Monday, May 23rd 2022
AMD Unveils 5 nm Ryzen 7000 "Zen 4" Desktop Processors & AM5 DDR5 Platform
AMD today unveiled its next-generation Ryzen 7000 desktop processors, based on the Socket AM5 desktop platform. The new Ryzen 7000 series processors introduce the new "Zen 4" microarchitecture, with the company claiming a 15% single-threaded uplift over "Zen 3" (16-core/32-thread Zen 4 processor prototype compared to a Ryzen 9 5950X). Other key specs about the architecture put out by AMD include a doubling in per-core L2 cache to 1 MB, up from 512 KB on all older versions of "Zen." The Ryzen 7000 desktop CPUs will boost to frequencies above 5.5 GHz. Based on the way AMD has worded their claims, it seems that the "+15%" number includes IPC gains, plus gains from higher clocks, plus what the DDR4 to DDR5 transition achieves. With Zen 4, AMD is introducing a new instruction set for AI compute acceleration. The transition to the LGA1718 Socket AM5 allows AMD to use next-generation I/O, including DDR5 memory, and PCI-Express Gen 5, both for the graphics card, and the M.2 NVMe slot attached to the CPU socket.
Much like Ryzen 3000 "Matisse," and Ryzen 5000 "Vermeer," the Ryzen 7000 "Raphael" desktop processor is a multi-chip module with up to two "Zen 4" CCDs (CPU core dies), and one I/O controller die. The CCDs are built on the 5 nm silicon fabrication process, while the I/O die is built on the 6 nm process, a significant upgrade from previous-generation I/O dies that were built on 12 nm. The leap to 5 nm for the CCD enables AMD to cram up to 16 "Zen 4" cores per socket, all of which are "performance" cores. The "Zen 4" CPU core is larger, on account of more number-crunching machinery to achieve the IPC increase and new instruction-sets, as well as the larger per-core L2 cache. The cIOD packs a pleasant surprise—an iGPU based on the RDNA2 graphics architecture! Now most Ryzen 7000 processors will pack integrated graphics, just like Intel Core desktop processors.
The Socket AM5 platform is capable of up to 24 PCI-Express 5.0 lanes from the processor. 16 of these are meant for the PCI-Express graphics slots (PEG), while four of these go toward an M.2 NVMe slot attached to the CPU—if you recall, Intel "Alder Lake" processors have 16 Gen 5 lanes toward PEG, but the CPU-attached NVMe slot runs at Gen 4. The processor features dual-channel DDR5 (four sub-channel) memory, identical to "Alder Lake," but with no DDR4 memory support. Unlike Intel, the AM5 Socket retains cooler compatibility with AM4, so the cooler you have sitting on your Ryzen CPU right now, will work perfectly fine.
The platform also puts out up to 14 USB 20 Gbps ports, including type-C. With onboard graphics now making it to most processor models, motherboards will feature up to four DisplayPort 2 or HDMI 2.1 ports. The company will also standardize Wi-Fi 6E + Bluetooth WLAN solutions it co-developed with MediaTek, weaning motherboard designers away from Intel-made WLAN solutions.
At its launch, in Fall 2022, AMD's AM5 platform will come with three motherboard chipset options—the AMD X670 Extreme (X670E), the AMD X670, and the AMD B650. The X670 Extreme was probably made by re-purposing the new-generation 6 nm cIOD die to work as a motherboard chipset, which means its 24 PCIe Gen 5 lanes work toward building an "all Gen 5" motherboard platform. The X670 (non-extreme), is very likely a rebadged X570, which means you get up to 20 Gen 4 PCIe lanes from the chipset, while retaining PCIe Gen 5 PEG and CPU-attached NVMe connectivity. The B650 chipset is designed to offer Gen 4 PCIe PEG, Gen 5 CPU-attached NVMe, and likely Gen 3 connectivity from the chipset.
