• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Speculation: Ryzen 3000 series

Page 207 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

What will Ryzen 3000 for AM4 look like?

  • 16-core monolithic CPU (four 4-core CCXs on SoC die)

    Votes: 13 5.7%

  • 12-core monolithic CPU (three 4-core CCXs on SoC die)

    Votes: 19 8.3%

  • 8-core monolithic CPU (two 4-core CCXs on SoC die)

    Votes: 24 10.4%

  • 8-core monolithic APU (two 4-core CCXs and an iGPU on SoC die)

    Votes: 8 3.5%

  • 4-core monolithic APU (one 4-core CCX and an iGPU on SoC die)

    Votes: 12 5.2%

  • 16-core modular CPU (two 8-core CPU chiplets + IO chiplet)

    Votes: 146 63.5%

  • 8-core modular CPU (one 8-core CPU chiplet + IO chiplet)

    Votes: 47 20.4%

  • 8-core modular APU (one 8-core CPU chiplet + IO chiplet + GPU chiplet)

    Votes: 23 10.0%

  • 8-core modular APU (one 8-core CPU chiplet + IO chiplet with iGPU)

    Votes: 15 6.5%

  • Something else (specify below)

    Votes: 11 4.8%
  • Total voters
    230
PotatoWithEarsOnSide said:
Boost implies not sustained. It isn't a difficult concept to understand.
If a clock is sustained at max boost then it is operating out of spec, as is the case with Intel CPUs.
Click to expand...

You are confusing all core and single core boosts. Single core boost can't use much power and thus boost is sustainable forever within power and thermal limits.
 
naukkis said:
Single core boost can't use much power and thus boost is sustainable forever within power and thermal limits.
Click to expand...

Well, that would imply that a core can run at max frequency at max temperature within nominal specifications, i.e. at thermal limit. Cores can achieve higher frequency at lower temperature (about 100 MHz for every 20 degrees C, according to DerBauer).

So, if sufficiently below max nominal temperature, a core can briefly boost to max frequency until the core heats up and forms a hot spot, which it probably does in a fraction of a second. At this point the OS may move the workload to an idle (and cold) core, allowing the hotspot to cool. Or the workload may ease up briefly (e.g. stalling on memory access or I/O), after which the core may have cooled down sufficiently to briefly boost higher again. This is seen as frequency spikes.

It appears AMD compensated for these spikes by lowering the advertised Max Boost frequencies by 50 to 100 MHz for the 1st and 2nd generation Ryzen and introducing the XFR concept accounting for the difference between the Max Boost advertised and the actual max frequency achievable by the CPU at stock settings. The XFR concept seems to have been dropped altogether for the 3rd generation, and the Max Boost now means max (or much closer to it, at least).

To remind everyone about XFR, I'll include this slide again:

AMD%20Ryzen%207%20Press%20Deck-10.jpg
 
Last edited:
Interesting. I wonder whether the 3500 and 3500X are coming to the USA, and how much? You know one of them is going to have to match the price of the i5-9400F, Intel's got a competitive chip @ $140 at that price point, though a 6C/6T Ryzen would be preferable to me.
 
VirtualLarry said:
Interesting. I wonder whether the 3500 and 3500X are coming to the USA, and how much? You know one of them is going to have to match the price of the i5-9400F, Intel's got a competitive chip @ $140 at that price point, though a 6C/6T Ryzen would be preferable to me.
Click to expand...

From what I know 3500X is priced pretty similar as 9400F. I guess if cache config is true 3500 would be cheaper than expected.


OTOH what surprise me is AMD rather release 6C6T SKUs than 4C8T.
 
cortexa99 said:
OTOH what surprise me is AMD rather release 6C6T SKUs than 4C8T.
Click to expand...
I think that the "marketing optics" are better, with AMD competing with Intel, with Intel now selling 6C/6T CPUs minus iGPU, rather than 4C/8T, that AMD should match them core-for-core, for the same price, and come out even or ahead on performance, on an arguably superior, upgradable platform.
 
