• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD EPYC "Bergamo" Uses 16-core Zen 4c CCDs, Barely 10% Larger than Regular Zen 4 CCDs

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
47,670 (7.43/day)
Location
Dublin, Ireland
System Name RBMK-1000
Processor AMD Ryzen 7 5700G
Motherboard Gigabyte B550 AORUS Elite V2
Cooling DeepCool Gammax L240 V2
Memory 2x 16GB DDR4-3200
Video Card(s) Galax RTX 4070 Ti EX
Storage Samsung 990 1TB
Display(s) BenQ 1440p 60 Hz 27-inch
Case Corsair Carbide 100R
Audio Device(s) ASUS SupremeFX S1220A
Power Supply Cooler Master MWE Gold 650W
Mouse ASUS ROG Strix Impact
Keyboard Gamdias Hermes E2
Software Windows 11 Pro
A SemiAnalysis report sheds light on just how much smaller the "Zen 4c" CPU core is compared to the regular "Zen 4." AMD's upcoming high core-count enterprise processor for cloud data-center deployments, the EPYC "Bergamo," is based on the new "Zen 4c" microarchitecture. Although with the same ISA as "Zen 4," the "Zen 4c" is essentially a low-power, lite version of the core, with significantly higher performance/Watt. The core is physically smaller than a regular "Zen 4" core, which allows AMD to create CCDs (CPU core dies) with 16 cores, compared to the current "Zen 4" CCD with 8.

The 16-core "Zen 4c" CCD is built on the same 5 nm EUV foundry node as the 8-core "Zen 4" CCD, and internally features two CCX (CPU core complex), each with 8 "Zen 4c" cores. Each of the two CCX shares a 16 MB L3 cache among the cores. The SemiAnalysis report states that the dedicated L2 cache size of the "Zen 4c" core remains at 1 MB, just like that of the regular "Zen 4." Perhaps the biggest finding is their die-size estimation, which puts the 16-core "Zen 4c" CCD just 9.6% larger in die-area, than the 8-core "Zen 4" CCD. That's 72.7 mm² per CCD, compared to 66.3 mm² of the regular 8-core "Zen 4" CCD.



The SemiAnalysis report states that the codename AMD assigned to the "Zen 4c" core itself, is "Dionysus," while the 16-core CCD is codenamed "Vindhya." The 128-core/256-thread "Begamo" EPYC 9754 processor is a chiplet-based multi-chip module, designed for existing Socket SP5 server infrastructure. The MCM features no more than eight "Zen 4c" CCDs to achieve its core-count of 128.

The Server I/O Die (sIOD) is built on the 6 nm process, and appears to be the same one found in EPYC "Genoa" processors. It features a 12-channel (24 sub-channel) DDR5 memory interface, and a PCI Express 5.0 x128 root-complex. The EYPC 9754 is a 400 W TDP-class processor, just like the top "Genoa" processor, but with much higher compute density. "Zen 4c" is shaping up to be AMD's answer to Intel's E-cores such as "Gracemont," the article notes.

View at TechPowerUp Main Site | Source
 
I wonder if we will see 24-core heterogeneous CPUs from AMD in the future.
 
I wonder if we will see 24-core heterogeneous CPUs from AMD in the future.
A socket AM5 chip with an 8-core + 3DV cache CCD and a 16-core Zen4c CCD? It's quite possible.
 
What is the actual difference between Zen 4 and Zen 4c?

Cache? Transistor layout designed to minimize space instead of increase clocks?
 
I think you forgot to link to the source. While it's behind a paywall, the first part covering the physical design is free to read. It's an impressive feat of physical design.
TLDR:
  1. reducing the number of timing critical regions to just 4 from well over 10 in Zen 4 as seen in the diagram below: this sacrifices clock speed for density
  2. a new SRAM bitcell developed by TSMC for memories outside L2. As a 6T design, it saves area compared to the usual 8T designs
  3. lower clock speed target allows denser circuits
1686157770423.png


What is the actual difference between Zen 4 and Zen 4c?

Cache? Transistor layout designed to minimize space instead of increase clocks?
The source goes over this in considerable detail.
 
That's interesting. But depending on how fast/these cores are.

It means amd could technically release a 32 core cpu for am5. But how fun thst would be and useful depends on the cores performance. If they suck in games and not really much faster than the 16 core cpu all ready out. It would not be a good solution.
 
That's interesting. But depending on how fast/these cores are.
This is for cloud workloads so it sacrifices clock speed for throughput. If it ever comes to the desktop, then its function would be analogous to Intel's E cores as Zen 4c should be slower than Zen 4 for single threaded or low threaded loads.
 
That's interesting. But depending on how fast/these cores are.

It means amd could technically release a 32 core cpu for am5. But how fun thst would be and useful depends on the cores performance. If they suck in games and not really much faster than the 16 core cpu all ready out. It would not be a good solution.
I'm thinking they probably do know that, and if this is to copy Intel's big little, this would be the little wouldn't it, Soooo, they do probably have something in mind for the Big, Maybe what @btarunr suggested is possible in his personal post above.
 