AMD is betting big on next-generation M.2 NVMe SSDs with PCI-Express Gen 5, and is gunning to be the first desktop platform with PCIe Gen 5-based M.2 slots. The company is said to be working with Phison to optimize the first round of Gen 5 SSDs for the platform.
All major motherboard vendors are ready with Socket AM5 motherboards. AMD showcased a handful, including the ASUS ROG Crosshair X670E Extreme, the ASRock X670E Taichi, MSI MEG X670E ACE, GIGABYTE X670E AORUS Xtreme, and the BIOSTAR X670E Valkyrie.
AMD is working to introduce several platform-level innovations like it did with Smart Access Memory with its Radeon RX 6000 series, which builds on top of the PCIe Resizable BAR technology by the PCI-SIG. The new AMD Smart Access Storage technology builds on Microsoft DirectStorage, by adding AMD platform-awareness, and optimization for AMD CPU and GPU architectures. DirectStorage enables direct transfers between a storage device and the GPU memory, without the data having to route through the CPU cores. In terms of power delivery Zen 4 uses the same SVI3 voltage control interface that we saw introduced on the Ryzen Mobile 6000 Series. For desktop this means the ability to address a higher number of VRM phases and to process voltage changes much faster than with SVI2 on AM4.
Taking a closer look at the AMD Footnotes, "RPL-001", we find out that the "15% IPC gain" figure is measured using Cinebench and compares a Ryzen 9 5950X processor (not 5800X3D), on a Socket AM4 platform with DDR4-3600 CL16 memory, to the new Zen 4 platform running at DDR5-6000 CL30 memory. If we go by the measurements from our Alder Lake DDR5 Performance Scaling article, then this memory difference alone will account for roughly 5% of the 15% gains.
The footnotes also reference a "RPL-003" claim that's not used anywhere in our pre-briefing slide deck, but shown in the video presentation. In the presentation we're seeing a live demo comparison between a "Ryzen 7000 Series" processor and Intel's Core i9-12900K "Alder Lake." It's worth mentioning here that AMD isn't disclosing the exact processor model, only that it's a 16-core part, if we follow the Zen 3 naming, that would probably be the Ryzen 9 7950X flagship. The comparison runs the Blender rendering software, which loads all CPU cores. Here we see the Ryzen 7000 chip finish the task in 204 seconds, compared to the i9-12900K and its 297 seconds time, which is a huge 31% difference—very impressive. It's worth mentioning that the memory configurations are slightly mismatched. Intel is running with DDR5-6000 CL30, whereas the Ryzen is tested with DDR5-6400 CL32—lower latency for Intel, higher MHz for Ryzen. While ideally we'd like to see identical memory used, the differences due to the memory configuration should be very small.
AMD is targeting a Fall 2022 launch for the Ryzen 7000 "Zen 4" desktop processor family, which would put this sometime between September thru October. The company is likely to detail the "Zen 4" microarchitecture and the Ryzen 7000 SKU list in the coming weeks.
Update 21:00 UTC: AMD has clarified that the 170 W PPT power numbers seen are the absolute max limits, not "typical" like the 105 W, on Zen 3, which were often exceeded during heavy usage.
Update May 26th: AMD further clarified that the 170 W number is "TDP", not "PPT", which means that when the usual x1.35 factor is applied, actual power usage can go up to 230 W.
You can watch the whole presentation again at YouTube:
Much like Ryzen 3000 "Matisse," and Ryzen 5000 "Vermeer," the Ryzen 7000 "Raphael" desktop processor is a multi-chip module with up to two "Zen 4" CCDs (CPU core dies), and one I/O controller die. The CCDs are built on the 5 nm silicon fabrication process, while the I/O die is built on the 6 nm process, a significant upgrade from previous-generation I/O dies that were built on 12 nm. The leap to 5 nm for the CCD enables AMD to cram up to 16 "Zen 4" cores per socket, all of which are "performance" cores. The "Zen 4" CPU core is larger, on account of more number-crunching machinery to achieve the IPC increase and new instruction-sets, as well as the larger per-core L2 cache. The cIOD packs a pleasant surprise—an iGPU based on the RDNA2 graphics architecture! Now most Ryzen 7000 processors will pack integrated graphics, just like Intel Core desktop processors.
