Question Intel Raptor Lake vs AMD Zen 4 vs Apple M2

[Carfax83]

These CPUs are all going to square off against each other at some point this year assuming nothing catastrophic occurs to delay any of the product launches. So going by what we know from official sources and informed rumor mongers (many of which were very accurate before Alder Lake and the M1 launched), which CPU do you think will win out in these categories?

1) Single threaded performance
2) Multithreaded performance
3) Gaming performance
4) Performance per watt
5) Overall performance (who wins the majority of applications)

While I've been keeping a close eye on rumors and leaks for Zen 4 and Raptor Lake, I have not admittedly been doing so for the M2; as I'm unrepentant Apple hater  At least I'm honest about it... That said, this is my ranking based on what I've seen and heard:

I think the single threaded crown will go to Raptor Lake, and I say this based on informed rumors that Raptor Lake will have up to 10% more IPC from microarchitectural updates, cache upgrades and higher clock speeds than Alder Lake. From what I've seen, gauging IPC performance isn't easy as it varies so much based on application, but I'd say Alder Lake already has at least a 15% across the board IPC advantage over Zen 3, so Raptor Lake could conceivably have 25% better IPC than Zen 3, which is similar to what Zen 4 will reportedly possess. But I doubt Zen 4 will match Raptor Lake in clock speeds and memory latency performance, which is why I'm predicting Raptor Lake will take the single threaded performance crown.

For multithreaded performance, Zen 4 should easily take it due to having more big cores than its Intel counterpart and similar IPC.

Gaming performance is more complicated because while some games are inherently more reliant on single core performance (strategy games for instance), more and more 3D engines are becoming increasingly parallel due to the adoption of Vulkan and DX12 in addition to modernized programming methods. Still, very few 3D engines can scale beyond 8 threads and 6 to 8 cores remains the sweet spot for gaming and will be for some time. So overall, I feel more comfortable going with Raptor Lake for the gaming crown. Also if rumors are correct, Raptor Lake will officially support DDR5-5600 off the bat while Zen 4 will reportedly use DDR5-5200. The raw memory speed won't likely be a significant factor, but Intel's memory controller will be right next to the CPU cores while Zen 4's will be in the I/O die which while still on the same package will definitely incur a significant latency penalty; which I'm sure will be offset by a massive L3 cache.

On performance per watt, one would think the M2 should take this category easily......but from the small amount of research that I've collected on it, it seems that there won't be much of a performance increase with the M2, if any. Some rumors are even suggesting there may be a bit of a regression in that aspect. Also since Zen 4 will be on TSMC's 5nm node, it will undoubtedly have excellent performance per watt and I believe it will also easily crush Apple's best in single core. So for performance per watt, I'm going to go with Zen 4.

When it comes to overall performance, I'm leaning towards Zen 4 but it will be close. Raptor Lake will supposedly double the amount of Gracemont efficiency cores which will certainly help in multithreaded performance per watt, but ultimately they won't be a match for Zen 4's 16 big cores with SMT. AMD will have the core count advantage and when that's combined with IPC parity with Raptor Lake, Zen 4 will win the majority of the benchmarks.

 

[biostud]

They need those watts, cause competing with 24C (32-40 cores in several years) is no joke, they are going to get outnumbered and beaten by atom cores in most MT throughput tests. 16C of Z4 can fight 8GC+24C no better than 8GC could fight 24 gracemonts today.
What happens when Intel reigns in their marketing idiots, iterates on Atom core further, puts it on proper voltage plane? And what is most important, once marketing morons are out, they will clock Gracemonts where they are incredibly efficient like 3.3Ghz or so and efficiency of whole chip suddenly moves ahead big time.

The real problem for AMD seems to be GC core and how strong it is in ST. Scenario where they have to use 5 dies of 5/6nm silicon to fight a single ~240mm^2 chip is not optimal for them and they might choose to not show up in that fight by diverting cache dies to server market.
So far from leaks Z4 does not seem like has many changes. AVX512 + L2 cache to feed said FP units. Without stacked L3 and without AVX512 workloads IPC gains might be dissapointing and below 25% some here are touting.

