Update 5/13 9:00 am PT: It appears that the listing on OpenBenchmarking.org has now disappeared.
Original Article:
One of AMD's approaching Ryzen 7000 (Raphael) processors appears to have surfaced. Twitter user Petykemano has unearthed the mysterious 5nm chip hidden deep inside the OpenBenchmarking.org database.
The Ryzen 7000 processor, which goes by the 100-000000666-20_Y identifier, wields eight Zen 4 cores with simultaneous multithreading (SMT). We had already seen a Zen 4 chip with a similar identifier back in January. Assuming that AMD doesn't rework its Ryzen hierarchy, the eight-core, 16-thread part should be the Ryzen 7 7800X, the successor to the popular Ryzen 7 5800X. During its CES 2022 keynote, AMD teased a Ryzen 7000 chip with an all-core boost clock over 5 GHz in Halo Infinite.
The Zen 4 chip from the OpenBenchmarking submission is an engineering sample, but that didn't make it less impressive, as the Zen 4 part boasts a 5.21 GHz boost clock. That's higher than every single chip on AMD's current mainstream roster. For example, the flagship Ryzen 9 5950X has a 4.9 GHz boost clock, whereas the Ryzen 7 5800X maxes out at 4.7 GHz. In addition, Raphael proves what AMD can manage with Zen 4 on TSMC's 5nm manufacturing process. The Taiwanese foundry claims that the N5 node boasts a 30% improvement in power efficiency or 15% higher performance with a 1.8X increase in density than the 7nm node used for AMD's existing Ryzen 5000 (Vermeer) lineup.
It's a shame that the anonymous user didn't test the Zen 4's processing prowess. Still, the person did at least help lend some credence to the rumor that AMD's next-generation Ryzen processors could sport integrated graphics, more specifically RDNA 2. According to the benchmark software, the Zen 4 chip has a graphics engine called "GFX1036" that runs between 1,000 MHz to 2,000 MHz. The clock speed looks reasonable, and there may be some headroom for improvement. AMD's Radeon 680M, which powers Ryzen 6000, tops out at 2,400 MHz.
As a refresher, AMD uses the GFX1030-series ID to refer to its RDNA 2 graphics engine, such as the Radeon RX 6000-series (Big Navi) graphics cards or the iGPU inside the Ryzen 6000 (Rembrandt) processors. Therefore, there is no doubt that GFX1036 belongs to the RDNA 2 family. Furthermore, as per a Phoronix report, AMD recently added the GFX1036 and GFX1037 graphics IP blocks to the RadeonSI Linux driver. The treatment for GFX1036 and GFX1037 is the same as AMD's other products, including Van Gogh APUs, Navi 24 (Beige Goby), and Rembrandt (Yellow Carp) mobile chips.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
The octa-core Zen 4 processor resided on a Splinter-RPL (WS22427N000 BIOS) system, which could be the codename for the reference AM5 platform. Sadly, the report only alluded to 16GB of memory. Still, various sources have told us that the AM5 platform will only support DDR5 memory, turning AM5 into a premium platform when considering the steep pricing for early DDR5 memory kits. On the other hand, Intel gives consumers a choice to pick between DDR4 and DDR5 with its 12th Generation Alder Lake chips.
AMD's latest roadmap confirmed that the Ryzen 7000 would be here before the end of the year. Dragon Range, which could be the mobile variant of Raphael, will not land until 2023, though.
Zhiye Liu is a news editor and memory reviewer at Tom’s Hardware. Although he loves everything that’s hardware, he has a soft spot for CPUs, GPUs, and RAM.
AMD crafts custom EPYC CPU with HBM3 memory for Microsoft Azure – CPU with 88 Zen 4 cores and 450GB of HBM3 may be repurposed MI300C, four chips hit 7 TB/s
AMD-powered El Capitan is now the world's fastest supercomputer with 1.7 exaflops of performance — fastest Intel machine falls to third place on Top500 list
-
dalauder So all this article said was that the previous claim of an 8c/16t Ryzen is accurate?Reply
I mean, I'm an AMD fanboy pretty much, but saying "Ryzen 7000 packs serious firepower" in the title is disingenuous journalism if your article has no performance content. "It's a shame that the anonymous user didn't test the Zen 4's processing prowess. " That sentence confirms the lack of actual content to support the title claim in the article.
Clarification: I know, all journalism is bait-and-switch or overblown jargon, but I'd prefer it had content like this was 2008. -
Neilbob dalauder said:So all this article said was that the previous claim of an 8c/16t Ryzen is accurate?
I mean, I'm an AMD fanboy pretty much, but saying "Ryzen 7000 packs serious firepower" in the title is disingenuous journalism if your article has no performance content. "It's a shame that the anonymous user didn't test the Zen 4's processing prowess. " That sentence confirms the lack of actual content to support the title claim in the article.
Clarification: I know, all journalism is bait-and-switch or overblown jargon, but I'd prefer it had content like this was 2008.
That's the way* of the world now, unfortunately.
And if (for example) Alder Lake happens to be 2% faster overall at 100% more power consumption, then we'll get something along the lines 'Core i9 Demolishes Zen 4 in new test!' -
VforV
Exactly, spot on!Neilbob said:That's the of the world now, unfortunately.
And if (for example) Alder Lake happens to be 2% faster overall at 100% more power consumption, then we'll get something along the lines 'Core i9 Demolishes Zen 4 in new test!'
Though if the leaks are true I expect the opposite, a 5% better Zen4 over Raptor Lake.
The question is: will the titles exaggerate in favor of AMD too, saying it "demolishes" intel?
I mean we can all like a company more than another (and fanboi all we want), but can we not be rational and have common sense too? +5% or less does not mean "demolishes" or "destroys" or whatever silly exaggeration is used these days. A simple "wins" or "beats" will suffice without exaggerating...
Tech journalism professionalism and integrity on most sites and YT channels has gotten to a pathetic low level these days. :cautious: -
jp7189
I hadn't seen 5.2GHz reported previously. Thats a healthy bump over the 5800x's 4.7GHzdalauder said:So all this article said was that the previous claim of an 8c/16t Ryzen is accurate?
I mean, I'm an AMD fanboy pretty much, but saying "Ryzen 7000 packs serious firepower" in the title is disingenuous journalism if your article has no performance content. "It's a shame that the anonymous user didn't test the Zen 4's processing prowess. " That sentence confirms the lack of actual content to support the title claim in the article.
Clarification: I know, all journalism is bait-and-switch or overblown jargon, but I'd prefer it had content like this was 2008. -
daredevil01
The frequency boost alone isn't as impressive as the IPC improvement. I hope AMD doesn't lose its efficiency advantage in favour of being 9% better than Intel instead of 3%.jp7189 said:I hadn't seen 5.2GHz reported previously. Thats a healthy bump over the 5800x's 4.7GHz -
jp7189
I haven't seen numbers re IPC improvements. Have they changed much in the core architecture? I was kinda thinking zen4 would be a die shrink, speed bump with big platform upgrade.daredevil01 said:The frequency boost alone isn't as impressive as the IPC improvement. I hope AMD doesn't lose its efficiency advantage in favour of being 9% better than Intel instead of 3%.