Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)

Page 388 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Vattila

Senior member
Oct 22, 2004
805
1,394
136
Except for the details about the improvements in the microarchitecture, we now know pretty well what to expect with Zen 3.

The leaked presentation by AMD Senior Manager Martin Hilgeman shows that EPYC 3 "Milan" will, as promised and expected, reuse the current platform (SP3), and the system architecture and packaging looks to be the same, with the same 9-die chiplet design and the same maximum core and thread-count (no SMT-4, contrary to rumour). The biggest change revealed so far is the enlargement of the compute complex from 4 cores to 8 cores, all sharing a larger L3 cache ("32+ MB", likely to double to 64 MB, I think).

Hilgeman's slides did also show that EPYC 4 "Genoa" is in the definition phase (or was at the time of the presentation in September, at least), and will come with a new platform (SP5), with new memory support (likely DDR5).



What else do you think we will see with Zen 4? PCI-Express 5 support? Increased core-count? 4-way SMT? New packaging (interposer, 2.5D, 3D)? Integrated memory on package (HBM)?

Vote in the poll and share your thoughts!
 
Last edited:
Reactions: richardllewis_01

Hans Gruber

Platinum Member
Dec 23, 2006
2,219
1,153
136
I think AMD needs an Extreme Power, Power and Eco mode for Zen 4 in bios. Extreme Power would be the new TDP for Zen 4. Power would be the same TDP as Zen 3 and eco would be 65w or less. That way people would not complain about high temps. I think AMD bumped the TDP because they were not getting the IPC in Zen 4 that they wanted. They knew Raptor Lake was coming and the increased TDP was their solution.

It seems obvious that 3D-vache should be standard on Zen 4.
 

eek2121

Diamond Member
Aug 2, 2005
3,051
4,273
136
I don't care that much for weight, but I certainly don't want a DTR like Schenker or XMG. That's a total overkill.
You can get a nice Razer Blade 14 weighting only 1.8kg (4 pounds) with Ryzen 9 6900HX + RTX 3070 Ti Laptop with decent temperatures. I expect they will update It to Dragon Ridge next year.
For Phoenix APU It would be best to choose at least a 45W TDP version, to not sacrifice too much CPU and IGP performance, but then weight won't be <1kg but ~1.4kg.
It will depend on the asked price, but Dragon Ridge would be more future-proof.


Note that in the case of the Razer laptop, I have the previous gen and the power brick is as heavy as the laptop is.

I think AMD needs an Extreme Power, Power and Eco mode for Zen 4 in bios. Extreme Power would be the new TDP for Zen 4. Power would be the same TDP as Zen 3 and eco would be 65w or less. That way people would not complain about high temps. I think AMD bumped the TDP because they were not getting the IPC in Zen 4 that they wanted. They knew Raptor Lake was coming and the increased TDP was their solution.

It seems obvious that 3D-vache should be standard on Zen 4.

I agree on your first paragraph except it should be a setting in Ryzen master instead. Strong disagree with the second comment. Why would AMD raise the cost of all their CPUs for something that really only helps gaming?
 
Reactions: Kaluan

Hans Gruber

Platinum Member
Dec 23, 2006
2,219
1,153
136
Note that in the case of the Razer laptop, I have the previous gen and the power brick is as heavy as the laptop is.



I agree on your first paragraph except it should be a setting in Ryzen master instead. Strong disagree with the second comment. Why would AMD raise the cost of all their CPUs for something that really only helps gaming?
Software based is no good. It needs to be in the bios. You start at the baseline which is the firmware aka. bios.

My 3D-Cache comment is based on Intel. They don't like to lose and they do not lose often or for long. Gaming performance is the most important area in consumer CPU's. The 7950x is basically a threadripper. That is the HEDT segment. Even those guys like to have the best in everything including gaming performance. 3D v-cache is proven technology.

I personally do not care much about temps. I do care about power consumption. One more thing to remember. Zen 4 is brand new and it will take AMD 2 or 3 months to iron out the bios. In the past, AMD has always listened to their customers. Whereas Intel says, who cares or we will address it in the next CPU iteration and/or motherboard.
 

HurleyBird

Platinum Member
Apr 22, 2003
2,726
1,342
136
I don't wanna risk ruining a CPU through delidding.

Perhaps lapping will also yield a major improvement. I'm certainly not going to delid and also use a special bracket. Too much hassle and risk. But if I can improve temps by >=10° C from lapping alone, I probably will (if I get a Ryzen 7000).
 

Panino Manino

Senior member
Jan 28, 2017
846
1,061
136
What? Nearly 30% ST performance Boost and Nearly 50% MT performance boost and you call this "Least Exciting"?