The next gen rumors for AMD is 8c zen5 + 16 zen4c cores, but depending on what Intel might bring to the table that might change. No matter what, competition is good and as long as it pushes both companies to put new products on the market, and keep the prices competitive, I'm a happy camper.

 

[coercitiv]

What happens when Intel reigns in their marketing idiots, iterates on Atom core further, puts it on proper voltage plane? And what is most important, once marketing morons are out, they will clock Gracemonts where they are incredibly efficient like 3.3Ghz or so and efficiency of whole chip suddenly moves ahead big time.

You probably meant this as a rhetorical question, but the answer may actually surprise you: AMD simply adds one or more chiplets on the mainstream desktop platform. Take a step back and look at what each company is making: Intel is growing a monolithic dinosaur bigger than 220mm2, AMD is baking a ~70mm2 Lego die. It's not Intel who has the actual Raptor, the agile pack hunter is on the other team.

The one and only way for Intel to capitalize on their hybrid approach is for the Coves to dominate Zen cores. They must use that massive transistor budget to break away in ST benchmarks. If that does not happen then the Monts will always be working in damage control mode, independent voltage plane or not.

 

[JoeRambo]

You probably meant this as a rhetorical question, but the answer may actually surprise you: AMD simply adds one or more chiplets on the mainstream desktop platform. Take a step back and look at what each company is making: Intel is growing a monolithic dinosaur bigger than 220mm2, AMD is baking a ~70mm2 Lego die. It's not Intel who has the actual Raptor, the agile pack hunter is on the other team.

I am well aware of this, but they need the rumored Zen5 + small core strategy to properly counter Intel's horde of atoms. 3 chiplets + potentially 3 stacked cache dies + IOD on 5/6nm is in fact a bad counter ( for AMD of course, @biostud and everyone on these forums will rejoice such chips). Let's not loose the focus, we are still talking casual desktop and laptop CPUs, not HEDT.

 

[coercitiv]

Let's not loose the focus, we are still talking casual desktop and laptop CPUs, not HEDT.

Friendly reminder I'm not the one talking about core counts that make little sense for the mainstream desktop.

with 24C (32-40 cores in several years) is no joke, they are going to get outnumbered and beaten by atom cores

Personally I already repeatedly warned that small cores don't make sense on mainstream consumer desktop, small cores become trully advantageous in swarms, really big number of cores... and consumer desktop won't be able to capitalize on that. It's a Catch-22 in all it's beauty.

 

[moinmoin]

I am well aware of this, but they need the rumored Zen5 + small core strategy to properly counter Intel's horde of atoms. 3 chiplets + potentially 3 stacked cache dies + IOD on 5/6nm is in fact a bad counter ( for AMD of course, @biostud and everyone on these forums will rejoice such chips). Let's not loose the focus, we are still talking casual desktop and laptop CPUs, not HEDT.

No, when talking about highest possible MT performance we are not talking casual desktop and laptop CPUs, but the niche high end high wattage desktop and laptop CPUs. The one thing AMD needs to worry about is ADL/RPL's ST performance. The E cores are little more than a distraction, a pretense Intel is competitive with AMD on core count already.

 

[CakeMonster]

The next gen rumors for AMD is 8c zen5 + 16 zen4c cores, but depending on what Intel might bring to the table that might change.

I'd frankly be a bit disappointed by that, I want more than 8 big cores, especially looking some years ahead. The main CCD would be rather small also, unless they are copying intel and making the cores really huge.

 

[AAbattery]

I'd frankly be a bit disappointed by that, I want more than 8 big cores, especially looking some years ahead. The main CCD would be rather small also, unless they are copying intel and making the cores really huge.

I don't look at the Zen 5 hybrid rumor as big and little cores. The 16 Zen 4c cores are rather big by themselves by today's standard, and will probably give similar multi-threaded performance as a 5950X, unless the 8 Zen 5 cores consume too much of the power budget. Those 8 'big' cores are likely to perform roughly 50% faster than a 5800X, plus the Zen 5 server chip rumor has the number of cores per socket increasing even above Bergamo, so a 16 core Zen 5 ccd is likely on its way, although it might take an extra year for something like that to come to AM5.

