation of a Q3 PRQ (production release qualification)". If you are going to nitpick over a few days, then that statement is not quite the same as claiming it will launch in Q3. https://download.intel.com/newsroom/2023/corporate/2Q2023-Earnings-Call-Comments.pdf That is the last really key step before release. The way I read your comments is as if they claimed it would launch then.

Edit: reading more into this, I think you are speaking of Emerald Rapids, not Meteor Lake.

Well, my post was made in tongue in cheek fashion. It's ok to have a little fun around here still, isn't it?

With that said, yes, I went back and checked and I had misremembered the MTL and EMR conflicting statements. MTL is PRQ in Q3. EMR is either Q3 or Q4 depending on if you ask Gelsinger at the earnings call or 4 days later it seems.

 

[Geddagod]

Well, my post was made in tongue in cheek fashion. It's ok to have a little fun around here still, isn't it?

With that said, yes, I went back and checked and I had misremembered the MTL and EMR conflicting statements. MTL is PRQ in Q3. EMR is either Q3 or Q4 depending on if you ask Gelsinger at the earnings call or 4 days later it seems.

Schrodinger's CPUs lol
MTL is PRQ, and launch, in 2023. Pat was trying to be very sly about when exactly MTL launches, but he slipped up in the Q2 2023 earnings call.

Meteor Lake will not be a pre-PRQ reserve in the third quarter because we expect to launch that this quarter.

He also stated

We see this as a true AI PC moment that begins with Meteor Lake in the fall of this year.

Fall is Sept, October, November. Only one in Q3 is Sept. Well, what else is Sept this year?

How much volume is actually going to be at launch though, is questionable.

 

[A///]

Pat also patted himself on the back with the rearview mirror bit. that backfired on intel fast.

 

[Hitman928]

Schrodinger's CPUs lol
MTL is PRQ, and launch, in 2023. Pat was trying to be very sly about when exactly MTL launches, but he slipped up in the Q2 2023 earnings call.

He also stated

Fall is Sept, October, November. Only one in Q3 is Sept. Well, what else is Sept this year?
View attachment 84247
How much volume is actually going to be at launch though, is questionable.

Depends on who you ask, fall can either be Sep 1 to Nov 30 or Sep 22 to Dec 21 this year. I think the Sep 22 to Dec 21 definition is the more common one, at least in the U.S. Not that this really matters, just being a bit pedantic.

I agree on the point though. I also am curious how quickly actual end products (i.e., laptops) will appear with MTL. Not launching until end of Q3, especially if in meme volumes, will make it difficult to hit the holiday season for laptop makers.

 

[Doug S]

Apple is reportedly typically taping out their SoCs a bit less than a year before the iPhones containing it launch and they're shipping in the tens of millions. So Intel doesn't have to take ~2 years from tape out to ship, that's an artifact of not doing the level of simulation others do and thus needing too many steppings before they get something production worthy.

 

[SiliconFly]

I'm going to say that was MTL just not being design ready rather than the fabs holding them back (this time at least).

Yep. Pat said last year that the nodes are ahead of schedule. And he repeated the same this year too. Meaning, Intel 4, 3, 20A & 18A are pretty much on track or even ahead of schedule. Nodes are the things we need to worry about.

It's the client cpu design teams that we shud be really worried about.

 

[SiliconFly]

I have no evidence that MTL design got tangled up.
BTW, in generally, tiled arch tends to show one digit fewer bugs than monolithic arch.

True. Tiled means less bugs in tiles. But integration is a b****!

 

[Thunder 57]

Yep. Pat said last year that the nodes are ahead of schedule. And he repeated the same this year too. Meaning, Intel 4, 3, 20A & 18A are pretty much on track or even ahead of schedule. Nodes are the things we need to worry about.

It's the client cpu design teams that we shud be really worried about.

I'm not confident on nodes after 14 and 10 whichc became Intel 7. If they can execute, I'll believe in them again.

 

[SiliconFly]

Do you have the corresponding dates for ARL, to get some range of expected ARL launch dates?

ARL should launch on schedule in Q3 2024 as planned. But are there going to be any "unexpected" delays... that we don't know yet!

 

[SiliconFly]

I'm not confident on nodes after 14 and 10 whichc became Intel 7. If they can execute, I'll believe in them again.

But we keep seeing leaks (just rumors) where MTL clocks above 5 GHz comfortably. If true, it speaks volumes. Considering Intel 4 is on it's first iteration, that wud be a big achivement.

