Question How ahead is Intel in CPU design compared to AMD?

[Adonisds]

If we could get both companies to design a CPU core now on the same process and using the same number of transistors, who would be ahead and by how much? I'm assuming Intel would be ahead because Skylake has a similar IPC to Zen 2, but Zen 2 uses a more dense node, and Skylake is from 2015 while Zen 2 is from 2019.

The both companies using the same process scenario is just to illustrate the problem and I know it will never happen, but maybe there is a smart way to compare current CPU designs and find out who is ahead and by how much. Wikichip has die sizes and die pictures of Sunny Cove and Zen 2, but I'm not sure how to interpret and compare them.

Another question: If Intel had to design future cores using the same process and the same die size as Skylake, how much further could the Skylake design be improved? Also assuming it will have to have to be x86 too and have all the same functions.

 

[zinfamous]

As the OP, I find it funny that this became the "Official Intel vs AMD thread". I didn't mean to or predict that it would happen, I just wanted to tackle a very specific thought experiment that almost no one here tried answering. Because of that I then tried answering my own question and no one cared too. 😂

I'm not complaining, I am fine with how it turned out, you guys can talk about whatever you want obviously. I guess I should have predicted what would happen.

You gave birth to the thread. Your baby. Now, all you can do is sit by helplessly and watch the internet tear it to pieces.

 

[Adonisds]

In all fairness that is a loaded question, with an assumption. At this point I don't think Intel is ahead with products in the market. Things you can actually buy. Marketing slides and roadmaps don't count. Currently, with shipping products AMD is better at SMT, much better at security, pretty much even in IPC and much better at connectivity PCIe 4 vs PCIe 3. Intel is ahead in memory controllers/latency. And semi-niche things like AVX512.

Depending on your use case, AMD is more likely to be a better pick than Intel today. What will next year bring in actually shipping parts? I have no idea, my crystal ball is broken. And while speculation and rooting for your favorite team is all well and good, the only parts that matter are the ones you can buy today.

It's a question about who can arrange the same number of transistors in the most efficient way, and if that skill matters. I assumed it was Intel.

I'm not an Intel fanboy, I would buy an AMD system today, I think they are in a better position. I think the chiplet design is amazing and they also have a process advantage. But I didn't mean to discuss who is making better products now.

 

[Adonisds]

You gave birth to the thread. Your baby. Now, all you can do is sit by helplessly and watch the internet tear it to pieces.

I am sorry, world

 

[DAPUNISHER]

This thread probably has lots of responses because I unintentionally used a clickbait title.

LULZ! sure it was. And I hope you are sincere that you have learned your lesson, and will not repeat the mistake in the future. Because speaking as a fellow member, I would much prefer to see you add value to the forums, rather than be THAT GUY/GAL.

 

[Markfw]

When people have poor capacity of abstraction we get jabs like these that totally misunderstand OP's text and still get lots of likes.

I need to try to remember that more, I did add "The both companies using the same process scenario is just to illustrate the problem and I know it will never happen" but I should have also removed anything concrete from the text and used instead "company X, company Y, Z nm process".

I am also team grey, I don't care who is ahead. This thread was not meant to be fanboy wars, it was not meant to discuss who is in a better position between Intel and AMD. It was not even meant to discuss who has better IPC.

The replies that try to be technical get much less response than stuff like this reply I am quoting.

This thread probably has lots of responses because I unintentionally used a clickbait title. Many people probably didn't even read the OP. It's my fault, I will try to avoid that in the future.

Even though you added "using the same process" thats a fantasy, one that will never happen. So its a mythical loaded clickbait question. What did you expect ?

 

[uzzi38]

Gonna be honest here - you guys are all focusing way too hard on his question ignoring some the intricacies of how things work. Even if it were intenionally loaded, it doesn't matter, it's a hypothetical question, there's absolutely no harm in asking. and there's also absolutely no harm in a straight answer assuming what OP's suggesting could be done.