 

[JoeRambo]

No, when talking about highest possible MT performance we are not talking casual desktop and laptop CPUs, but the niche high end high wattage desktop and laptop CPUs.

Can't agree here. 12900K has nasty TDP, but it is by Intel's marketing decisions. They can choose to reign it in, while loosing minor performance. And those 8+16 and higher chips will be just that, a chip that can go to desktop or 65W laptop.
For AMD they can release AM5 three chiplet solution, no big deal, what about laptop tho? Can't bolt a die to current crop of laptop chips.

Things are gonna get fun, breakneck pace of advance during next 3-4 years.

 

[dr1337]

It's not Intel who has the actual Raptor, the agile pack hunter is on the other team.

Careful, I hear intel is planning to kill the dinosaurs with meteor lake.

 

[Abwx]

I am well aware of this, but they need the rumored Zen5 + small core strategy to properly counter Intel's horde of atoms. 3 chiplets + potentially 3 stacked cache dies + IOD on 5/6nm is in fact a bad counter ( for AMD of course, @biostud and everyone on these forums will rejoice such chips). Let's not loose the focus, we are still talking casual desktop and laptop CPUs, not HEDT.

One Zen 4 core is supposed to roughly match 2 E cores at same frequency, 16 such cores will have no trouble matching 8 P + 16 E at much lower TDP given the process advantage.

RPL will perform something like 20% better at same power than ADL while a Zen 4 based 16C will perform 40% better at isopower than a 5950X wich is already much less power hungry at isoperf than ADL.

 

[diediealldie]

One Zen 4 core is supposed to roughly match 2 E cores at same frequency, 16 such cores will have no trouble matching 8 P + 16 E at much lower TDP given the process advantage.

RPL will perform something like 20% better at same power than ADL while a Zen 4 based 16C will perform 40% better at isopower than a 5950X wich is already much less power hungry at isoperf than ADL.

Intel is planning to put more cache memories(according to rumors), some unlocks in GC and +8 monts so probably they'll perform more than 20% better compared to Alder Lake. If rumors are true then all core clusters(GC or 4x GM) will get 1MB more cache(2MB => 3MB) and twice more L2 cache for GM clusters(2MB => 4MB L2).

3 times more L3 cache memory gave 10~15% boost to Zen 3D so I think we can expect like 6~10% improvements due to cache memory size alone. So I think RPL vs Zen 4 will be similar to ADL vs Zen 5950X. Zen 4 gains will be coming from IPC + process advantage but Intel will somehow manage it 30% ish MT improvements + DLVR power delivery.
Zen 4 will be superior in many ways thanks to homogeneous architecture and superior tech node but Intel will be a much better position than it was with 10900K and 11900K(gigantic HEDT die size sold 500$)

 

[moinmoin]

12900K has nasty TDP, but it is by Intel's marketing decisions. They can choose to reign it in, while loosing minor performance. And those 8+16 and higher chips will be just that, a chip that can go to desktop or 65W laptop.

Intel's hybrid design in ADL just isn't very good. Golden Cove's inane power draw is holding Gracemont's efficiency back. And Gracemont's low peak performance is often lowering MT performance below what Golden Cove alone would be able to manage. Instead managing that in hardware (like doing a coherent design to begin with) Intel tried to solve that in software by letting Microsoft push an unexpected all new Windows version, with the results as we know them.

For AMD they can release AM5 three chiplet solution, no big deal, what about laptop tho? Can't bolt a die to current crop of laptop chips.

Raphael-H will likely be the straight laptop counterpart to the desktop chips.

 

[DrMrLordX]

Intel's hybrid design in ADL just isn't very good.

It appears that Intel can't reliably fabricate dice with significantly more than 8 Golden Cove cores. At least not on a monolithic die. A tiled approach would let them compete with AMD on a more-even footing, but for whatever reason, Intel has chosen to wait until Meteor Lake for tiles.

(also tiles don't seem to be working out all that well for Sapphire Rapids. Why tho?)

 

[Henry swagger]

Raptor Lake will be outclassed by AMD Zen 4. Unsure why people think that adding more cores will make a huge difference when power limits exist. AMD gets to take advantage of a new node. Rumors point to a greater than 20% increase in single core performance, and possibly even more. In addition, AMD may be increasing TDP on some SKUs up to 170W.