This is single handedly the most performance AMD has been able to extract since Bulldozer and you are unimpressed? Get Out Of Here...!

I'm a few pages back yet but I have to say, I'm also a bit disappointed.
Yes, compared with Zen 3 is good, but the hype here on Anandtech made me expect more in comparison to Alder Lake, specially because Zen 4 amazingly have a clock advantage over Alder Lake.
I was seeing the results and feeling a bit disappointed.
It got better when I remembered that Alder Lake have bunch of extra cores helping, but I'm still a bit unsure.

Looks like Rapter Lake will really be ahead of Zen 4, at least in performance ("at what coast?!" Yes, despite that).
 
Last edited:

Hans Gruber

Platinum Member
Dec 23, 2006
2,219
1,153
136
Perhaps lapping will also yield a major improvement. I'm certainly not going to delid and also use a special bracket. Too much hassle and risk. But if I can improve temps by >=10° C from lapping alone, I probably will (if I get a Ryzen 7000).
I do not think lapping or using liquid metal (maybe 2 or 3C) would do much good. Zen 4 is based on 5nm vs. 7nm for Zen 2 and 3. Hot spots on CPU's became an issue with 7nm. With 5nm, hot spotting on the CPU is going to be a much more significant issue. They need to switch from silicon to something that conducts heat better.
 

moinmoin

Diamond Member
Jun 1, 2017
4,994
7,765
136
Not sure if it was noted somewhere else, but Computerbase quoted AMD to expect a bigger v-cache than with 5800X3D:

"AMD hat angedeutet, dass der zusätzliche L3-Cache in den für Anfang 2023 erwarteten Ryzen 7000X3D größer ausfallen könnte als aktuell beim Ryzen 7 5800X3D (64 MB)."

"AMD has hinted that the additional L3 cache in the Ryzen 7000X3D, expected in early 2023, could be larger than currently in the Ryzen 7 5800X3D (64 MB)."

Woaah mate, same time.
You were one minute late according to the timestamp.
 

inf64

Diamond Member
Mar 11, 2011
3,764
4,222
136
Looks like Rapter Lake will really be ahead of Zen 4, at least in performance ("at what coast?!" Yes, despite that).

I think you will be (again) disappointed. Raptor Lake is basically E core spam with slight clock uplift vs 12900KS. It will not match the MT performance of 7950X (stock vs stock) and it might edge out ST and gaming by a hair. At what cost? At the cost of consuming 100W more and being on a dead end platform.

Raptor Lake will shine in the mid-range (7700X/7600X) and AMD might have a problem there.

btw, we missed the most important question: Can it run Crysis

Video by Dr. Cutress


7950X is ~25% faster than 12900K in Crysis. Almost there at magical 25fps barrier!

 

Hitman928

Diamond Member
Apr 15, 2012
5,600
8,790
136
Maybe a special water block with the CPU die integrated into it. The CPU cannot be taken out. You take the waterblock and attach it to the socket and then connect the tubes to it. Only for those who demand the best.

The amount of NRE that would go into such a solution would drive the cost way higher than even the most extreme would go for. There's no way you could make this remotely commercially viable. I would also have huge concerns of shipping a bare die CPU attached to a waterblock. As the package is tossed around shipped around, there's a significant chance that a very large percentage of chips arrive DOA. You would literally be better of just offering special order CPUs without an IHS but that's not going to happen either for reasons already mentioned.
 

Hitman928

Diamond Member
Apr 15, 2012
5,600
8,790
136
They could also put the connections on four sides of the die and let the heat be exchanged from both sides of the die (top and bottom).

Physics do not agree that this would help in any meaningful way. No significant heat would travel out the sides of the die. Heat will already conduct out through the socket but there isn't a great thermal path there so the vast, vast majority of heat goes out the side of the IHS.
 
Reactions: Kaluan and ftt

itsmydamnation

Platinum Member
Feb 6, 2011
2,863
3,417
136
In Linux at least, turning off AVX-512 actually increases thermals!

View attachment 68133

I hope someone comes up with a really good explanation for that.

My own guess is that without AVX-512, the various execution units of the CPU get hammered hard, forcing the CPU to use max power in an effort to deliver the best performance with max boost. With AVX-512, the CPU doesn't need to boost way past its efficiency curve and manages to deliver the needed performance at lower boost clocks.
no its really simple 1/2 the work needed for the front end for the same throughput. Intel with their AVX-512 design struggle with spike power which needs down clocking to handle. AMD just seem to have design a well rounded Beast
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,101
136
At the cost of consuming 100W more and being on a dead end platform.
Raptor Lake and Raphael have effectively the same power limit. I don't understand the infatuation with this 350W "rumor". And you think that applies for gaming of all things? Come on, we have numbers now. Let's use them.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |