AMD "Zen 5c" CCDs Made On More Advanced 3 nm Node Than "Zen 5"

[Metroid]

No, it doesn't.

Zen xC doesn't have the ability to have 3DVCache.

Do you want to explain it better, could not understand this "Zen xC doesn't have the ability to have 3DVCache.".

 

[dgianstefani]

Do you want to explain it better, could not understand this "Zen xC doesn't have the ability to have 3DVCache.".

The C cores can't have it.

 

[Metroid]

The C cores can't have it.

Okay, so you are saying the Zen 5C can't have x3dcache, interesting as rumours said a x3dvcache zen5 cpu would also launch this time around, maybe could be something else.

 

[R0H1T]

The C cores can't have it.

And why is that? The reason seems purely based on economics, because AMD cut their cache size, but anything technological?

 

[dgianstefani]

And why is that? The reason seems purely based on economics, because AMD cut their cache size, but anything technological?

The C cores don't have the through silicon via connections to send data vertically to the X3D cache.

They're space optimized like E cores. Extra fluff in the architecture is removed.

Would also make little sense to halve the cache on a core then put more on top, making heat transfer more difficult.

 

[john_]

Intel: Here, get some cut down cores with butchered specs and let's call them just Efficiency Cores, because thanks to their cut down specs they use less energy.

AMD: Here, get some identical cores with less cache and lower frequencies. Let's also use a more advance manufacturing node to get REAL Efficiency out of them.

 

[dgianstefani]

Intel: Here, get some cut down cores with butchered specs and let's call them just Efficiency Cores, because thanks to their cut down specs they use less energy.

AMD: Here, get some identical cores with less cache and lower frequencies. Let's also use a more advance manufacturing node to get REAL Efficiency out of them.

Intel E cores use 1/4 of the die area as P cores, yet are still as fast as a Skylake core. AMD C cores use a bit more than 1/2 of the die area as a normal Zen core, so the die space savings are nowhere near as great. This means that Intel has plans to release 8P+32E, whereas I doubt AMD will release more than an 8+16 for their Zen consumer platforms.

 

[john_]

Intel E cores use 1/4 of the die area as P cores, yet are still as fast as a Skylake core. AMD C cores use a bit more than 1/2 of the die area as a normal Zen core, so the die space savings are nowhere near as great. This means that Intel has plans to release 8P+32E, whereas I doubt AMD will release more than an 8+16 for their Zen consumer platforms.

You aren't saying something new here. Intel designed the E cores to manage to get the upper hand in marketing against AMD, by advertising more cores, while having a big disadvantage in node manufacturing. They tried to build a 10 P core CPU and failed, so they realized that they couldn't compete with only big cores. AMD had done the same 15 years ago with Bulldozer, where it was marketing 1.5 cores as dual cores, in an effort to somehow nullify Intel's "more threads" advantage with Hypertheading.

In any case you come to my words by pointing at Skylake. It's exactly what I am saying. Cores with lower performance and newest features missing to make them smaller. AMD doesn't really need much smaller cores, because Zen P cores are already smaller than Intel's P cores and also it enjoys the advantage of node manufacturing.

Intel was always planning to keep that 8 P cores configuration for years and only increase the E cores number, because 8 P cores are more than enough today for about everything in consumer space.

 

[dgianstefani]

You aren't saying something new here. Intel designed the E cores to manage to get the upper hand in marketing against AMD, by advertising more cores, while having a big disadvantage in node manufacturing. They tried to build a 10 P core CPU and failed, so they realized that they couldn't compete with only big cores. AMD had done the same 15 years ago with Bulldozer, where it was marketing 1.5 cores as dual cores, in an effort to somehow nullify Intel's "more threads" advantage with Hypertheading.

In any case you come to my words by pointing at Skylake. It's exactly what I am saying. Cores with lower performance and newest features missing to make them smaller. AMD doesn't really need much smaller cores, because Zen P cores are already smaller than Intel's P cores and also it enjoys the advantage of node manufacturing.

Intel was always planning to keep that 8 P cores configuration for years and only increase the E cores number, because 8 P cores are more than enough today for about everything in consumer space.

The Intel designs are faster than the Zen 7950X though in multithreading, even with a node disadvantage and monolithic, so the approach works, despite your feelings on "slow" cores.

 

[john_]

The Intel designs are faster than the Zen 7950X though in multithreading, even with a node disadvantage and monolithic, so the approach works, despite your feelings on "slow" cores.

"My feelings"......Oh my........... Fine. Whatever. Viva la 300W!