Lets wait for benchmarks before you declare victory..

 

[mikk]

Though I cannot but think that RC cores are uninteresting based on what the speculation about caching has indicated so far. Every time there has been a hyping up of cache hierarchies from the Intel camp, it has been a disappointment. Cue Skylake > Skylake-X, Sunny Cove > Willow Cove.

The difference this time is that Intel claims the improved CPU cache improves gaming experience, they never made such direct claims on SKL-X or Willow Cove. SKL-X is a bad example also because overall the cache capacity had been decreased and the mesh design suffered in gaming workloads. Neither a mesh nor a L3 cache decrease comes with Raptor.

 

[Henry swagger]

They need those watts, cause competing with 24C (32-40 cores in several years) is no joke, they are going to get outnumbered and beaten by atom cores in most MT throughput tests. 16C of Z4 can fight 8GC+24C no better than 8GC could fight 24 gracemonts today.
What happens when Intel reigns in their marketing idiots, iterates on Atom core further, puts it on proper voltage plane? And what is most important, once marketing morons are out, they will clock Gracemonts where they are incredibly efficient like 3.3Ghz or so and efficiency of whole chip suddenly moves ahead big time.

The real problem for AMD seems to be GC core and how strong it is in ST. Scenario where they have to use 5 dies of 5/6nm silicon to fight a single ~240mm^2 chip is not optimal for them and they might choose to not show up in that fight by diverting cache dies to server market.
So far from leaks Z4 does not seem like has many changes. AVX512 + L2 cache to feed said FP units. Without stacked L3 and without AVX512 workloads IPC gains might be dissapointing and below 25% some here are touting.

Plus gracemont cores will have more l2 and raptor cove will have more ipc than golden.. amd will be massacred in laptops

 

[Markfw]

Plus gracemont cores will have more l2 and raptor cove will have more ipc than golden.. amd will be massacred in laptops

Lets wait for benchmarks.... Sound familiar ? You have been on this forum for 3 days, and you act like you have been here forever. Not to mention that you are most likely completely WRONG.

 

[tamz_msc]

The difference this time is that Intel claims the improved CPU cache improves gaming experience, they never made such direct claims on SKL-X or Willow Cove. SKL-X is a bad example also because overall the cache capacity had been decreased and the mesh design suffered in gaming workloads. Neither a mesh nor a L3 cache decrease comes with Raptor.

Well Skylake-X might not have been the most appropriate example, but Sunny Cove > Willow Cove resulted in increased latency for both L2 and L3. So unless Intel has found a way to keep latency in check, their claims are a bit vague.

 

[Carfax83]

And whose fault is that? Intel won't produce any consumer CPUs on 10ESF with more than 8 Golden Cove/Raptor Cove cores.

Supposedly they will be resurrecting their HEDT line based on Sapphire Rapids. These CPUs will likely have higher core counts and probably won't have any efficiency cores.

Who said anything about that? Yes the 12600k and 12700k are positioned better overall than the 12900k relative to the competition from last year, but that also wasn't really the topic of discussion . . .

The 12600K and 12700K still pertain to the discussion because they share the same architecture. The point is, that Golden Cove is a very strong architecture that is able to completely dominate Zen 3 in many respects. Code compilation is probably the most salient of them all. Raptor Cove will likely build on Golden Cove's strengths while further mitigating its weaknesses.

The 12600K in that LLVM benchmark I posted a few pages back, is a whopping 75% faster than the 5600x. That's a complete and total ass whooping! And yeah, Zen 3 is an older architecture, but the node its on is on parity with Intel so it's not the total mismatch that was Comet Lake vs Zen 3.

 

[tamz_msc]

Supposedly they will be resurrecting their HEDT line based on Sapphire Rapids. These CPUs will likely have higher core counts and probably won't have any efficiency cores.



The 12600K and 12700K still pertain to the discussion because they share the same architecture. The point is, that Golden Cove is a very strong architecture that is able to completely dominate Zen 3 in many respects. Code compilation is probably the most salient of them all. Raptor Cove will likely build on Golden Cove's strengths while further mitigating its weaknesses.

