Speculation: Ryzen 3000 series

[Fanatical Meat]

Well, I have to say this LOOKS legit. So at some point a 3800x will be in my house. 4.7 turbo on 16 cores ? Yea baby ! I bet I could get 5 out of it on a good 240 or 360 AIO.

Edit: while I was posting, FlanK3r was also. Well, how could they be selling fake CPU's ?

.......but can it run Crysis?


sorry couldn't resist.

 

[moinmoin]

View attachment 5806



De Ryzen 9 3800X box from a different angle on this site:

https://nguyencongpc.vn/cpu-amd-ryzen-9-3800x-3-8-ghz/

View attachment 5805

The leakers/fakers can't agree on a consistent font family for that "9" on it yet.

 

[Hans de Vries]

The leakers/fakers can't agree on a consistent font family for that "9" on it yet.

Neither one of the "9" is consistent with the "RYZEN" font, while the "3", "5" and "7" on other boxes are. The "RYZEN" font may well be a custom font.

 

[Veradun]

Neither one of the "9" is consistent with the "RYZEN" font, while the "3", "5" and "7" on other boxes are. The "RYZEN" font may well be a custom font.

I'm still intrigued at how Ryzen and Radeon haven't the same font.

 

[Atari2600]

As far i see nguyencongpc.vn Ryzen 9 photo, this is very likely low resolution photo of real pasteboard Ryzen 9 retail box package.

https://nguyencongpc.vn/cpu-amd-ryzen-9-3850x-4-3-ghz/

Pretty sure that 9 is not aligned with the box - looks like its normal to the viewers POV to me - or at best slightly isometric, rather than correctly represented in a perspective view.

 

[PingSpike]

Any idea when the x570 motherboards are coming out? Last time I heard it was June but I'm not sure anymore.

 

[Fanatical Meat]

Any idea when the x570 motherboards are coming out? Last time I heard it was June but I'm not sure anymore.

AMD is having a partners meeting at the end of the month. We will likely know more after that.

Speculative time frame for release here:

https://www.pcgamesn.com/amd/ryzen-3000-release-date-ryzen-3rd-gen-specs-performance-price

 

[DrMrLordX]

More 7/7 release date speculation. AMD really doesn't need to wait that long . . .

 

[Seronei]

7/7 is a sunday, extremely unlikely to release on that day.

 

[Fanatical Meat]

7/7 is a sunday, extremely unlikely to release on that day.

Smartest thing I have seen regarding the supposed 7/7 release for 7nm
You are a smart man

 

[Vattila]

7/7 is a sunday, extremely unlikely to release on that day.

Are you sure? The numerology is strong on this one. Sunday is the 7th day of the week (on some calendars, at least), and 2019-07-07 falls at the end of the 27th week. Shame, I hear you say, that the year 2019 has no sevens at all — but then I point you to the fact that 2019 is equal to 7*7*7*7 - 7*7*7 - 7*7 + 7/.7. Booyakasha!

 

[Hitman928]

Are you sure? The numerology is strong on this one. Sunday is the 7th day of the week (on some calendars, at least), and 2019-07-07 falls at the end of the 27th week. Shame, I hear you say, that the year 2019 has no sevens at all — but then I point you to the fact that 2019 is equal to 7*7*7*7 - 7*7*7 - 7*7 + 7/.7. Booyakasha!

#Confirmed

 

[Ajay]

Makes sense. If we are lucky a June release perhaps?

That's the rumor. The WCCF rumor is misrepresenting the earnings call. Su was vague on when 7nm Ryzen comes to market.
AMD needs to announce something at Computex, hence the rumor that it'll be Ryzen with mass market availability some time later.

 

[krumme]

That's the rumor. The WCCF rumor is misrepresenting the earnings call. Su was vague on when 7nm Ryzen comes to market.
AMD needs to announce something at Computex, hence the rumor that it'll be Ryzen with mass market availability some time later.

How long before first shipping do you typically know fmax?

 

[RTX2080]

As for the doubtful R9-3800x box photo, I posted two box photoes to my local forum but got deleted 1 day later. I think if these photoes are fake, why the administrators so serious & nervous about them, why delete my post? hmmmmm

 

[exquisitechar]

Some guy messaged one Turkish store that lists the Ryzen 3k SKUs, they said that they got the information from their supplier and that they anticipate a launch in August.

Honestly, I think it's almost certainly fake. These stores have lied before about these kind of things, I guess it gives them a lot of attention. The fact that they perfectly match Adored's leak makes you wonder, even if the leak is legit, wouldn't an SKU or two be changed compared to what was planned 6 months ago? They probably just copied it. The L3 caches being gimped, I also find unlikely, but it's not impossible, I suppose.

 

[Asterox]

Some guy messaged one Turkish store that lists the Ryzen 3k SKUs, they said that they got the information from their supplier and that they anticipate a launch in August.

Honestly, I think it's almost certainly fake. These stores have lied before about these kind of things, I guess it gives them a lot of attention. The fact that they perfectly match Adored's leak makes you wonder, even if the leak is legit, wouldn't an SKU or two be changed compared to what was planned 6 months ago? They probably just copied it. The L3 caches being gimped, I also find unlikely, but it's not impossible, I suppose.

What is unlikely or impossible, well look at this from a diferent view.

Threadripper 2950X max boost singlecore is 4.4ghz, all this on Samsung 14nm.We all now what is Samsung 14nm, or not high performance node+not designed for significantly higher cpu frequencies.

https://www.amd.com/en/products/cpu/amd-ryzen-threadripper-2950x

TSMC 7nm, someone said that Zen 2 engineering CPU sample can hit 4.5ghz. That is all and nothing higher, let say only 4.5ghz for singlecore max boost well that is very unlikely.

