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

[Vattila]

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!

 

[cellarnoise]

This CPU is so good you can run it with an insulator between the die and the cooler and you still kick ass with it 😂

These CPUs like to run hot. Don't want to risk overheating and damaging the cooler 😜

 

[Markfw]

These CPUs like to run hot. Don't want to risk overheating and damaging the cooler 😜

He is referring to my "boo-boo". The chip slows down until it can maintain the set temp, 95c by default. In my case no damage, thank god. And now its running a lot faster.

 

[eek2121]

They exist. They're expensive and require loud compressors or multi-stage compressor setups to function, and the peak heat flux from the heat source (read: CPU) can't be that high either. The more power these CPUs shed, the less-attractive phase gets.

You know, I have this little compressor-less desktop fridge that is awesome at cooling down beer and soda and stuff while using very little electricity (drinks are ice cold after an hour or so). I wonder why nobody has come up with some type of solution that does this for water in an AIO. I imagine it could shave 5-10 degrees off, depending on how much water is circulating per hour.

Proper chillers use too much juice, but it seems like something designed to shave 5-15 degrees off would be useful.

 

[igor_kavinski]

I wonder why nobody has come up with some type of solution that does this for water in an AIO.

One word: condensation.

I read that the Intel approved peltier solution for their Comet Lake i9 (and maybe Rocket Lake too) used heating in conjunction with cooling to avoid the formation of droplets that could damage hardware.

I think the CPU would need to be put inside a vacuum chamber to prevent condensation.

 

[Kocicak]

You know, I have this little compressor-less desktop fridge .... I wonder why nobody has come up with some type of solution that does this for water in an AIO.

Fridges without compressors use peltier thermoelectric coolers, which are extremely energy inefficient compared to compressor fridges.

Intel introduced recently AIO with peltier cooler. There is no problem with condensation, when you regulate the thing carefuly not to get the cold side below dew point.

The only problem is, that if your CPU consumes 200W, peltier cooler will ask for MORE than 200W on top of that. No wonder why nobody uses these.

I do not think that Intel sold many of these coolers.

EDIT: it seems that there IS a problem with condensation, because the cooler I mentioned cools below the dew point, it is also sold as Coolermaster ML360 sub-zero.

 

[maddie]

Fridges without compressors use peltier thermoelectric coolers, which are extremely energy inefficient compared to compressor fridges.

Intel introduced recently AIO with peltier cooler. There is no problem with condensation, when you regulate the thing carefuly not to get the cold side below dew point.

The only problem is, that if your CPU consumes 200W, peltier cooler will ask for MORE than 200W on top of that. No wonder why nobody uses these.

I do not think that Intel sold many of these coolers.

EDIT: it seems that there IS a problem with condensation, because the cooler I mentioned cools below the dew point, it is also sold as Coolermaster ML360 sub-zero.

I don't think its a peltier, or at least only a peltier.. They can't cool that fast, at least in small sizes. It probably uses a phase change storage with maybe a peltier to charge it over extended time periods.

 

[DrMrLordX]

You know, I have this little compressor-less desktop fridge

Fridge PC never works the way people think they will. Refrigerator compressors aren't designed for 24/7 operation, and those little peltiers in mini-fridges are at risk of thermal runaway if you force them to deal with any kind of heat flux.

 

[maddie]

You know, I have this little compressor-less desktop fridge that is awesome at cooling down beer and soda and stuff while using very little electricity (drinks are ice cold after an hour or so). I wonder why nobody has come up with some type of solution that does this for water in an AIO. I imagine it could shave 5-10 degrees off, depending on how much water is circulating per hour.

Proper chillers use too much juice, but it seems like something designed to shave 5-15 degrees off would be useful.

I'm curious, is it German made?

 

[Asterox]

2CU RDNA2 iGPU/2200mhz, 1080p 30-40FPS it is ok or nothing to complain about.

Vega 3 iGPU/1100mhz, 720p

 

[cellarnoise]

Amazing how far graphics have come the past 1,3,5,10 years!

Though both those pre-recorded Boobtube videos look crappy when stretched to 4k on my screen

 

[Timmah!]

FIRST I want to thanks everyone here for saying "something is wrong" After a zillion builds, I did the worst thing ever. I forgot to take the tape off the head for the AIO.

So... NOW a lot changed. Still 95c. BUT, NOW I am running 5.1 ghz, and vcore went up from .9 to 1.2. And watts went from 55 to 150.

Now what remains to be seen is what happens to performance. It was already kicking butt at 3100. So now at 5100, how good is it.

Its 11 pm at night, but I could not sleep thanks to your advice, so I fixed it. I am going to bed, and will report in the morning.

Getting too old (will be 69 soon)

Edit: The one thing I will say, is that I learned something. At 55 watts, its kicks butt compared to every CPU out there, including Alderlake. (my 12700F) At 150 watt ? Well, I will let you know in the morning, but lets just say I know its the efficiency king at least.

So its 95C with Arctic Freezer 420? I was planning to get that one as well, hoping it will do somewhat better, but i guess no is there any other AIO on the market, that would do better? This one seemed to be the highest rated, from what i could gather.