The 12600K in that LLVM benchmark I posted a few pages back, is a whopping 75% faster than the 5600x. That's a complete and total ass whooping! And yeah, Zen 3 is an older architecture, but the node its on is on parity with Intel so it's not the total mismatch that was Comet Lake vs Zen 3.

Eh, you're conflating the product itself with the characteristics of the P-core when you make the comparison between the 12600K and 5600X. In terms of the latter, Golden Cove is not that exciting considering the extent of changes it has compared to Cypress Cove, in my opinion.

 

[JoeRambo]

So unless Intel has found a way to keep latency in check, their claims are a bit vague.

Latest investigation and simulation by Chips And Cheese outlines some options they could take:

1) For L3 it's obvious that L3 cache clocks are hurting them. AMD also has ring and 32MB ( and now 96MB ) of L3 and they have 10ns cache when Intel deals with ~16-17ns one. So that is an obviuos target of great performance gains. Imagine shaving 7ns from each LLC hit and miss.
There is obviuos problem of 3.7Ghz uncore clock when Atom cores are active also

2) For GC L2, the obviuos "fix" is make it larger, since they are already paying latency tax, might as well make it 2MB. This plays into (1) as less requests will miss into horrible L3 they have.

3) For atom cores they have 2MB and soon will have 4MB of SRAM coupled to L3 slice. Good for performance in both single threaded and especially in multi threaded when cores are fighting for capacity and cache tag conflicts can ruin the day. This would benefit the most crazy workloads like compiles and non-CB renders.

 

[Carfax83]

Eh, you're conflating the product itself with the characteristics of the P-core when you make the comparison between the 12600K and 5600X.

This sentence to me is nonsensical. The reason why I drew the comparison between the 5600x and the 12600K is because they are priced similarly, and the big core count is the same. Actually, going by Newegg prices, the 5600x is more expensive than the 12600K.

In terms of the latter, Golden Cove is not that exciting considering the extent of changes it has compared to Cypress Cove, in my opinion.

It's very rare to see Intel or AMD have such a large gap between them in any workload not factoring core count these days. Cypress Cove looks paltry to me compared with Golden Cove.

 

[coercitiv]

The 12600K in that LLVM benchmark I posted a few pages back, is a whopping 75% faster than the 5600x. That's a complete and total ass whooping!

The benchmarks from a few pages back show 12600K being 50% faster in LLVM while using 27% more power.

With that out of the way @tamz_msc probably had an issue with you extracting conclusions about the strength of Golden Cove from a product that wins against the competition using a larger mix of P+E cores. You compare 6+0 from AMD with 6+4 from Intel and conclude Intel Cove architecture is much better because.... 12600K is priced equally with 5600X. It doesn't get more apples and oranges than this.

 

[Zucker2k]

The benchmarks from a few pages back show 12600K being 50% faster in LLVM while using 27% more power.

With that out of the way @tamz_msc probably had an issue with you extracting conclusions about the strength of Golden Cove from a product that wins against the competition using a larger mix of P+E cores. You compare 6+0 from AMD with 6+4 from Intel and conclude Intel Cove architecture is much better because.... 12600K is priced equally with 5600X. It doesn't get more apples and oranges than this.

Let's match process nodes, cache sizes, and latencies as well, while we're at it. 5950x v 12900k is 32 threads vs 24 threads all day, not to mention the 5950x is more expensive but you don't call out people for comparing to two, not to mention the 5600x is more expensive on Newegg.

 

[DrMrLordX]

Supposedly they will be resurrecting their HEDT line based on Sapphire Rapids.

An HEDT product is a poor substitute for innovation in the consumer sector; furthermore, Sapphire Rapids-X (or whatever they call it) will face off against consumer Raphael. And Raptor Lake ironically enough.

The 12600K and 12700K still pertain to the discussion because they share the same architecture.

Maybe? They aren't as fast as the 5950X in compilation. They're good for people that need compiler performance but can't afford something better, I suppose.

5950x v 12900k is 32 threads vs 24 threads

Intel is welcome to put more than 8 Golden Cove or Raptor Cove cores on their consumer CPUs using tiles, and charge a premium for the product. But they won't do it.