 

[Abwx]

Dunno for the effective max frequencies but for base frequencies the alleged leaks are perfectly in line with the sample AMD displayed at CES.

The sample was running at 4GHz or so, with the chiplet taking 45-50W and the I/O 10-15W for a 60W grand total, doubbling the chiplet would require 90-100W and say 20W for the I/O, downclocking the thing to 3.9GHz would allow to be within 125W with Prime 95.

 

[PotatoWithEarsOnSide]

The IO is taking up much more than that, unless you're assigning IF power consumption to the chiplets which IMO is the incorrect approach.
We saw from last gen Ryzen that the cores were consuming barely any power, and since they're the part on 7nm then they'll be consuming even less this time around.
It's this complete lack of power consumption per core that makes the chiplet design so scalable upwards, but less viable at the bottom end.
IMO, each chiplet is consuming 30w or about 3.5w per core, with the IO being around 30-40w.

 

[amd6502]

Dunno for the effective max frequencies but for base frequencies the alleged leaks are perfectly in line with the sample AMD displayed at CES.

The sample was running at 4GHz or so, with the chiplet taking 45-50W and the I/O 10-15W for a 60W grand total, doubbling the chiplet would require 90-100W and say 20W for the I/O, downclocking the thing to 3.9GHz would allow to be within 125W with Prime 95.

Good estimate. The final silicon could be improved and refined. So I see base as 4ghz and to even a few hundred MHz above, especially for single chiplet products. I also don't see multi-chiplet products available in the initial launch. It's going to be a paper launch for Ryzen 9 while Ryzen 7 and hexacores will have a month to three head start.

 

[coercitiv]

The IO is taking up much more than that, unless you're assigning IF power consumption to the chiplets which IMO is the incorrect approach.
We saw from last gen Ryzen that the cores were consuming barely any power, and since they're the part on 7nm then they'll be consuming even less this time around.
It's this complete lack of power consumption per core that makes the chiplet design so scalable upwards, but less viable at the bottom end.
IMO, each chiplet is consuming 30w or about 3.5w per core, with the IO being around 30-40w.

From Anandtech's review of the 2700X:
Package Power - 106W
Cores only - 85W

Power delta between package and cores --> 21W.

 

[Guru]

I fully expect AMD's top CPU's to reach 5GHz. I don't see why people are so suspicious about this? Ryzen 2700X can boost to 4.3Ghz on 14nm GloFo process, 2950X at 12nm can boost to 4.4GHz, so I don't see why new and improved CPU's on 7nm can't reach 5GHz, its seems very likely to me. If anything we might even see a CPU that comes later on, something like a Ryzen 9 3900x that can boost up to 5.2GHz.

 

[realibrad]

I fully expect AMD's top CPU's to reach 5GHz. I don't see why people are so suspicious about this? Ryzen 2700X can boost to 4.3Ghz on 14nm GloFo process, 2950X at 12nm can boost to 4.4GHz, so I don't see why new and improved CPU's on 7nm can't reach 5GHz, its seems very likely to me. If anything we might even see a CPU that comes later on, something like a Ryzen 9 3900x that can boost up to 5.2GHz.

Here is a good breakdown of why people expect lower than 5ghz.

https://www.reddit.com/r/Amd/comments/a8b5aa

 

[Guru]

Here is a good breakdown of why people expect lower than 5ghz.

https://www.reddit.com/r/Amd/comments/a8b5aa

Everything you posted is pure speculation, though we do have actual factual evidence and that is Radeon 7 and MI60. Its essentially tweaked Vega on 7nm. What it does on 7nm is being able to provide 35% more performance at the same power consumption, with less CU's compared to Vega 64.

Radeon 7 is essentially a cut down MI60 running at 50MHz less, but MI60 runs at 1800MHz or 264MHz more than Vega 64. Again this is the same architecture with minor tweaks, no major redesign, nothing of the sort. Its a slightly tweaked Vega 64 on 7nm process essentially.

And again why is Nvidia able to run GPU's at 2100MHz, why is Intel able to run CPU's at 5GHz and AMD can't? Why can't they have improved their technology, improved their tools, with TSMC improved the 7nm process from the MI60 days, and we had 7nm sampled in Q2 2018, so its been a full year for the process to mature, it's going to be a year and a quarter when Ryzen and Navi hit the market.

TSMC recently announced their 6nm process node and they claim it's up to 10% faster than 7nm or 15% more efficient. So to me it's highly likely that AMD are using an advanced 7nm process, 7+nm if you will for their next Ryzen and Navi chips.

Again I don't see how it's unbelievable that they've been able to increase clock by only 600MHz on 7nm as opposed to 12nm, even if it means increasing the power consumption somewhat. It doesn't really matter if AMD's final power consumption is say 20W more than the leaked documents, it won't matter, as Intel's 8core chips consume much more power already, if AMD releases 16 cores at 150W that would still devastate Intel and blow them out of the water. Their perf per power per chip would be out of this world.

Again to me it seems very likely that on at least a one year mature 7nm process, on an upgraded Ryzen architecture they've been able to increase clocks by 600MHz.

 

[amd6502]

The advantage for GPU is different. The 5GHz barrier isn't relevant for GPUs in the foreseeable future. Localized heat is also less of a problem.

I expect high single digits or around ~10% gains for 7nm at 4GHz+ over 12nm, which is a performance oriented version of 14LPP+

Shintel is using a highly refined mature 12nm-like node with high speed libraries. They've also gotten more aggressive with elite binning. That's how they are able to reach high frequencies.

I am guessing we'll beat 4.5ghz boosts on good binnings. 5ghz would be a really pleasant surprise, and a big surprise. I do think those speeds will be surpassed in 2020 with a more enhanced and mature 7nm node.