Discussion Intel current and future Lakes & Rapids thread

[Hulk]

That has nothing to do with it. For AMD it was the right way to go. If the BoD didn't hire CEOs laser focused on nothing but "moolah" and/or they had a semi competent CEO Intel wouldn't be in the trouble they are in today.

Paul Otellini did a decent job considering his background but overall he still fell into that same trap(The famous passage of declining iPhone/iPad chip production come into mind).

Otellini should have prepped Gelsinger for the next CEO rather than firing him over the Larrabbee debacle. Yes, people make mistakes unfortunately. Larrabbee failure was big but not something that would break the company.

"Pride is before the fall"

They had the best times with 22nm before they screwed up on 14nm. I don't know how many remember that? Because the 10nm delay completely overshadowed 14nm failures and delays. That's why Broadwell got a shoddy launch.

14nm also started the crazy focus on DensityTM, the definition of where the MTr/mm2 metric is used as epeen competition but was useless for the Core chips as it never got the scaling benefit anyway.

Then they decided to DOUBLE DOWN on the DensityTM with 10nm.

I think the density focus had *some* merit but they went all the way. Traditionally Intel transistors were slightly less dense than TSMC but was much more performant. Then they decided it wasn't enough. The density focus benefitted GPUs and the E core chips but they should have weighed the options more carefully. Something a competent management team would have done, so you don't bet the whole farm.

Broadwell process issues was a foreshadowing of 10nm. I still wonder if the reason we never really saw a widespread Broadwell desktop release was due to 14nm issues.

 

[mikk]

Broadwell process issues was a foreshadowing of 10nm. I still wonder if the reason we never really saw a widespread Broadwell desktop release was due to 14nm issues.

For sure it was, the first 14nm version was really bad. Low clock speeds and bad yields, Broadwell mobile was delayed by about 6 months if not more. Skylake was too close to Broadwell, there was no point for Intel working on a full desktop Broadwell lineup. With Skylake 14nm was in much better shape but still not that great in regards to clock speeds. Kabylake fixed this.

 

[Hulk]

For sure it was, the first 14nm version was really bad. Low clock speeds and bad yields, Broadwell mobile was delayed by about 6 months if not more. Skylake was too close to Broadwell, there was no point for Intel working on a full desktop Broadwell lineup. With Skylake 14nm was in much better shape but still not that great in regards to clock speeds. Kabylake fixed this.

And Kabylake R fixed it some more.

 

[Exist50]

I don't see why it can't be Gracemont, but fair enough.

Heavily modified, means it'll take additional time and a branch from the original Gracemont core. You could say that's Crestmont but it won't be as they'll modify it specifically to enhance HPC, while Crestmont will be more for client. Also if they are talking 2023 that's not a long time, and their mont chips come every 2 years.

Grand Ridge comes later but also uses Gracemont and is on Intel 4.

Rest assured that every part of Nosta's description of Sierra Forest is wrong. Every part.

 

[Exist50]

@mikk I think the whole Haifa vs. Oregon made a difference as late as Haswell but after that I'm pretty sure it matters little.

If you see the subsequent designs, there has been lot of engineers from both sides that have been reassigned or moved to the other team. Whatever team/thing/idea was there up to the Otellini era has been rebuilt/reimagined after the Kraznich debacle.

We only cared about "Haifa vs Oregon" since they made names after Core 2 and we got sneak peak into the two teams' philosophies. Oregon was about systems design(Hyperthreading, QPI, memory controller, etc) and what you may call uncore while Haifa was about the uarch.

Looking back into Brian Kraznich's legacy he sounded like the Hulk gone crazy and just wanted to gut everything good about the company. Before you say one person can't make all the difference, that's true, but it's far easier to destroy than build. The Intel you see today can't be compared to Intel pre-2014.

At least in the past decade or so, the split has been more about power, at least on the client side. The Israeli team is about performance at any cost, while the Oregon team (heavily influenced by Broxton) is more power sensitive. You can also see this split in Core (IDC) vs Atom (DDG). I'm told that at one point, the company even considered splitting the client architectures along this line, though that plan was abandoned.

In any case, what I said 2 years ago holds true. The Core line is owned by IDC, and will likely remain that way forevermore. But over the course of 2019 and onward, I'm told there was a change. Something new is afoot. Call it part of Keller's legacy, if you want, but let's just say that Glenn Hinton (and others) did not return to work on Core.

 

[igor_kavinski]

But over the course of 2019 and onward, I'm told there was a change. Something new is afoot. Call it part of Keller's legacy, if you want, but let's just say that Glenn Hinton (and others) did not return to work on Core.

Nova Lake with Royal Cores (double the IPC of Golden Coves). Rumored to be the culmination of the work started by Jim Keller..

