Intel Raptor Lake-S Cache Sizes Confirmed in Blurry CPU-Z Screenshot: 68MB L2+L3

[btarunr]

Back in January, we heard the first reports of Intel significantly increasing the on-die cache sizes on its 13th Gen Core "Raptor Lake-S" desktop processor, with the sum total of L2 and L3 caches on the silicon being 68 MB. A CPU-Z screenshot from the same source as the January story, confirmed the cache sizes. The "Raptor Lake-S" die in its full configuration features eight "Raptor Cove" performance cores (P-cores), and sixteen "Gracemont" efficiency cores (E-cores), making it a 24-core/32-thread chip.

Each "Raptor Cove" P-core features 2 MB of dedicated L2 cache even in its client variant, as previously reported, which is an increase from the 1.25 MB L2 cache of the "Golden Cove" P-cores on "Alder Lake-S." The sixteen "Gracemont" E-cores are spread across four E-core clusters, just like the eight E-cores of "Alder Lake-S" are spread across two such clusters. The four cores in each cluster share an L2 cache. Intel has doubled the size of this L2 cache from 2 MB on "Alder Lake" chips, up to 4 MB. The shared L3 cache on the silicon has increased in size to 36 MB. Eight P-cores with 2 MB each, and four E-core clusters with 4 MB, each, total 32 MB of L2 cache. Add this to 36 MB of L3 cache, and you get 68 MB of L2+L3 cache. Intel is expected to debut "Raptor Lake" in the second half of 2022 alongside the 700-series chipset, and backwards compatibility with 600-series chipset. It could go down as Intel's last client processor built on a monolithic silicon.



View at TechPowerUp Main Site | Source

 

[Chaitanya]

So Intel is going to start using "glue" instead of sniffing it.

 

[Crackong]

ok cool
But the competition has 100MB cache on a 8core CPU, and maybe 200MB on a 16 core
Good luck going that route

 

[DeathtoGnomes]

So Intel is going to start using "glue" instead of sniffing it.

Intels been using it for some time now, sniffing all the way.

 

[Minus Infinity]

ok cool
But the competition has 100MB cache on a 8core CPU, and maybe 200MB on a 16 core
Good luck going that route

But Raphael-X aren't coming immediately from what I've read, there will be Zen4 without the v-cache at first. The thing is will v-cache come to all models eventually?

 

[Taraquin]

But Raphael-X aren't coming immediately from what I've read, there will be Zen4 without the v-cache at first. The thing is will v-cache come to all models eventually?

Probably not as it is more expensive and only holds advantage in gaming. I bet entry and very high end with many cores will have less cache to save money and improve productivity. Maybe we get a 'Gaming' brand of AMD CPUs?

 

[tabascosauz]

So Intel is going to start using "glue" instead of sniffing it.

Raptor Lake is monolithic. Where's the glue?

14th gen, however...EMIB has a lot of promise to live up to. Can't be as bad as just running traces across the substrate though

 

[Chaitanya]

Raptor Lake is monolithic. Where's the glue?

14th gen, however...EMIB has a lot of promise to live up to. Can't be as bad as just running traces across the substrate though

I did try to quote last sentence by unfortunately on cell phone quote/multiquote doesnt seem to work at all.

 

[tabascosauz]

I did try to quote last sentence by unfortunately on cell phone quote/multiquote doesnt seem to work at all.

I'm not sure it's possible to direct quote a news post on any device.

Amusingly, the glue remark only really fits intel's own tech, so calling substrate Fabric links "glue" both shoots themselves and also gives AMD more credit than it deserves......way too many counterproductive PR choices from intel in 9th-10th gen years

 

[Crackong]

will v-cache come to all models eventually?

Eventually, yes.
In 2022-2023? Probably not.

 

[Deleted member 24505]

So Intel is going to start using "glue" instead of sniffing it.

Typical crap anti Intel bullshit in every thread about Intel. If Intel fans did the same in every thread about AMD it would be frowned upon.
This shit should be marked as a low value post

 

[dragontamer5788]

ok cool
But the competition has 100MB cache on a 8core CPU, and maybe 200MB on a 16 core
Good luck going that route

AMD's L3 cache is split between different clusters / CCX though, so the giant L3 caches are really 96MB L3 x (number of CCX), and not a monolithic 60+MB like what's offered here. From the perspective of a singular Zen3 core, only 96MB L3 exists. You need multithreaded programming to "unlock" the other caches. So Intel's design here is superior from the perspective of L3 cores sharing space with everyone else.