 

[Kapone33]

The Intel designs are faster than the Zen 7950X though in multithreading, even with a node disadvantage and monolithic, so the approach works, despite your feelings on "slow" cores.

In everything? I did not think so.

 

[dgianstefani]

In everything? I did not think so.

 

[john_]

In everything? I did not think so.

Don't bother. I mean 3% faster on average, if you choose specific apps, at probably 50-100% more power consumption and having to train the OS to manage multiple types of cores, is a great design. Go figure.....

 

[dgianstefani]

Don't bother. I mean 3% faster on average, if you choose specific apps, at probably 50-100% more power consumption and having to train the OS to manage multiple types of cores, is a great design. Go figure.....

Choose specific apps?

My dude there's 12 pages of application testing done to get that average.

Training the OS? Unlike Zen with a software based driver scheduler for X3D that works most of the time, but is based on manual whitelists, likely for Zen C hybrid CPUs as well, Intel has a hardware based thread director.

You're talking about power consumption as if its relevant to your argument against E cores, when you say yourself AMD has a node advantage. 5nm TSMC is not comparable to 10nm Intel 7 superfin.

I guess we'll see when both Intel and AMD have their CPU "tiles" or "CCDs" next generation.

From what I understand there's major advancements made to tech with Arrow Lake, including RibbonFET gate all around transistors, foveros packaging and backside power delivery. Zen 6 is moving to comparable packaging tech which should fix the idle and low load efficiency issues with Zen 1-5.

 

[R0H1T]

Yeah no

Intel Core i5 14600K & Intel Core i9 14900K Linux Benchmarks Review - Phoronix

www.phoronix.com

 

[dgianstefani]

Yeah no

Intel Core i5 14600K & Intel Core i9 14900K Linux Benchmarks Review - Phoronix

www.phoronix.com

An average of 25 W more is pretty interesting.

TPU forum though, so I'm using TPU testing, and synthetic server benchmarks isn't exactly "application" testing in the way people on this site understand it.

 

[Noyand]

You aren't saying something new here. Intel designed the E cores to manage to get the upper hand in marketing against AMD, by advertising more cores, while having a big disadvantage in node manufacturing. They tried to build a 10 P core CPU and failed, so they realized that they couldn't compete with only big cores. AMD had done the same 15 years ago with Bulldozer, where it was marketing 1.5 cores as dual cores, in an effort to somehow nullify Intel's "more threads" advantage with Hypertheading.

In any case you come to my words by pointing at Skylake. It's exactly what I am saying. Cores with lower performance and newest features missing to make them smaller. AMD doesn't really need much smaller cores, because Zen P cores are already smaller than Intel's P cores and also it enjoys the advantage of node manufacturing.

Intel was always planning to keep that 8 P cores configuration for years and only increase the E cores number, because 8 P cores are more than enough today for about everything in consumer space.

Ehh, saying that it's about marketing shows your feelings about Intel . The first iteration of the hybrid design had the same core count, but lower threads count vs AMD. And while the 13900k does have more cores, it was necessary to stay competitive at the high end. Intel 16 P-core Xeon is noticeably slower at similar Power limits (Puget doesn't do reviews with unlimited PL they use PL1 125w and pl2 250w). Nowadays, the consumer platform is also a prosumer platform, Buying into a TR or Xeon platform is not being to bring any ROI for some jobs where the core scaling isn't that great, memory/expansion cards requirement not that heavy (I'm saying this before I get called out by "real pros are ALWAYS buying xeons and Threadripper.) There are still many apps that like a balance of fast cores and Multithreading. Intel was starting to lose the prosumer market before ADL. The 3900x/3950x and the 5900x/5950x had no competition in that sector for a looong time. Intel had a real problem on their hands and e-cores was the solution. Now there's even XEON made up of e-cores.

But yhea, AMD doesn't really have any reason to bring C-cores to the desktop unless Intel get a meaningful in MT.

 

[R0H1T]

Spoiler

TPU forum though, so I'm using TPU testing, and synthetic server benchmarks isn't exactly "application" testing in the way people on this site understand it.

You said 12 apps which isn't really that big of a number, although it's Linux so not exactly comparable even if you'd generally get better results for Intel (on Linux) as well.

 

[dgianstefani]

View attachment 335411


You sad 12 apps which isn't really that big, although it's Linux so not exactly comparable even if you'd generally get better results for Intel (on Linux) as well.

12 pages of apps.

Read the actual TPU review, it has the AMD results too.