I'm usually not one for making comments like this, but I can't help feel this is getting borderline-toxic here. Dude asked a question - a hypothetical one at that - what's so wrong with giving him a straight answer? The vast majority of people here have just gone and flipped at him for suggesting you can port uArchs directly over to other nodes. Of course you can't do so easily, nobody's suggesting you can. But for the love of god, it's just a hypothetical question, no need to hang the guy on a stake.

 

[Adonisds]

LULZ! sure it was. And I hope you are sincere that you have learned your lesson, and will not repeat the mistake in the future. Because speaking as a fellow member, I would much prefer to see you add value to the forums, rather than be THAT GUY/GAL.

Yes it was. I even admitted I made mistakes, and I haven't been behaving as if I wanted this tread to be fanboy wars. I barely even posted here until now. Why do people always have to assume the people they disagree with have the worst intentions?

 

[Adonisds]

Gonna be honest here - you guys are all focusing way too hard on his question ignoring some the intricacies of how things work. Even if it were intenionally loaded, it doesn't matter, it's a hypothetical question, there's absolutely no harm in asking. and there's also absolutely no harm in a straight answer assuming what OP's suggesting could be done.

I'm usually not one for making comments like this, but I can't help feel this is getting borderline-toxic here. Dude asked a question - a hypothetical one at that - what's so wrong with giving him a straight answer? The vast majority of people here have just gone and flipped at him for suggesting you can port uArchs directly over to other nodes. Of course you can't do so easily, nobody's suggesting you can. But for the love of god, it's just a hypothetical question, no need to hang the guy on a stake.

In my post where I tried pathetically to answer my own question because no one else was even trying to I even reached the conclusion that the assumption of the loaded question is probably wrong. I changed my mind.

If the mods want, we could make changes to the OP to try to change the direction of the thread, or close it. I am fine either way

 

[uzzi38]

In my post where I tried pathetically to answer my own question because no one else was even trying to I even reached the conclusion that the assumption of the loaded question is probably wrong. I changed my mind.

If the mods want, we could make changes to the OP to try to change the direction of the thread, or close it. I am fine either way

For the record, I tried to right at the beginning, (second half of this post) but kinda just stopped paying attention to the thread shortly after (I tend to reply from my phone, and it's easy to miss stuff), and I just came back to this now.

I'm not sure if you replied to what I wrote again afterwards, so if there was, mind if I ask you to ask again? I swear I'll reply this time xd

EDIT: And yeah, if you want me to provide power consumption figures or something, ask away, I can do that.

 

[Markfw]

In my post where I tried pathetically to answer my own question because no one else was even trying to I even reached the conclusion that the assumption of the loaded question is probably wrong. I changed my mind.

If the mods want, we could make changes to the OP to try to change the direction of the thread, or close it. I am fine either way

If you post a thread in moderator discussions, and ask for something specific, I am sure we can help you !

As for answering the question, even ignoring the node challenge, we don't know jack about Intel "current" technology since we have had no new chips (to speak of) in years.

So again, no answer can be made IMO, aside from wild guesses.

 

[uzzi38]

As for answering the question, even ignoring the node challenge, we don't know jack about Intel "current" technology since we have had no new chips (to speak of) in years.

Well, Ice Lake-U exists, and actually you can draw quite a lot of information from it and just a few Cinebench runs alongside media coverage, which is enough to give you information on clocks sustained in real workloads, power consumption, IPC uplift... pretty much everything needed to know for the most part.

You just gotta search for the info in the right places

 

[DAPUNISHER]

Yes it was. I even admitted I made mistakes, and I haven't been behaving as if I wanted this tread to be fanboy wars. I barely even posted here until now. Why do people always have to assume the people they disagree with have the worst intentions?