At its launch, in Fall 2022, AMD's AM5 platform will come with three motherboard chipset options—the AMD X670 Extreme (X670E), the AMD X670, and the AMD B650. The X670 Extreme was probably made by re-purposing the new-generation 6 nm cIOD die to work as a motherboard chipset, which means its 24 PCIe Gen 5 lanes work toward building an "all Gen 5" motherboard platform. The X670 (non-extreme), is very likely a rebadged X570, which means you get up to 20 Gen 4 PCIe lanes from the chipset, while retaining PCIe Gen 5 PEG and CPU-attached NVMe connectivity. The B650 chipset is designed to offer Gen 4 PCIe PEG, Gen 5 CPU-attached NVMe, and likely Gen 3 connectivity from the chipset.
AMD is working to introduce several platform-level innovations like it did with Smart Access Memory with its Radeon RX 6000 series, which builds on top of the PCIe Resizable BAR technology by the PCI-SIG. The new AMD Smart Access Storage technology builds on Microsoft DirectStorage, by adding AMD platform-awareness, and optimization for AMD CPU and GPU architectures. DirectStorage enables direct transfers between a storage device and the GPU memory, without the data having to route through the CPU cores. In terms of power delivery Zen 4 uses the same SVI3 voltage control interface that we saw introduced on the Ryzen Mobile 6000 Series. For desktop this means the ability to address a higher number of VRM phases and to process voltage changes much faster than with SVI2 on AM4.
Update 21:00 UTC: AMD has clarified that the 170 W PPT power numbers seen are the absolute max limits, not "typical" like the 105 W, on Zen 3, which were often exceeded during heavy usage.
Update May 26th: AMD further clarified that the 170 W number is "TDP", not "PPT", which means that when the usual x1.35 factor is applied, actual power usage can go up to 230 W.
You can watch the whole presentation again at YouTube:
211 Comments on AMD Unveils 5 nm Ryzen 7000 "Zen 4" Desktop Processors & AM5 DDR5 Platform
What happened to 3D-CACHE... not incorporated on Zen 4? i didnt see any mention or am i missing something Or is AMD to use this feature on one-off refresh novelty chips?
For me, the graphics integration is a BIG +1. On my personal gaming/work build I've never opted for anything otherwise... i like the idea of having a trouble-shoot iGPU, needed more than ever with these crazy power consuming modern graphics cards and their higher than usual fail rate.
"a job finished in 1s" is 20000% faster than "a job finished in 200s"
Now did you learn anything from that?
I'm guesstimating the bigger cache variants would likely ditch the IGP with massive (L3?) caches near the cores or on the IoD, maybe even an L4 cache.
>>It is the other way around
Apologies, apparently I cannot type and nest quotes properly. Valantar was right.
Raptor lake just doubles the e-cores (so as most real-world loads hit a scaling wall, Raptor lake will also hit that same scaling wall earlier than Zen 4 (8P + 16e versus 16 P!)
it's going to take a perfectly-scaling application for Raptor Lake to rape 7950!
"a job finished in 1s" is how many % faster than "a job finished in 200s" ?
Answer that with common sense and compare with Valantar's formula.
I am tired explaining primary school maths online.Same
www.pcgamesn.com/intel/raptor-lake-40-percent-faster
With Raptor Lake, Intel is said to be improving the performance of these cores, and the leaked road map published by Videocardz suggests that we should see “new hybrid CPU core changes for improved performance” as well as “improved CPU cache for gaming” improvements for desktop Raptor Lake. It’s unclear what these changes will be at this time, however.
www.digitaltrends.com/computing/intel-raptor-lake-cpu-rumors-news-specs/
Pretty sure they are not "just" adding more E cores
wccftech.com/roundup/intel-13th-gen-raptor-lake-cpus/
With that said, my own 5950X can stock boost up to 5.05 ghz on the best cores. With some tingering in bios the 3 best cores can go to 5.25 ghz and all cores capable of reaching 5 ghz or above at light load. Clock Will be lower at high load and anything above 5.25 ghz and it crashes at just the slides load, it can boot but not handle much stress over 5.25 ghz. By the way my cpu is aircooled, so it doesn't need exotic cooling to reach these clocks, just a good motherboard.