 

[SiliconFly]

Intel Moving to Chiplets: ‘Client 2.0’ for 7nm

www.anandtech.com

https://images.anandtech.com/doci/16021/Section%208%20%2820%29.jpg

In theory, on the one hand, ADL needs "3-4 years of Dev time", but on the other hand, MTL needs "2-3 years of Dev time".
RPL diverted ADL design. So, it is'nt an instance appropriate for this debate.

MTL has been in the making for more than 4 years. The MTL cpu tile tape-in happened in early 2021.

 

[SiliconFly]

Just for fun, based on the discussion back over at the Zen 5 thread, I was interested in how GLC effected clocks in relation to WLC.
Because of how rare power limits/scaling tests for non-high end parts are, and how GLC P-core only CPUs are rare, I had to do some digging and ended up comparing the 65watt 12400f vs a power-limit removed 11400h and found the 12400f to clock slightly higher (Jarrods Tech), while also having significantly higher IPC (18%). Sure, Intel 7 is supposed to be a 10-15% jump in perf too, but I still think it's pretty nice. Maybe I shouldn't bully GLC so hard online lol (I say this while typing on a 12900h laptop).

GLC has had it's time. Time to let go of existing fat & inefficient P-cores that take up way too much die space. Lets hope they go away sooner rather than later.

 

[coercitiv]

But we keep seeing leaks (just rumors) where MTL clocks above 5 GHz comfortably. If true, it speaks volumes. Considering Intel 4 is on it's first iteration, that wud be a big achivement.

Leaks don't build trust in execs, truth does. People pointing out what Pat says keep forgetting that. If MTL comes in on time and delivers the goods, that's great news. It's also the first step of a minimum of two successful consecutive launches on new nodes before we start believing Intel PR again.

And please, for the sake of the thread, stop saying things like "Intel 4 is on it's first iteration". You're just inviting another derail.

 

[SiliconFly]

Leaks don't build trust in execs, truth does. People pointing out what Pat says keep forgetting that. If MTL comes in on time and delivers the goods, that's great news. It's also the first step of a minimum of two successful consecutive launches on new nodes before we start believing Intel PR again.

And please, for the sake of the thread, stop saying things like "Intel 4 is on it's first iteration". You're just inviting another derail.

If Intel delivers on all fronts with MTL, it's a very huge sign that the design teams have adapted to the new Intel foundry rules and are doing a good job. And it also means the nodes are healthy and progressing well as Pat claims. They don't have to deliver for the next 10 years to build enough confidence. MTL, if it delivers good, it itself can do a lot.

 

[coercitiv]

They don't have to deliver for the next 10 years to build enough confidence.

They have to deliver everything promised until 2025, including a healthy 18A product. After that they will have a repaired reputation.

You're confusing a successful delivery with a track record.

 

[SiliconFly]

They have to deliver everything promised until 2025, including a healthy 18A product. After that they will have a repaired reputation.

You're confusing a successful delivery with a track record.

Intel already had a 40 yr successful track record and it didn't help them when they fumbled in the last 5 years.

AMD had a trashy and very questionable track record for many decades and their success with Zen has already skyrocketed their reputation in a very short span. Don't forget the fact that by the time Zen 2 came out, they were already rock stars and people were worshiping them!

Truth is, public memory is very short. No one really cares about the past. In the next 2 years, we can see comments in this very same thread stating how things were very bad for Intel 2 years ago.

 

[DrMrLordX]

I don't see the point in a customer needing to look at other things aside from the processor to know if it's good or not.

Some customers don't care as much about the CPU as they do the rest of the features that come with the platform. Which is why Intel created the Centrino label in the first place.

…and yet Meteor Lake and Intel 4 are on time

8+16/8+32 Meteor Lake has been completely cancelled, and 6+8 (or 2+8 or other configs) were supposed to be out last year. Intel 4 was supposed to debut with Ponte Vecchio as the pipecleaner in late 2020.

Its ok to admit to being wrong instead of doubling down on a silly take.

And is it okay to be right?

AMD had a trashy and very questionable track record

Wow don't even know what to say to that one.

 

[Abwx]

AMD had a trashy and very questionable track record for many decades

Neither the Athlon in 1999 nor the Athlon 64 few years later ever actually existed...

Indeed 64bit X86 CPUs came from nowhere.