• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."
  • Community Question: What makes a good motherboard?

Speculation: Ryzen 4000 series/Zen 3

Page 63 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

uzzi38

Golden Member
Oct 16, 2019
1,220
2,250
96
Wat.

Dude, _rogame just spends ages scouring for benchmarks etc. I called him my guy because we spend a lot of time discussing this stuff in DMs.

He's not the kind of guy to make this stuff up. But for the most part, just take this as confirmation that Renoir is up to 8 cores.
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
Wat.

Dude, _rogame just spends ages scouring for benchmarks etc. I called him my guy because we spend a lot of time discussing this stuff in DMs.

He's not the kind of guy to make this stuff up. But for the most part, just take this as confirmation that Renoir is up to 8 cores.
OK, got ya. When you said "my guy" it sounded like you had an inside source. Not saying this guy isn't credible though.
 

IntelUser2000

Elite Member
Oct 14, 2003
7,249
1,839
136
Interesting how all these years later, people are still trying to make that point. Because we got the ARM fans or the RISC-V fans or whoever trying to crap on x86. x86 is sill around. That's what I meant when I said "You know how long I've been hearing that?".
Know how this is true?

Because even for GPU ARM SoCs are beating AMD/Intel. The performance of the GPU is absolutely fantastic for their die size and power consumption. Qualcomm's SD855 is in the 10mm2 range and A13's GPU is 15mm2.

What happens is the execution of ARM-based vendors are stellar and AMD/Intel suck. In the desktop GPU world, Nvidia executes better than AMD/Intel.
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
Know how this is true?

Because even for GPU ARM SoCs are beating AMD/Intel. The performance of the GPU is absolutely fantastic for their die size and power consumption. Qualcomm's SD855 is in the 10mm2 range and A13's GPU is 15mm2.

What happens is the execution of ARM-based vendors are stellar and AMD/Intel suck. In the desktop GPU world, Nvidia executes better than AMD/Intel.
I'm not exactly a GPU guy but that sounds plausible. I don't think AMD's execution sucks though, at least not on the CPU side. NVIDIA has been on a roll, I'll give them that.
 

IntelUser2000

Elite Member
Oct 14, 2003
7,249
1,839
136
I'm not exactly a GPU guy but that sounds plausible. I don't think AMD's execution sucks though, at least not on the CPU side. NVIDIA has been on a roll, I'll give them that.
No they suck.

I did some simple comparisons. I estimate while Gen11/Vega iGPU is twice as fast as the SD855's GPU, Gen 11 is 4x the size, and Vega is 6.5x the size. Granted Vega will get a die shrink with 7nm, but still will be 3x the size.

Apple is of course at the top of their game. 15mm2 but with similar performance. I do think their vertical integration strategies may be saving them here but nevertheless their execution is simply stellar.

Unfortunately the x86 duoplies go through serious ups and downs and that have hampered them too much. Intel with Netburst, then this. AMD with Bulldozer. Each time they falter, they are at a standstill. What happened during that time is that ARM vendors steadily progressed. Now their A77 is at Zen 2 level, if not more.

What the ISA differences give is people excuses on why Intel/AMD suck so much on execution. I think its a little beyond that now.
 
  • Like
Reactions: Tlh97

Richie Rich

Senior member
Jul 28, 2019
470
227
76
We knew about chiplets long before Zen 2. That's a much larger shift then SMT4. We knew plenty about Netburst and K7 before they came out. How long did we know about 64 bit in K8 before it came out? That was huge. Extending an ISA vs adding a couple more threads/core. So to say AMD/Intel has never disclosed a major next gen feature that far in advance is flat out false. AMD announced x86-64 a full three years before it came out. K7 and Netburst were about a year each before release. I'm willing to hear your arguments, but make sure they are accurate if you want to change mine (or anyone's) mind.
You are mixing two very different things here: ISA extension and Processor uarch features. One is interface thing (HW and SW needs to adapt), the other is blackbox thing (no SW adaptation needed). You desperately try to defend your naive SMT4 feature revealing during Zen2 presentation. You give a hard time to yourself by trying to defend such a things.


OK, so your just inventing a metric (pts/GHz) using just SPEC2006 and are using that to claim ARM is far superior to x86. IPC on its own means nothing. GHz on its own means nothing. ARM excels in low power performance. x86 does well in overall performance. Can these lower power ARM cores scale up to 32 or 64 core CPU's without blowing up power usage? I don't know. I'd like to see them try.
Metric like SPECint2006/GHz is not an invention, it's just basic engineering to normalize values to be comparable. IPC on its own means nothing? Maybe for you. However for engineers an IPC is the most important metric about CPU architecture performance.