But a 5950X3D and I would be more than happy and dosent need a new socket. But I don't know, guess a 3d version dosent make sense as v-cashe mostly benefits games only.
Guess it's for some thermal advantages, right?
TL:DR Valantar was right. Go Away and think about this.
PS I also like beer. Try it, it helps.
I mean, this is even included in your hackneyed reformulation, which tries to avoid this by reformulating the variables in question to "units of work per time" (which might be zero, but only at zero work done) rather than "time spent per one unit of work", which is what the slide here (and nearly all such benchmarks) presents.
You're arguing as if it's better if a change between, say, 20s and 10s compared to a 200s baseline were presented as "10x faster" and "20x faster", despite the fact that this grossly exaggerates the difference between the two. You see that, right? Presenting those two as 90% and 95% faster is a far more accurate representation of their absolute time expenditure.But all of those are still true. Your "in reality" statements, which seem to be meant as rebuttals, are literally the same ratio. They're saying the same thing. And all are equally valid - but which is more useful or appropriate depends on context, of course. And the context is what you're misapplying here. The context is not a question looking for a rate of work, but a time till completion of work. And in terms of marketing, the application you're arguing for is one that exaggerates the actual improvements. When you're comparing two things to see how fast they can finish a task, it's the reduction in task completion time that matters, not the fact that a 100% reduction is impossible. That's just how the world works.
"AI compute acceleration" is vague enough to mean precisely nothing - why would consumers care?
"up to 24 PCI-Express 5.0 lanes from the processor" - that's not nearly as many as I was hoping, it's the same number as ADL (I'm including the latter's chipset DMI link here). Granted, 8 PCIe 5.0 lanes are superior to 8 4.0 lanes, but if AMD has to spend 4 lanes on the chipset(s) then you're back to parity with ADL. Which means it should be easy for RKL to match or even exceed this count.
"the AM5 Socket retains cooler compatibility with AM4" - I wonder how many idiots are going to reuse their shitty $20 tower coolers on Zen 4 CPUs, then complain the CPUs are slow because they throttle.
"up to 14 USB 20 Gbps ports" - lovely marketing weasel-words, you still need an entire lane of PCIe 5.0 or 2 lanes of 4.0 to reach 20Gbps. Unless the chipset(s) themselves are 5.0-capable, which I strongly doubt due to cost implications, they will need to have a shitton of 4.0 lanes to be able to provide that level of USB connectivity. I'm expecting the same thing that we saw on X370, namely one or two USB-C ports at the highest speed and the rest still being ye olde 3.1 gen 1 type-A.
No explicit mention of USB4 anywhere, which is ominous. I can't imagine AMD would be stupid enough to launch a platform that lacks USB4, but also... AMD.
"will also standardize Wi-Fi 6E + Bluetooth WLAN solutions it co-developed with MediaTek, weaning motherboard designers away from Intel-made WLAN solutions" - that MediaTek-branded solution will have to be incredibly good to pull board manufacturers away from Intel's tried-and-true WiFi hardware. My fear is that it'll instead be incredibly cheap and instead of having decent Intel WiFi on everything, we'll get crappy MediaTek on lower-end SKUs and need to pay more for Intel on the better ones.
"AMD is betting big on next-generation M.2 NVMe SSDs with PCI-Express Gen 5" - nobody cares, PCIe 4.0 SSDs are stupidly fast already, no ordinary consumer wants or needs 5.0 SSDs, what they want and need are cheaper and more energy-efficient SSDs. Console tards will lap this up though.
"The new AMD Smart Access Storage technology builds on Microsoft DirectStorage" - something else nobody cares about.