 

[mikk]

Nova Lake with Royal Cores (double the IPC of Golden Coves). Rumored to be the culmination of the work started by Jim Keller..

Yes the old reddit leak claimed that Intel "is working to build entirely new architecture from the ground up" for Nova Lake and this is why Glenn Hinton returned.

 

[Exist50]

Nova Lake with Royal Cores (double the IPC of Golden Coves). Rumored to be the culmination of the work started by Jim Keller..

I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher, though what they'll achieve in reality, who can say.

 

[mikk]

I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher, though what they'll achieve in reality, who can say.

Maybe he doesn't even refer to Nova Lake, hard to say really.

 

[dmens]

I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher, though what they'll achieve in reality, who can say.

He is laughing at your gullibility.

 

[lobz]

I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher, though what they'll achieve in reality, who can say.

In all fairness, he may have laughed at how unbelievably unrealistic a 100% increase in IPC would be.

 

[TESKATLIPOKA]

I am also very sceptical. We most likely can't expect 100% increase in IPC with Nova Lake, that would mean 15% IPC increase on average.

	Alder Lake	Raptor Lake	Meteor Lake	Arrow Lake	Lunar Lake	Nova Lake
IPC (%)	100	115	132	152	175	201

We already know Raptor Lake won't have 15% better IPC compared to ADL, big core is only a modified Golden Cove. Meteor Lake (Lunar Lake) and Arrow Lake could be the same thing from the CPU side, both using the same cores, only aimed for different segment (mobile vs desktop).

 

[mikk]

15% every generation is not out of reach. I believe we will see more than 15% with Lunar and a lot more with Nova. Meteor might be less than 15% and Raptor most likely as well. Intel can invest in more P core size+performance because the P core count increase is on hold thanks to the E cores. And there is big competition with AMD, I think we will see bigger IPC increases the next few years.

 

[BorisTheBlade82]

Maybe some of their R&D guys just reinvented the good old hockey stick chart 😉

Hockey Stick Chart: Overview and Examples in Investing

A hockey stick chart is a line chart in which a sharp increase occurs suddenly after a short period of quiescence.

www.investopedia.com

 

[Exist50]

In all fairness, he may have laughed at how unbelievably unrealistic a 100% increase in IPC would be.

Nah. Skepticism is more than warranted, clearly, given the history here, but it was a laugh with a smile, if you get my drift.

Though I do find it pretty amusing to have people telling me what my own friend meant, lol.

He is laughing at your gullibility.

How are your Alder Lake and Gracemont predictions fairing again

We most likely can't expect 100% increase in IPC with Nova Lake, that would mean 15% IPC increase on average.

I think amortized is the wrong way to look at it. I expect a single, large jump, though how large, I really don't know. But that's an independent effort from the Core/Cove line. Clearly Raptor Lake will have minimal IPC gains, so there's no plausible intercept from the Core line, unless you expect them to suddenly pull 20%+ annually out of thin air.

 

[igor_kavinski]

I have a feeling that it's some performance enhancing trick they have never utilized before because it would have been wasteful of their transistor budget. With Nova Lake, they will finally have enough transistor density to consider it.

 

[dmens]

Heh, an annual 15% jump to reach 2x is absurd enough, but now it is a single 2x leap in one moonshot? Man, you twitterati will say anything for clicks.

I suppose that is one way to move on from your delusional fantasy predictions about Alderlake. Quick reminder: 10 watt Gracemont cores.

 

[mikk]

Heh, an annual 15% jump to reach 2x is absurd enough, but now it is a single 2x leap in one moonshot? Man, you twitterati will say anything for clicks.

I suppose that is one way to move on from your delusional fantasy predictions about Alderlake. Quick reminder: 10 watt Gracemont cores.

He didn't say he expects a 2x jump from one to another generation, he expects one large jump from the supposedly new architecture. Talking about delusional fantasy predictions: your predictions in this thread were one of the worst in the past and you don't even understand basic things like roadmaps. Furthermore your are easily the most toxic user in this thread, you are completely deluded thanks to your immense Intel hate, flaming others is your favoured thing when you don't like what the other said. You are so pathetic.


Personal attacks and insults are not allowed.


esquared
Anandtech Forum Director

 

[dmens]

you don't even understand basic things like roadmaps

Well, I don’t bother divining the future from non-technical powerpoint slide decks like you do. It is not my fault you have to stoop that low to come up with garbage scraps.

He didn't say he expects a 2x jump from one to another generation

“I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher”

“I expect a single, large jump”

So… he didn’t say he expects a 2x jump, it is really a >2x jump. Sorry about that.

 

[eek2121]

Wait

Intel is taking N4 wafers too? Hmm. Wonder what they're doing with those . . .

I mean it should be pretty obvious. Until they manage to have enough capacity to produce enough chips on Intel 4 (with EUV) on their own they will co-source.

I am going to go out on a limb, however, and say that the 4nm parts are GPU orders. The day we actually see TSMC roll out a Core design I will be shocked. Definitely not saying it isn't happening or won't happen, but rather, those chips will be produced LAST. TSMC will get the orders Intel can't fill. That should be obvious at any rate. Intel won't use TSMC unless they have to. If they suddenly start producing 40+% of their Core designs using TSMC we know they are behind. I suspect that won't happen, however. I strongly suspect TSMC will feature a prominent, yet minority role going forward with Intel. They are a no-brainer for the GPU business, for example, and for other projects...

The single biggest way for Intel's margins to get eaten alive are by using TSMC, that is why I think they will avoid doing it for ANY high margin product. As GPUs are a new product and not included in revenue, the hit they take for producing those won't matter at all. that is also why I expect the GPUs to not significantly undercut the competition.

 

[Exist50]

I have a feeling that it's some performance enhancing trick they have never utilized before because it would have been wasteful of their transistor budget. With Nova Lake, they will finally have enough transistor density to consider it.

I think the big question is, what is the goal for Royal? If it's to replace Core, then they'll still have to care about area efficiency, for server if nothing else. Apple shows that you can achieve much higher IPC (and power efficiency) without overly bloating area thanks to higher transistor densities and utilization, so that's one possibility. But if Royal is to live alongside Core, then that could give them freedom to make a "huge" design and back off Core to more of a middle ground.

 

[dmens]

Yes, Intel should spend engineering effort on an even more obese CPU (when their p-core already has the worst area efficiency on the market), and end up with a product which would be ridiculously expensive to manufacture, be worthless for servers due to high area/dollar cost for core scaling, and therefore can only go into the ever shrinking x86 client market with its low margins relative to servers so it has no hope of paying back that increased die area.

Brilliant, keep the jokes coming.

 

[TESKATLIPOKA]

He didn't say he expects a 2x jump from one to another generation, he expects one large jump from the supposedly new architecture.

“I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher”

“I expect a single, large jump”

So… he didn’t say he expects a 2x jump, it is really a >2x jump. Sorry about that.

Really?

First, he wrote this:

Nova Lake with Royal Cores (double the IPC of Golden Coves). Rumored to be the culmination of the work started by Jim Keller..

I wasn't able to get confirmation of the name, but when I mentioned that 2x IPC target to my source, he just laughed. I'm thinking the target might be significantly higher, though what they'll achieve in reality, who can say.

Second, I wrote this:

I am also very sceptical. We most likely can't expect 100% increase in IPC with Nova Lake, that would mean 15% IPC increase on average.
.....

Third, his reply was this:

I think amortized is the wrong way to look at it. I expect a single, large jump, though how large, I really don't know. But that's an independent effort from the Core/Cove line. Clearly Raptor Lake will have minimal IPC gains, so there's no plausible intercept from the Core line, unless you expect them to suddenly pull 20%+ annually out of thin air.

As shown, Exist50 didn't say he expects a single 2x IPC jump or higher in a single generation.
He said, instead of 15% IPC increase on average, he expects that there will be a single large IPC jump, the rest will be much smaller jumps. That means something like this for example.

	Alder Lake	Raptor Lake	Meteor Lake	Arrow Lake	Lunar Lake	Nova Lake
IPC	100	108(+8%)	122(+13%)	134(+10%)	144(+7.5%)	200(+39%)

* I think 15% on average is more believable than 39% increase in a single generation Only my opinion.

The one who started this nonsense about 2x jump(leap) in a single generation was actually you.

Heh, an annual 15% jump to reach 2x is absurd enough, but now it is a single 2x leap in one moonshot? Man, you twitterati will say anything for clicks.
......

Sorry about that.

 

[deasd]

It would be crazy and even laughable if a company target 'IPC x 2' in the future.
At least, GoldenCove to me already is a disapointment when comes to some theoretical test... I think nobody could precisely foresee future when comes to technology and especially semiconductor development which has been already stuck. I just clearly remember Intel already miss its process node target when comes to 10nm... Everybody should go back to reality.

 

[Cardyak]

I think the big question is, what is the goal for Royal? If it's to replace Core, then they'll still have to care about area efficiency, for server if nothing else. Apple shows that you can achieve much higher IPC (and power efficiency) without overly bloating area thanks to higher transistor densities and utilization, so that's one possibility. But if Royal is to live alongside Core, then that could give them freedom to make a "huge" design and back off Core to more of a middle ground.

From what I’ve heard, Royal core is about offering the efficiency of the Atom line up whilst also maintaining the performance of the Core line-up, through clever engineering and trickery.

Intels acquisition of SoftMachines back in 2016 offers a MASSIVE hint of how this is achievable.