Intel Core i9-14900K Review - Reaching for the Performance Crown

Core i9-14900K is Intel's new flagship with clock speeds of up to 6 GHz. It's actually clocked even higher than the 13900KS, thanks to an extra 100 MHz when more than two cores are active. Our review confirms that Raptor Lake Refresh is amazing for both applications and gaming, if you can live...

www.techpowerup.com

 

[R0H1T]

So 43(?) apps if I counted them right. Phoronix still has a larger data set & not all of them are server or HPC benchmarks.

 

[Kapone33]

12 pages of apps.

Read the actual TPU review, it has the AMD results too.

Intel Core i9-14900K Review - Reaching for the Performance Crown

Core i9-14900K is Intel's new flagship with clock speeds of up to 6 GHz. It's actually clocked even higher than the 13900KS, thanks to an extra 100 MHz when more than two cores are active. Our review confirms that Raptor Lake Refresh is amazing for both applications and gaming, if you can live...

www.techpowerup.com

Yep. This is from the review you posted. Uh oh looks like the 12 core you love to hate is faster too.

 

[dgianstefani]

Yep. This is from the review you posted. Uh oh looks like the 12 core you love to hate is faster too.

View attachment 335414

OK. Now look at the average not a single example.

 

[Kapone33]

OK. Now look at the average not a single example.

You are moving goal posts. This was your statement "The Intel designs are faster than the Zen 7950X though in multithreading," That implies everything. It is too bad that you did not add the word nuanced as that is the truth. There are applications that prefer either or. Ryzen has been around long enough to have software developed around it. I used one after looking through the review for 10 seconds, I am sure I can find more and surfing the Net find even more. It comes back to the fact that we should be discussing Zen5 and if the IPC gain is true there is nothing, even Multithreaded where the 14900K will win handily. There is also the power draw argument to apply to Intel. A 14900K can draw more power than a 6800XT........

 

[john_]

Choose specific apps?

My dude there's 12 pages of application testing done to get that average.

Training the OS? Unlike Zen with a software based driver scheduler for X3D that works most of the time, but is based on manual whitelists, likely for Zen C hybrid CPUs as well, Intel has a hardware based thread director.

You're talking about power consumption as if its relevant to your argument against E cores, when you say yourself AMD has a node advantage. 5nm TSMC is not comparable to 10nm Intel 7 superfin.

I guess we'll see when both Intel and AMD have their CPU "tiles" or "CCDs" next generation.

From what I understand there's major advancements made to tech with Arrow Lake, including RibbonFET gate all around transistors, foveros packaging and backside power delivery. Zen 6 is moving to comparable packaging tech which should fix the idle and low load efficiency issues with Zen 1-5.

Did I hurt your feelings? If so, I am sorry.

Ehh, saying that it's about marketing shows your feelings about Intel

It is marketing. How do you think AMD started winning market share? By offering more cores than Intel, even when those cores where lower IPC cores. That's how Ryzen became a success. Where Intel was offering 4 cores, AMD started offering 6 and 8 cores.
It's not about feelings.
Thinking that every post is about feelings, only shows how YOU think and why YOU post, in this case. It's not about me. Don't project yourself on me. It doesn't work.

AMD doesn't really have any reason to bring C-cores to the desktop

They do. Marketing. They can't sell a 16 core CPU when Intel will be selling a "24 core" or a "32 core" CPU at the same price point, or even at a little higher price.

 

[dgianstefani]

You are moving goal posts. This was your statement "The Intel designs are faster than the Zen 7950X though in multithreading," That implies everything. It is too bad that you did not add the word nuanced as that is the truth. There are applications that prefer either or. Ryzen has been around long enough to have software developed around it. I used one after looking through the review for 10 seconds, I am sure I can find more and surfing the Net find even more. It comes back to the fact that we should be discussing Zen5 and if the IPC gain is true there is nothing, even Multithreaded where the 14900K will win handily. There is also the power draw argument to apply to Intel. A 14900K can draw more power than a 6800XT........

Multithreading performance is calculated by TechPowerUp, and every other review site, by averaging many tests of many tasks, 43 here on TPU and 173 on Phoronix, this is because singular examples are not representative of general performance. Phoronix is more of a Linux server site, with tests that are relevant to that usage. Your single example of AI upscaling, where a GPU would be more appropriate to use, doesn't change that the Intel chips are faster in multithreading. This statement is true despite your attempts to find examples that support your choice of CPU.

They do. Marketing. They can't sell a 16 core CPU when Intel will be selling a "24 core" or a "32 core" CPU at the same price point, or even at a little higher price.

And that's exactly why TPU tests these products.

Marketing ≠ performance.

The testing doesn't lie.