What thermalpaste are you using? Conductive or non-conductive one? I guess i would be scared to use conductive one cause of all the stuff thats on top of the cpu now on the side of the IHS…

 

[igor_kavinski]

So its 95C with Arctic Freezer 420? I was planning to get that one as well, hoping it will do somewhat better, but i guess no

You missed an important fact about Zen 4. No matter what solution you use, it will boost as high as possible and try to stay at or below 95 degrees. Less than that, you would need a Coolermaster Subzero AIO maybe.

 

[maddie]

We might be stuck in time regarding temps. Assuming total ignorance, why is 95 C so terrible? Is it that our past experience is preventing us from seeing a new model? 95 C in a vacuum is meaningless. Give me context, why bad?

 

[igor_kavinski]

Give me context, why bad?

Only thing I can think of is someone not liking a source of such heat warming up the space around their PC.

 

[maddie]

Only thing I can think of is someone not liking a source of such heat warming up the space around their PC.

Maybe that's the issue. We conflate heat and temperature.

You can have an extremely hot object emitting small amounts of heat, as in joules.
You can also have a slightly hot object emitting huge amounts of heat.
And all possible combinations in between.

Temperature, by itself, cannot indicate heat output.

 

[PJVol]

Give me context, why bad?

Offhand, at least two things comes to mind:
1. Unnecessarily high static power which is ~ leakage current * temperature, thus leaving less room for the dynamic part.
2. If reliability tracker (FIT) is on, then the higher temps, the lower maximum voltage.
Both result in decreasing of operating frequency and hence performance.

 

[Asterox]

Amazing how far graphics have come the past 1,3,5,10 years!

Though both those pre-recorded Boobtube videos look crappy when stretched to 4k on my screen

Yes, but 2CU is very small iGPU=if you get 1080p/30-40fps in any game you should jump for joy.

GPU frequency is high 2200mhz, but i wonder why AMD didn't put 4CU iGPU it wouldn't take too much space.

 

[Kocicak]

i wonder why AMD didn't put 4CU iGPU it wouldn't take too much space.

They will sell CPUs with "proper" integrated GPU later, perhaps they want to support their sale.

 

[nicalandia]

The highest Performance of a single socket CPU so far posted by Genoa(not even the top model)

ASUSTeK COMPUTER INC. RS500A-E12-RS12U - Geekbench

Benchmark results for an ASUSTeK COMPUTER INC. RS500A-E12-RS12U with an AMD EPYC 9654P processor.

browser.geekbench.com

For comparison this is what the 5995WX gets

 

[Abwx]

We might be stuck in time regarding temps. Assuming total ignorance, why is 95 C so terrible? Is it that our past experience is preventing us from seeing a new model? 95 C in a vacuum is meaningless. Give me context, why bad?

Because at 95°C an AMD CPU is hotter than an Intel CPU at 102°C...

That s a typical viral marketing campaign based on total clulessness of the competing CPUs, it is assumed that Intel s CPUs are running cooler even if it s documented that they have higher temps..

AMD Ryzen 7000 im Test: So schnell sind 7950X und 7700X: Temperaturen und Kühlung

Ryzen 9 7950X & Ryzen 7 7700X im Test: Temperaturen und Kühlung / TJMax und Temperaturen unter Last / TJMax = 95 °C gilt auch für Zen 4

www.computerbase.de

So the 12900K run at higher temps using the same cooler, but seems that it was not a problem, more likely that ignorance is the argument here...

 

[maddie]

Offhand, two things comes to mind:
1. Unnesessarily high static power which is ~ leakage current * temperature, thus leaving less for the dynamic part.
2. If reliability tracker is on, then the higher temps, the lower maximum voltage, so is the operating frequency.

Concerning (1).

This has always been true. So, if not 95, why 85, or 75 or 65? Where do you stop?
Chemistry certainly affects leakage. Past gens 95 C leakage does not have to equal latest 95 C leakage.

All of this compresses to my saying that we're looking at 95 C in a vacuum as some absolute value versus one that can change with new materials and technologies.

I have problems with number values containing no context.

Concerning (2)

How do you know this is worse than before? Node chemistry matters. What applied once does not necessarily apply always and ALL design values have compromises. They existed before and will in the future.

 

[moinmoin]

I'd say people who find 95°C to be too high should set their desired temperature limit in the BIOS and be done with that topic.

I'd still be interested in hearing at what low point such a temperature limit starts to seriously limit performance, but the high-ish numbers so far indicate one can go quite low until then.

 

[PJVol]

I have problems with number values containing no context

Well... as of
(1) I was talking about high temperatures in general, not a specific temperature limit value, which I assume is the result of the same trade-offs as when choosing TDP/PPT for a particular SKU.
(2) sorry, I don't quite understand what you mean by "worse than before", but assuming I may have misunderstood the point of your question, which apparently was about a specific value of 95°C, this also applies to high temperatures in general.

 

[igor_kavinski]

is there any other AIO on the market, that would do better?

From computerbase:



I don't know what the ambient temperature was in their test, though.

 

[nicalandia]

96 Core Genoa vs 96 Core Sapphire Rapids.

ASUSTeK COMPUTER INC. RS500A-E12-RS12U vs Nettrix N/A - Geekbench

browser.geekbench.com