Almost certainly, if not with zen4 then maybe zen5 but whether they release it in this slagging market is anyone's guess. They don't really need it right now but if sales go down further they might just shut shop on AM4 & go with hybrid &/or xC chips to boost demand.

I wonder if we will see 24-core heterogeneous CPUs from AMD in the future.
Was addressed to both.
This is for cloud workloads so it sacrifices clock speed for throughput. If it ever comes to the desktop, then its function would be analogous to Intel's E cores as Zen 4c should be slower than Zen 4 for single threaded or low threaded loads.
 
I wonder if we will see 24-core heterogeneous CPUs from AMD in the future.
We will probably first see smaller SKUs for mobility segment, as those have been tested in recent months and leaks came out.
For desktop, we have no leaks at the moment, but as the orignal article by Semianalysis shows, 16 c-core CCD is just a tad bigger than current 8 core CCD. So, 8+16 is very possible on current socket, and 16+16 is possible too, though this could require a new socket and new packaging for three CCDs and I/O dies.
 
A socket AM5 chip with an 8-core + 3DV cache CCD and a 16-core Zen4c CCD? It's quite possible.
It will also not be very good unless they can figure out the scheduler.
 
It will also not be very good unless they can figure out the scheduler.
This would be simpler than the current cases. Just prefer the regular Zen 4 CCD until you run out of cores. Of course, if it uses the stacked cache, then that would probably complicate things.
 
A socket AM5 chip with an 8-core + 3DV cache CCD and a 16-core Zen4c CCD? It's quite possible.
I doubt we would see the first heterogenous SKU 8+16 with V cache, but it's all in play for this to happen soon.

What is the actual difference between Zen 4 and Zen 4c?
Cache? Transistor layout designed to minimize space instead of increase clocks?
Did you read the article and the link to original article? It's all there.

It means amd could technically release a 32 core cpu for am5. But how fun thst would be and useful depends on the cores performance. If they suck in games and not really much faster than the 16 core cpu all ready out. It would not be a good solution.
Gaming is irrelevant. You already have up to 16 big cores for gaming. Plenty! Additional CCD with c-cores would primarily boost MT workloads.

This is for cloud workloads so it sacrifices clock speed for throughput. If it ever comes to the desktop, then its function would be analogous to Intel's E cores as Zen 4c should be slower than Zen 4 for single threaded or low threaded loads.
C-cores will be more performant than e-cores because c-cores support HT and all instrucitons, including AVX512, whereas e-cores have a single thread and reduced instructions.
 
C-cores will be more performant than e-cores because c-cores support HT and all instrucitons, including AVX512, whereas e-cores have a single thread and reduced instructions.
I concur; these have none of the drawbacks of the current E cores. They have the same IPC and ISA as Zen 4; the tradeoff is lower clock speeds for lower die rea and presumably lower power.
 
Should be slightly lower IPC than current chips with bigger caches, depending on the workload as well.
If a workload is too large for the 16 MiB L3, then it would have lower IPC than regular Zen 4.
 
Should be slightly lower IPC than current chips with bigger caches, depending on the workload as well.
Not really in the target market. Independent cache is the same amount.

L3 suffers slightly, but most loads will be pretty independent anyways, so that shouldn't matter much.

The main way it could be reduced is the access to the SRAM cells, which could have a slight cache latency penalty.
A socket AM5 chip with an 8-core + 3DV cache CCD and a 16-core Zen4c CCD? It's quite possible.
I imagine it would require a redesign of the package, due to the difference in GMI layout. I would think it would be more likely in the next generation, once the design is tested.


I would be curious to see how performance of Zen 4 and 4c compare at the extremely low power levels available in a laptop. Especially single core performance at the same core cound and power budget. It might have a clock speed advantage sufficient to make up for the lack of L3, in a Dragon Range package, and more then enough in a Phoenix package.
 
That's interesting. But depending on how fast/these cores are.

It means amd could technically release a 32 core cpu for am5. But how fun thst would be and useful depends on the cores performance. If they suck in games and not really much faster than the 16 core cpu all ready out. It would not be a good solution.

I'd prefer a single CCD over a dual CCD all day and night.

But yeah it's there to create a 32 core / 64 thread consumer CPU.
 
would be nice if they made a tiny 8 core ccd for a cheap low power chip
Sounds like they are doing so, in the mobile market.

The trouble is, the CCD is 16 core, so I am not sure if that can be made much cheaper as an 8 core.
 
I just realized that the code names for the Zen 4 and Zen 4c cores are emblematic of their origins. Persephone is the mother of Dionysus in some sources.
 
I concur; these have none of the drawbacks of the current E cores. They have the same IPC and ISA as Zen 4; the tradeoff is lower clock speeds for lower die rea and presumably lower power.
True that. The tradeoff is necessary to compete with ARM designs for cloud server. I would not even call it a "tradeoff". It's a different type of core, like those diverse cores on ARM SoCs. c-cores are much better in performance/watt, which is a fundamental feature in increasing number of systems in variety of segments.

I just realized that the code names for the Zen 4 and Zen 4c cores are emblematic of their origins. Persephone is the mother of Dionysus in some sources.
And riotous son of Zeus :)
 
Back
Top