What do we disagree on? We both agree it is a click bait title, and you clarified why it was indeed unintentional. I am not contesting that. We also agree you made mistakes here, and you have assured us you have learned from them. Great! Next mistake to learn from is characterizing anyone that posted in this thread as being "fanboys". Despite the title, and despite the odd premise underlying it, which I failed to understand the purpose of. That is, until you explained it was a failed thought experiment on your part. Despite all that, I learned from reading this thread. There are some quality posts about CPU design, manufacturing, etc. that are all interesting to me.

As to assuming worst intentions: In 20 years of lurking and being a member here, thread titles like yours', with few exceptions, are made by attention seekers that will get that attention anyway they can. Usually they lack the technical acumen to do it in a positive manner, so they resort to negative behavior. Fortunately, you have been forthcoming about your intent, and showed contrition for having engaged in that type of behavior, albeit unintentionally/naively. That makes you okay by me. And I do thank you for initiating a discussion which I found informative, and that despite the title, I personally did not find to be a flame war at all. Instead, it was a constructive debate with good info for people like myself, that read these discussions, but lack the expertise necessary to engage in them.

 

[Adonisds]

If you post a thread in moderator discussions, and ask for something specific, I am sure we can help you !

As for answering the question, even ignoring the node challenge, we don't know jack about Intel "current" technology since we have had no new chips (to speak of) in years.

So again, no answer can be made IMO, aside from wild guesses.

I deeply regret ever using the words "Intel" and "AMD". I am still being misunderstood because of that. It would be cool but it doesn't need to use current tech in an answer. All I wanted was to ask if two different relevant companies using the same process and same number of transistors and cores can reach vastly different performance numbers because design matters. It doesn't seem to be a question that stupid.

To be clear, I think the thread as it is currently is OK. People are finding value in it so it can continue to exist. I just realized long ago that my question will not get answered. That's why I didn't post much until now and I probably shouldn't post much more since people are not being too kind with me.

 

[Carfax83]

History is meaningless. Intel is behind. I have my reasons to believe they will stay behind in the foreseeable future. And in my humble opinion, Intel's historic process supremacy is the single most important factor in its competitiveness in this market. You can extrapolate their lack of supremacy to the market landscape in the next five years.

Well, only time will tell I suppose. I just don't see Intel staying behind for any particularly long length of time. They already have great microarchitectures in their lineup, they just need to nail the process node. If 10nm doesn't deliver on that front, then perhaps 7nm will.

Then why did you say you can have both without qualifications?

Because both AMD and Intel attempt to design CPUs with the highest degree of IPC and clock speed that they can manage. There is a balance between the two which leads to greater overall performance, which is the ultimate objective.

IPC isn't enough by itself, and neither is clock speed if you want to achieve the highest performance.

No, it doesn't.

This is the Wiki page entry on Simultaneous Multithreading under details:

However, in most current cases, SMT is about hiding memory latency, increasing efficiency, and increasing throughput of computations per amount of hardware used.

Source

My understanding has always been that SMT hides memory latency in extreme cases of thread stalls and or long latency events because if one thread stalls, the other thread's resources can still be used.

Maybe I'm wrong though.

 

[Carfax83]

He said multiple times Sunny Cove will go into a wide range of products, including standard PCs (mentioned desktops separately too, for some reason) among a lot of things. That right there was the first straight up lie in so many ways.
The vulnerability mitigations is a pure gem too. After being deliberately silent and building cores for 5-10 years using degenerate speculation, suddenly a pretty, professionally looking, deep blue colored .ppt slide that says 'ongoing research to help us stay ahead of evolving security landscape' is enough for you?
Should we really pick apart the whole interview?

I don't know, this seems extremely cynical if you ask me. That interview was posted well over a year ago, and as we all know, things can change extremely fast in the tech industry.

I'm not trying to apologize for or cover up Intel's failure, I'm just saying it's likely due to incompetence or unrealistic expectations rather than anything malignant like outright lies.

 

[UsandThem]

Locked at OP's request.

AT Mod Usandthem