Question AMD Rembrandt/Zen 3+ APU Speculation and Discussion

[izaic3]

Alright, so we've had some leaks so far. I don't know if any of it's been confirmed yet, as it's pretty early, but here is what I've surmised so far (massive grain of salt of course):

• Zen 3 "enhanced"
• DDR5 and LPDDR5 support
• PCIE4
• 6nm
• CVML
• Navi 2/RDNA/RDNA 2
• 12 CU
• Socket FP7/AM5
• Possible USB4
• 2021(unlikely) -2022 (more likely) release date

If if turns out to have RDNA 2 and 12 CU, I could see iGPU performance potentially almost doubling over Cezanne.

If I've made any mistakes or gotten anything wrong, please let me know. I'd also love to hear more knowledgeable people weigh in on their expectations.

 

[eek2121]

5800U DDR4-3200 22-22-22 vs 6800U LPDDR5-6400 19-15-17
5800U 15W vs 6800 28W
Zero about the Iris Xe device, probably DDR4

https://www.amd.com/en/products/ryzen-processors-laptop

But it's AMD, it's fine. If Intel does this the forum explodes.

Fail to see the issue.

Also, those timings.

 

[Abwx]

5800U DDR4-3200 22-22-22 vs 6800U LPDDR5-6400 19-15-17
5800U 15W vs 6800 28W
Zero about the Iris Xe device, probably DDR4

https://www.amd.com/en/products/ryzen-processors-laptop

But it's AMD, it's fine. If Intel does this the forum explodes.

A 28W 5800U perform exactly 28% better that it does at 15W, so it s unlikely that their 5800U was ran at 15W, it would mean that RMB is also 7nm with no CPU MT perf enhancement.

 

[mikk]

A 28W 5800U perform exactly 28% better that it does at 15W, so it s unlikely that their 5800U was ran at 15W, it would mean that RMB is also 7nm with no CPU MT perf enhancement.

The CPU comparison is useless until we get the real power consumption, assuming 15W vs 28W wasn't. This is a really bad presentation. And of course they should have used a LPDDR4x device for 5800U, there are some Lenovo Yoga and Thinkpad devices in the market.

 

[DrMrLordX]

Does anyone even sell 5800u devices with DDR4-3200 CAS/CL 22?

 

[Shivansps]

So in desktop, Zen4 is Ryzen 7000 and launching in H2 along with AM5, but they skipped the 6000 series that are likely reserved for RMB, i think it is safe to say desktop RMB by the end of the year, if it ever comes out.

 

[Abwx]

The CPU comparison is useless until we get the real power consumption, assuming 15W vs 28W wasn't. This is a really bad presentation. And of course they should have used a LPDDR4x device for 5800U, there are some Lenovo Yoga and Thinkpad devices in the market.

Fortunately they used Cinebench, wich is straightforward for comparison of a same or so design.

A 6nm fabbed theorical 5800U ran at 28W would be at least 40% faster in MT than a 7nm based one fed with 15W, so it s likely that the 5800U was tested at its usual 25W cTDP, still a 3W advantage for RMB but not that significant overall.

 

[mikk]

Does anyone even sell 5800u devices with DDR4-3200 CAS/CL 22?

Yes there are a lot of DDR4 devices and CL22 is standard for DDR4-3200 sodimm. The odd part is 19-15-17 on LPDDR5-6400, how is this even possible? Maybe it's doubled and it's CL38 in real. AMD really failed on this.

 

[insertcarehere]

Fortunately they used Cinebench, wich is straightforward for comparison of a same or so design.

AMD's footnotes here literally says:

Based on testing by AMD as of 12/14/2021. CPU performance evaluated with a geomean of 9 multi-threaded content creation and CPU tests.

That's not "just Cinebench".

Frankly, the greater concern is the 6800u Laptop's use of very aggressive LPDDR5 modules, which is 99% not going to translate to laptop OEMs, and thus put a damper to claimed gpu gains.

 

[eek2121]

Does anyone even sell 5800u devices with DDR4-3200 CAS/CL 22?

Unsure about the U series, but my H series laptop has it.

 

[Abwx]

AMD's footnotes here literally says:


That's not "just Cinebench".

Separate numbers for Cinebench...

AMD Keynote and Dr. Lisa Su at CES 2022: Live Blog (7am PT, 15:00 UTC)

www.anandtech.com

And again :

AMD Ryzen 6000 Mobile: Rembrandt bringt Zen 3+, RDNA 2, DDR5 und vieles mehr

Aufgewertete Zen-3-Kerne, DDR5, RDNA-2-Grafik, TSMC-N6-Fertigung und eine neue Plattform, das ist Rembrandt alias AMD Ryzen 6000 Mobile.

www.computerbase.de

 

[leoneazzurro]

I think the one big thing nobody is spaeaking about is not performance, but battery life. Now, 24h is certainly not the average result, but the comparisons with 5800U in workloads are quite good fir such a powerful APU.

 

[eek2121]

The CPU comparison is useless until we get the real power consumption, assuming 15W vs 28W wasn't. This is a really bad presentation. And of course they should have used a LPDDR4x device for 5800U, there are some Lenovo Yoga and Thinkpad devices in the market.

So maybe you are confused about cTDP.

The 5800U had a 10-25W cTDP: AMD Ryzen™ 7 5800U | AMD
The 6800U has a cTDP of 15-28W: AMD Ryzen™ 7 6800U | AMD

 

[TESKATLIPOKA]

If they compared 5800U 15W vs 6800U 28W then the increase in Cinebench is not surprising. If It was 25W vs 28W then there would be barely any difference.