That being said: AMD's 96MBL3 for a single thread is still far more than all of this L3 combined. On the other hand, AMD's default chips only have 32MB L3 cache, so Intel very possibly has a value-argument, depending on how cheap this chip here is manufactured.

 

[DeathtoGnomes]

Typical crap anti Intel bullshit in every thread about Intel. If Intel fans did the same in every thread about AMD it would be frowned upon.
This shit should be marked as a low value post

There was a time when the glue reference was in every AMD thread, so yea yours should be low value as well as mine for taking the thread off topic.

so Intel very possibly has a value-argument

The value is not there until performance reviews are in. /2c

 

[Deleted member 24505]

There was a time when the glue reference was in every AMD thread, so yea yours should be low value as well as mine for taking the thread off topic.


The value is not there until performance reviews are in. /2c

Was my reply aimed at you? No, so what has low quality posts go to do with you, don't see no moderator title.

My comment is valid, go look in most Intel threads and see the low brow childish crap. Maybe I should do the same in AMD threads. Guess what would happen, some double standards here for sure.

The glue comment was crap as mussels pointed out when a monolith has no glue, so was the usual childish shit thrown in most Intel threads.

 

[Richards]

I'm not sure it's possible to direct quote a news post on any device.

Amusingly, the glue remark only really fits intel's own tech, so calling substrate Fabric links "glue" both shoots themselves and also gives AMD more credit than it deserves......way too many counterproductive PR choices from intel in 9th-10th gen years

Its not glue its advanced packaging ponte vechio proves that

AMD's L3 cache is split between different clusters / CCX though, so the giant L3 caches are really 96MB L3 x (number of CCX), and not a monolithic 60+MB like what's offered here. From the perspective of a singular Zen3 core, only 96MB L3 exists. You need multithreaded programming to "unlock" the other caches. So Intel's design here is superior from the perspective of L3 cores sharing space with everyone else.

That being said: AMD's 96MBL3 for a single thread is still far more than all of this L3 combined. On the other hand, AMD's default chips only have 32MB L3 cache, so Intel very possibly has a value-argument, depending on how cheap this chip here is manufactured.

Plus intel has bigger l2 cache which is way faster than even 96mb l3

 

[DeathtoGnomes]

Was my reply aimed at you? No, so what has low quality posts go to do with you, don't see no moderator title.

My comment is valid, go look in most Intel threads and see the low brow childish crap. Maybe I should do the same in AMD threads. Guess what would happen, some double standards here for sure.

The glue comment was crap as mussels pointed out when a monolith has no glue, so was the usual childish shit thrown in most Intel threads.

I was just pointing out that the glue nonsense started out in AMD threads. No need to be hostile.

 

[Deleted member 24505]

I was just pointing out that the glue nonsense started out in AMD threads. No need to be hostile.

So what about this nugget of intelligence then? Intels been using it for some time now, sniffing all the way.

Whatever. I don't see intel users posting shit in every AMD thread. Just wait till Intel is back in front. TPU will be a different place.

 

[Blaeza]

I'm an amateur to PCs and they both do there job. Let me play some games and watch YouTube, so Intel and AMD both win for me.

 

[Minus Infinity]

Its not glue its advanced packaging ponte vechio proves that


Plus intel has bigger l2 cache which is way faster than even 96mb l3

Zen 4 and Raptor Lake will both have 2MB L2 cache.

 

[AlwaysHope]

I'm betting only DDR5 support for RL?, who knows what the costs of a dual channel kit will be then & not at 4800MHz either.
I can do 32GB kit 4800MHz already on my old tech DDR4 kit under RKL!

 

[yeeeeman]

zen 4 will also ONLY support ddr5, so...

 

[AlwaysHope]

zen 4 will also ONLY support ddr5, so...

so? prices & supply had better stabilize by then or a significant number of enthusiasts & gamers will be on DDR4.
The way the world's economy is going atm, I'm not optimistic.