Problem no. 1 is that you don't want to see that x86 AMD Ryzen 3950X is loosing vs. Apple A13 in overall performance (50.02 vs 52.82) despite its much higher frequency 4.6GHz vs 2.65 GHz. Any x86 CPU is terribly slower than Apple's ARM in therms of IPC.
Problem no. 2 is that generic Cortex core A77 is also a bit faster than Zen 2. This is not a prediction of x86 will be in trouble in the future. This is actually measured values confirming the fact that x86 is in trouble already IMHO. Feel free to show us your own measurement to proving x86 superiority. I'd prefer less talking and more graphs and numbers please.


IPC calculations of SPECint2006:
  • - 9900K .... 54.28/5 GHz = 10.86 pts/GHz
  • - 3950X .... 50.02/4.6 GH = 10.87 pts/GHz
  • - A76 ........ 26.65/2.84 GHz = 9.38 pts/GHz
  • - A77 ........ 33.32/2.84 GHz = 11.73 pts/GHz ...... +8% IPC over 9900K
  • - A11 ........ 36.80/2.39 GHz = 15.40 pts/GHz .... +42% IPC over 9900K
  • - A12 ........ 45.32/2.53 GHz = 17.91 pts/GHz .... +65% IPC over 9900K
  • - A13 ........ 52.82/2.65 GHz = 19.93 pts/GHz .... +83% IPC over 9900K
  • - A14 ........ 66.00/3.00 GHz = 22.00 pts/GHz (estimated +10%IPC)
  • - Zen3 ...... 60.00/4.60 GHz = 13.04 pts/GHz (estimated +20% IPC)


For example Ampere eMag 80-core ARM server CPU with A76 will be hard competitor to AMD's 64c EPYC. Those 16-cores advantage is exactly +25% more. This compensates SMT2 feature (usually +15-25% performance boost) while having same performance with just 80 threads in compare to Epyc's 128 threads (SMT2). AMD needs Zen 3 to be something much better than just poor Zen 2.
 

soresu

Golden Member
Dec 19, 2014
1,532
757
136
I did some simple comparisons. I estimate while Gen11/Vega iGPU is twice as fast as the SD855's GPU, Gen 11 is 4x the size, and Vega is 6.5x the size. Granted Vega will get a die shrink with 7nm, but still will be 3x the size.

Apple is of course at the top of their game. 15mm2 but with similar performance. I do think their vertical integration strategies may be saving them here but nevertheless their execution is simply stellar.
I wonder how the new PowerVR A Series compares size wise?
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
No they suck.

I did some simple comparisons. I estimate while Gen11/Vega iGPU is twice as fast as the SD855's GPU, Gen 11 is 4x the size, and Vega is 6.5x the size. Granted Vega will get a die shrink with 7nm, but still will be 3x the size.

Apple is of course at the top of their game. 15mm2 but with similar performance. I do think their vertical integration strategies may be saving them here but nevertheless their execution is simply stellar.

Unfortunately the x86 duoplies go through serious ups and downs and that have hampered them too much. Intel with Netburst, then this. AMD with Bulldozer. Each time they falter, they are at a standstill. What happened during that time is that ARM vendors steadily progressed. Now their A77 is at Zen 2 level, if not more.

What the ISA differences give is people excuses on why Intel/AMD suck so much on execution. I think its a little beyond that now.
Like I said I'm not really all that GPU savvy. I would agree there is is poor execution there.


You are mixing two very different things here: ISA extension and Processor uarch features. One is interface thing (HW and SW needs to adapt), the other is blackbox thing (no SW adaptation needed). You desperately try to defend your naive SMT4 feature revealing during Zen2 presentation. You give a hard time to yourself by trying to defend such a things.



Metric like SPECint2006/GHz is not an invention, it's just basic engineering to normalize values to be comparable. IPC on its own means nothing? Maybe for you. However for engineers an IPC is the most important metric about CPU architecture performance.