2+2 and 4+0 die salvage is going to be pretty rare, and it seems that Epyc is taking a some of these dies too with their high L3:cores parts. With 80%+ yields on 8c chiplets it would take a while to accumulate.

Also, given this 3500 series launch, if AM4 gets any 4c/8t chiplet based parts they would have to get numbered with the 3400 series. like 3430x or so.

In my opinion this is a good move, to get rid of some of the dies that don't make the freq cutoff for the 3600x and 3700. It allows you to price the 6c/6t parts a good bit under the 3600 price and helps the 3600 out in moving volume of low bins.
 
Last edited:
Besides, AMD already has 4c4t and 4c8t parts with their APUs. They can also salvage 2c4t and 2c2t if they want. They could potentially salvage ones with bad gpus.

I wonder if the 2300/2500x chips are still being made?
 
They are sold in retail but at hugely inlated prices, something like 150/195€ respectively for the 2300X/2500X, so basically yes, it s OEM only...

geizhals.de

AMD Ryzen 3 2300X | Preisvergleich Geizhals Deutschland

✔ Preisvergleich für AMD Ryzen 3 2300X ✔ Produktinfo ⇒ Kerne: 4 (4C) • Threads: 4 • Turbotakt: 4.00GHz • Basistakt: 3.50GHz… ✔ AMD ✔ Testberichte ✔ Günstig kaufen
geizhals.de geizhals.de

geizhals.de

AMD Ryzen 5 2500X | Preisvergleich Geizhals Deutschland

✔ Preisvergleich für AMD Ryzen 5 2500X ✔ Produktinfo ⇒ Kerne: 4 (4C) • Threads: 8 • Turbotakt: 4.00GHz • Basistakt: 3.60GHz… ✔ AMD ✔ Testberichte ✔ Günstig kaufen
geizhals.de geizhals.de
 
Looks like 24c/48t might be Threatripper 3000 flagship. If there is a higher core count Zen2 part for TR4 I wouldn't expect it till next year and if so, possibly branded under Opteron. Chiplets seem under high demand now. It could be a combo of healthy server and AM4 demand. I think it would be a possibility if parts were still added to the 3000 line in 2020, including Pinnacle ridge; during Q2, close to when the first 4000 series start arriving.
 
amd6502 said:
Looks like 24c/48t might be Threatripper 3000 flagship. If there is a higher core count Zen2 part for TR4 I wouldn't expect it till next year and if so, possibly branded under Opteron. Chiplets seem under high demand now. It could be a combo of healthy server and AM4 demand. I think it would be a possibility if parts were still added to the 3000 line in 2020, including Pinnacle ridge; during Q2, close to when the first 4000 series start arriving.
Click to expand...
I can't believe it will premere with less than the 2990wx. I would say 48core tops, 24 core min. Then 64 core later.
 
Markfw said:
I can't believe it will premere with less than the 2990wx. I would say 48core tops, 24 core min. Then 64 core later.
Click to expand...

24 cores fully loaded at mid 4's is going to draw like 250+ W as it is. You may as well make an "Epyc W" platform (and charge Epyc prices) if you want to sell 300 W+ monsters.
 
jpiniero said:
24 cores fully loaded at mid 4's is going to draw like 250+ W as it is. You may as well make an "Epyc W" platform (and charge Epyc prices) if you want to sell 300 W+ monsters.
Click to expand...
Dont bet on that. My 2970wx does not draw 250 watt and its the old version.
 
jpiniero said:
24 cores fully loaded at mid 4's is going to draw like 250+ W as it is. You may as well make an "Epyc W" platform (and charge Epyc prices) if you want to sell 300 W+ monsters.
Click to expand...
Now sure now. My 3900x (one of them) is doing 134 watts @4.1 ghz, so 268 for 4.1... Maybe
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top