R7 5800U at 25W is 24% faster than at 15W.

 

[DrMrLordX]

Yes there are a lot of DDR4 devices and CL22 is standard for DDR4-3200 sodimm.

Interesting. I thought CL20 was the typical JEDEC standard used by OEM machines. Maybe that's more for standard DIMMs. The LPDDR5 timings are odd. Surely something isn't right there.

 

[Abwx]

If they compared 5800U 15W vs 6800U 28W then the increase in Cinebench is not surprising. If It was 25W vs 28W then there would be barely any difference.

R7 5800U at 25W is 24% faster than at 15W.

You forgot that RMB is 6nm fabbed; so i guess that you re saying that this node didnt bring any perf/Watt improvement, as well as AMD engineers being so lazy that they didnt even bother to improve the design.

 

[insertcarehere]

I think the one big thing nobody is spaeaking about is not performance, but battery life. Now, 24h is certainly not the average result, but the comparisons with 5800U in workloads are quite good fir such a powerful APU.

That's because the battery claims aren't test results, but projections from AMD themselves:

Engineering projections are not a guarantee of final performance. Performance projection by AMD engineering staff based on expected battery life for idle, video playback, modern standby, and esports gaming vs. Ryzen 5000 series mobile processors with the same TDP. Power reduction projections based on pre-silicon engineering analysis . Estimated battery life increases assume identically HW & SW configured systems. Specific projections are based on reference design platforms and are subject to change when final products are released in market. RMB-3

 

[izaic3]

MX450 Doa?

 

[jpiniero]

MX450 Doa? View attachment 55427

That's why nVidia did the mx570 and 2050L.

 

[Abwx]

That's why nVidia did the mx570 and 2050L.

Are you sure that it s not rather because of this :

AMD Radeon RX 6000 mobile specifications leak from the RX 6300M to the RX 6850M XT

A substantial leak has revealed specifications for eight upcoming AMD Radeon RX 6000M laptop GPUs. All based on the company's RDNA 2 architecture, the GPUs will range from the RX 6300M to the RX 6850M XT. AMD will also release the RX 6000S series, seemingly an equivalent to NVIDIA's old Max-Q GPUs.

www.notebookcheck.net

 

[tamz_msc]

5800U DDR4-3200 22-22-22 vs 6800U LPDDR5-6400 19-15-17
5800U 15W vs 6800 28W
Zero about the Iris Xe device, probably DDR4

https://www.amd.com/en/products/ryzen-processors-laptop

But it's AMD, it's fine. If Intel does this the forum explodes.

The Intel test platform was a Dell XPS 9310 with LPDDR4x, but it doesn't matter; 2x faster than 5800U means its barely catching up to MX450.

 

[tamz_msc]

MX450 Doa? View attachment 55427

Nah, look at the Witcher 3 result. That's the only one which is actually GPU-limited, rest are all CPU-limited.

 

[leoneazzurro]

That's because the battery claims aren't test results, but projections from AMD themselves:

Projections based on their internal testing with preproducion silicon and identical configurations, of course they cannot guarantee for an OEM.

 

[jpiniero]

Are you sure that it s not rather because of this :

AMD Radeon RX 6000 mobile specifications leak from the RX 6300M to the RX 6850M XT

A substantial leak has revealed specifications for eight upcoming AMD Radeon RX 6000M laptop GPUs. All based on the company's RDNA 2 architecture, the GPUs will range from the RX 6300M to the RX 6850M XT. AMD will also release the RX 6000S series, seemingly an equivalent to NVIDIA's old Max-Q GPUs.

www.notebookcheck.net

I suppose the 2050L could be a response to Navi 24.

 

[insertcarehere]

Projections based on their internal testing with preproducion silicon and identical configurations, of course they cannot guarantee for an OEM.

Again, the footnotes from AMD themselves say:

Performance projection by AMD engineering staff based on expected battery life for idle, video playback, modern standby, and esports gaming vs. Ryzen 5000 series mobile processors with the same TDP. Power reduction projections based on pre-silicon engineering analysis.

How does the above somehow translate into internal testing with actual silicon? Especially when the performance footnotes with testing are significantly more specific.

 

[TESKATLIPOKA]

You forgot that RMB is 6nm fabbed; so i guess that you re saying that this node didnt bring any perf/Watt improvement, as well as AMD engineers being so lazy that they didnt even bother to improve the design.

So you are saying that HP ProBook 635 Aero G8 with R7 5800U was actually set to 25W even though they stated in footnotes:

Based on testing by AMD as of 12/14/2021. CPU performance evaluated with a geomean of 9 multi-threaded content creation and CPU tests. GPU performance evaluated with 3DMark® Time Spy. System configuration for Ryzen™ 7 5800U CPU/GPU performance: HP ProBook 635 Aero G8 configured with 2x8GB DDR4-3200 (22-22-22), Windows® 11 22000.282, Samsung 980 Pro 1TB SSD, 15W nominal processor TDP, GPU driver 27.20.21026, BIOS T83. System configuration for Ryzen™ 7 6800U CPU/GPU performance: AMD reference motherboard configured with 4x4GB LPDDR5-6400 (19-15-17), Windows® 11 22000.282, Samsung 980 Pro 1TB SSD, 28W nominal processor TDP, GPU driver 30.0, BIOS TRM0081D. RMB-13

that It was only at 15W, right?
OK, so where is your proof? I guess It was just too much of a bother searching for It.

Unlike you, I found a review of HP ProBook 635 Aero G8 with R7 5800U and yes, the score in CB R23 is 9576 points, which would mean It was set to 25W.
Now the question is If they have a typo in their footnotes or they manually set TDP to 15W.