Problem no. 1 is that you don't want to see that x86 AMD Ryzen 3950X is loosing vs. Apple A13 in overall performance (50.02 vs 52.82) despite its much higher frequency 4.6GHz vs 2.65 GHz. Any x86 CPU is terribly slower than Apple's ARM in therms of IPC.
Problem no. 2 is that generic Cortex core A77 is also a bit faster than Zen 2. This is not a prediction of x86 will be in trouble in the future. This is actually measured values confirming the fact that x86 is in trouble already IMHO. Feel free to show us your own measurement to proving x86 superiority. I'd prefer less talking and more graphs and numbers please.


IPC calculations of SPECint2006:
  • - 9900K .... 54.28/5 GHz = 10.86 pts/GHz
  • - 3950X .... 50.02/4.6 GH = 10.87 pts/GHz
  • - A76 ........ 26.65/2.84 GHz = 9.38 pts/GHz
  • - A77 ........ 33.32/2.84 GHz = 11.73 pts/GHz ...... +8% IPC over 9900K
  • - A11 ........ 36.80/2.39 GHz = 15.40 pts/GHz .... +42% IPC over 9900K
  • - A12 ........ 45.32/2.53 GHz = 17.91 pts/GHz .... +65% IPC over 9900K
  • - A13 ........ 52.82/2.65 GHz = 19.93 pts/GHz .... +83% IPC over 9900K
  • - A14 ........ 66.00/3.00 GHz = 22.00 pts/GHz (estimated +10%IPC)
  • - Zen3 ...... 60.00/4.60 GHz = 13.04 pts/GHz (estimated +20% IPC)


For example Ampere eMag 80-core ARM server CPU with A76 will be hard competitor to AMD's 64c EPYC. Those 16-cores advantage is exactly +25% more. This compensates SMT2 feature (usually +15-25% performance boost) while having same performance with just 80 threads in compare to Epyc's 128 threads (SMT2). AMD needs Zen 3 to be something much better than just poor Zen 2.
I'm not mixing anything. The slide said "Max Cores / Threads". That's such a big secret that it must be hidden until launch, meanwhile AMD announced x86-64 and it took them three years to get there? I don't want to play this ARM game here either, this isn't an ARM relevant thread. I'll gladly talk to you about ARM in another thread.

Regarding Problem #1; Why would I hate to see a miraculous CPU? Let Apple keep developing ARM and win the market. I have no loyalty to AMD or x86. I just want the best. Also, I don't know who taught you, but IPC is application dependent. A higher frequency raises all boats. So don't tell me IPC is the most important metric. It is one of many things that need to be considered.
 

soresu

Golden Member
Dec 19, 2014
1,532
757
136
That's such a big secret that it must be hidden until launch, meanwhile AMD announced x86-64 and it took them three years to get there?
As much as I agree with you on many other things, I think you may be reaching a bit too far back in time for reference to their business strategy.

AMD changed massively between then and now - including divesting their fabs, buying ATI, over promising on and botching BD, dramatically cutting their prices for years to become basically the higher profile x86 value brand (VIA being th lower profile one), and the resultant scaling back of their facilities and staff to its leaner state prior to Zen/Ryzen's runaway success.

Suffice to say there has been some significant turn over too in the management during those times.
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
As much as I agree with you on many other things, I think you may be reaching a bit too far back in time for reference to their business strategy.

AMD changed massively between then and now - including divesting their fabs, buying ATI, over promising on and botching BD, dramatically cutting their prices for years to become basically the higher profile x86 value brand (VIA being th lower profile one), and the resultant scaling back of their facilities and staff to its leaner state prior to Zen/Ryzen's runaway success.

Suffice to say there has been some significant turn over too in the management during those times.
Fair enough. Still, AMD and Intel love to show figures beforehand when they are good. Intel did with Conroe and Sandy Bridge. AMD did with Zen. They generally don't leave many secrets. With the original Zen we knew a lot other than final clocks.

Now when things aren't going great, they tend to hid numbers.Intel never boasted the numbers of the Pentium D before launch, nor did AMD boast anything regarding BD IIRC. I was kind of out of the enthusiast space when BD was in the works. Hence my long join date but relatively few posts.

I guess what I am trying to say is that AMD has no reason to keep this mythical SMT4 a secret at this point. It just doesn't exist on Zen 3 (opinion). It's not like AMD is beating Intel in just about every aspect in the server space already, right? They don't need some "secret sauce" to get on top, they are on top.
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
Excepting dat thicc 52% IPC figure - you got to hand it to them, that was certainly a nice surprise to say the least.
Indeed. Even they were saying 40%. I still remember reading the leaks about an 8C 65W CPU and all I thought was, "No way". But they did it.
 