Don't get me wrong, i'm ECSTATIC about AMD being in THE GAME again and insanely happy about the competition;
also LOVING the new AM5/DDR5/Pci-E5.0 stuff.
Now i'll just enjoy this AM4/DDR4/Pci-E4.0 system for the next 5/6 years while at the same time watching the AMD/INtel/nGreedia wars.
Next upgrade is bound for 2027/2028, unless smth goes poof outside the warranty period but even then i'm not switching platform.
They need a 3D cache version ASAP or AMD is AMDead.
7000 a series is gonna lose to LoveLace as well
Your point beeing?
We( consumers) still win, or should win.
I'm still gaming at 1080p 144Hz, had a RX 5700 XT $385 in 2020 including shipped from 2000miles, under ran it by both power (sometimes at 50% for the kicks of it) and gpu clocks / gpu Tension supply , card mustered trough exceding expectations and then some.
A pretty much similar story( mind the price) after upgrading to a RX 6900XT, heavily under run , not much heat output , fans speed set on minimum( only because I currently find it better compared to the fan stop feature).
It might of well been a couple of nVidia based gpu graphics cards my last two, but it wasnt.
Notable mention : PSU rated 550 Watts.
I dont see the issue(s) here. New platform with future CPU releases.
They need some time, but they will come.
PCi-E 5.0 is'nt a requirement either...Theres no card taxing even PCI-E 4.0.
It's quite likely that Zen4c has less L3 cache, yes - that's one of the easiest ways of cutting down on area. But it's also likely tweaked in other ways - just like Zen2 was significantly smaller than Zen3 on the same process, Zen4 is another die size increase, so Zen4c might be closer to Zen2/3 in various areas to keep it slim. It's meant for applications where the sheer number of threads matters far more than their absolute peak performance after all, so some concessions are expected.We'll see. If that's 15% IPC, that's okay, if that's 15% including the clock speed increase it's a let-down for sure. Acceptable overall performance boost, barely, but only through pushing clocks ridiculously high, which kills efficiency.Yep. Guess they're matching Intel there though; they've been advertising the same for at least the past generation.It's not the same as ADL - ADL has 5.0 x16 PEG and 5.0 for the chipset (IIRC), but no 5.0 m.2. Not that 4 lanes less matters much, but ADL prioritizing 5.0 for GPUs rather than storage never made sense in the first place - it's doubtful any GPU in the next half decade will be meaningfully limited by PCIe 4.0 x16.... is that any more likely than them buying a shitty $20 AM5 tower cooler? There are plenty of great AM4 coolers out there after all. Retaining compatibility reduces waste in a meaningful and impactful way. You don't fix people being stupid by forcing obsolescence onto fully functional parts.X670E is literally marketed as "PCIe 5.0 everywhere", providing 24 more lanes of 5.0 (and, presumably, another 4 of which go to the CPU interconnect, leaving a total of 40). X670 most likely retains the 5.0 chipset uplink even if it runs its PCIe at 4.0 speeds. The main limitation to this is still the cost of physically implementing this amount of high speed IO on the motherboard, as that takes a lot of layers and possibly higher quality PCB materials.Several announced motherboards mention it explicitly, so no need to worry on that front. The only unknown is whether it's integrated into the CPU/chipset or not. Support is there.This is mostly true, and I agree that PCIe 5.0 SSDs are pretty dumb, but that's how competition works in tech - if your competitor has a feature, you need another feature on top of that again.On this I'd have to disagree with you. DS has a lot of potential - current software just can't make use of our blazing fast storage, and DS goes a long way towards fixing that issue. It just needs a) to be fully implemented, with GPU decompression support, and b) be adopted by developers. The latter is pretty much a given for big name titles given that it's an Xbox platform feature though.Depends how that increase is reached, and whether the same thing is maintainable in MT. If it's only from pushing clocks and that means increasing power, it might not. If it's from improved efficiency and IPC, most likely yes. But there's tons of gray area and nuance.