  • Like
Reactions: soresu

soresu

Golden Member
Dec 19, 2014
1,532
757
136
Indeed. Even they were saying 40%. I still remember reading the leaks about an 8C 65W CPU and all I thought was, "No way". But they did it.
Still running my R7 1700 to write this on.

They went one better with the 2700E at 45W, but sadly chose to keep it out of the hands of Joe public.

It makes me wonder what an equivalently clocked 8C CPU only TDP would be with Zen2, 15-25W would be nice.
 

RetroZombie

Senior member
Nov 5, 2019
464
383
96
They went one better with the 2700E at 45W, but sadly chose to keep it out of the hands of Joe public.

It makes me wonder what an equivalently clocked 8C CPU only TDP would be with Zen2, 15-25W would be nice.
It would be great if they used something like that in gaming laptops.
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
Still running my R7 1700 to write this on.

They went one better with the 2700E at 45W, but sadly chose to keep it out of the hands of Joe public.

It makes me wonder what an equivalently clocked 8C CPU only TDP would be with Zen2, 15-25W would be nice.
Wow, 45W? Ryzen Master is/was supposed to have an eco mode that limits the TDP to the next range down. A 95W becomes a 65W, a 65W becomes a 35W, etc. I don't see it on my 2600X. I can limit wattage in the BIOS though.
 
Last edited:

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
Still running my R7 1700 to write this on.

They went one better with the 2700E at 45W, but sadly chose to keep it out of the hands of Joe public.

It makes me wonder what an equivalently clocked 8C CPU only TDP would be with Zen2, 15-25W would be nice.
Well at 65W my 2600X settles at 3.55GHz in Furmark. Wonder what it can do at 25W?

EDIT

OK, for reference,
at 95W it does 4.0GHz
at 65W it does 3.55GHz
at 25W it does 1.25GHz
 
Last edited:
  • Like
Reactions: lightmanek and OTG

itsmydamnation

Platinum Member
Feb 6, 2011
2,153
1,674
136
Strange, it's usually a hard set clock limit and wandering voltage/watt draw.
Not how amd cards have worked for a few generations now. its why undervolting generally leads to higher clocks without changing target clock and why AMD introduce the terminology of gaming clock without actually changing the way Navi does its clocking compared to Vega.
 

uzzi38

Golden Member
Oct 16, 2019
1,220
2,250
96
  • Like
Reactions: lightmanek

RetroZombie

Senior member
Nov 5, 2019
464
383
96
Well at 65W my 2600X settles at 3.55GHz in Furmark. Wonder what it can do at 25W?

EDIT

OK, for reference,
at 95W it does 4.0GHz
at 65W it does 3.55GHz
at 25W it does 1.25GHz
What about 35W and 45W?
Because the uncore eats up too much power leaving almost nothing to the cpu cores. Even 8 core Epycs uses lot's of power.

I wonder why amd never released some Epyc 5000 series (TR platform based) and Epyc 3000/Opteron 3000 series (am4 platform based). It seams they have a large gap (read market) to fill in servers.
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
13,922
3,584
136

Leaks.

Leaks everywhere
About dang time 7nm laptop CPUs and APUs launched. I've had no incentive to upgrade from a i7-6700HQ laptop, but a 8c/16t laptop CPU with 7nm power efficiency would give me a good reason to upgrade.
 
  • Like
Reactions: krumme and Markfw

Veradun

Senior member
Jul 29, 2016
564
780
136

Leaks.

Leaks everywhere
CES is in less than a month :>

Sent from my ONEPLUS A6013 using Tapatalk
 

Thunder 57

Golden Member
Aug 19, 2007
1,572
1,550
136
What about 35W and 45W?
Because the uncore eats up too much power leaving almost nothing to the cpu cores. Even 8 core Epycs uses lot's of power.

I wonder why amd never released some Epyc 5000 series (TR platform based) and Epyc 3000/Opteron 3000 series (am4 platform based). It seams they have a large gap (read market) to fill in servers.
at 95W it does 4.0GHz
at 65W it does 3.55GHz
at 45W it does 3.05GHz
at 35W it does 2.5GHz

at 25W it does 1.25GHz

45W numbers are interesting if the Renoir rumors of 45W 'H' models are true.
 

ASK THE COMMUNITY