AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

[KentMan]

My own calculations suggest 4.1-4.3GHz max for the Ryzen ES, based upon known 2700X scores and the fairly widely predicted 9-13% IPC increase for Zen 2.
1900 for 2700X at 4.3GHz.
2057 for ES at unknown GHz.
1900*1.09= 2061 which implies clock parity.
1900*1.13= 2147 which implies ES was running lower than 4.3GHz. <5% difference in score, implying maximum clock difference of 0.2GHz.

Crude calculations of course, and CB runs do have varying results.

If we look at it from a 9900K perspective, with 2700X supposedly having a 5% MT IPC lead:

2700X at 4.3GHz = 1900
2700X at 4.7GHz = 2077
2700X at 4.3GHz with 9900K IPC = 1806
2700X at 4.7GHz with 9900K IPC = 1973
9900K at 5.0GHz = 2138 as per easyrider's image, and at 2700X IPC it'd be 2245.
In order for the ES to score 2057 at 4.7GHz there would need to be IPC regression from the last gen. That is simply not plausible given the core changes that have been made.
Given that the ES did only equal the 9900K, it could not have been clocked higher than 4.5GHz.

So we have an absolute upper limit for how that ES was clocked, and I very much doubt that it was even near to it.

I said this earlier in the thread, that i suspected the rumored 15% IPC increase they showed off with Epyc2 vs Epyc1 would translate to approximately that on the demo, that the ES chip was probably clocked to the same as the 2700x. Your maths just reinforces this for me.

So if the chips can actually clock to 4.7 or higher, there are going to be some very nervous people sweating at Intel.

 

[Potatowithearsontheside]

Just as a fun little side note, notice how every tech journalist is now talking about Ryzen 3000 but not PCPer whose owner now 'officially' works for Intel are knocking out video after videos about cases....

I do hope they aren't court cases.

 

[Hagar]

Just as a fun little side note, notice how every tech journalist is now talking about Ryzen 3000 but not PCPer whose owner now 'officially' works for Intel are knocking out video after videos about cases....

Ryan has now dropped to insignificance for me, when he used to be the go to for information on AMD.

 

[Potatowithearsontheside]

I'd like to know more about the rumour of a 29% IPC uplift in certain workloads. Specifically, what workload that applies to, and how different that workload is to a CB 15 MT bench.

Edit: my suspicion is it being AVX workloads, where AMD is a long way behind anyway.

 

[hominid]

It's was obvious for years that PCPer were shills for both Intel and NVidia, it was good that they were shown what they were.. Ryan and Alan jumped to Intel officially soon afterwards - PCPer will die in due course.

 

[humbug]

I do hope they aren't court cases.

Ryan has now dropped to insignificance for me, when he used to be the go to for information on AMD.

My opinion of him has been pretty low for a while but i didn't know him back in those days, for me he's been shilling for both nVidia and Intel for years, every time nVidia had something about AMD be it true like stuttering on the 7950 or utter BS like Ghosting on Free-Sync Ryan Shrout was there with all the specialised equipment and all the knowledge of how to prove nVidia are right.

He's a slime-ball

 

[subbytna]

As I replied to beany above, and others have also suggested, I think a draw was entirely the point.

If you buy something, then soon after something new comes along that matches it. You are then told that the "new" thing isn't even the finished article, but a test product. You would know that the finished article will be "better" or more polished than the test product was. I too think that it was done to almost match the 9900k with the reasoning being that people will KNOW it's not the finished article and will be better still come the launch time, thus generating chat about "just how good will this be when it's finalised with proper BIOS"

 

[Devrij]

I think what people should consider is that regardless of whether you can overclock a 9900k to beat this chip, and even if the finished article is no better than the engineering sample, you can drop a new cpu in later without having to upgrade your whole system as AM4 will have support for another gen. With a 9900k your chances of that are very slim given intels history. So in 2 or 3 years time when I get the itch, I can sell on my 2700x and drop in the latest AM4 chip. I might lose some platform features, but I save the cost of a new mobo, which is at least £100. That's not something to be ignored.

 

[djkav]

How do you think it will OC? Any better than Ryzen1000/2000? Any techy guys know why ryzen was a fairly poor (you could says "consistent" [consistently small]) overclocker and would that change in zen 2?

I think Zen 1 (Ryzen 1xxx) and Zen 1+ (Ryzen 2xxx), were single die designs. Zen 2 (Ryzen 3xxx) utilises the Threadripper/Epyc chiplet design. So it's difficult to compare Zen 2 OC potential to the Zen 1 design, due to this design difference.

I think the HEDT Ryzen 9 3850x 16/32 will be around £499.

 

[Troezar]

If you buy something, then soon after something new comes along that matches it. You are then told that the "new" thing isn't even the finished article, but a test product. You would know that the finished article will be "better" or more polished than the test product was. I too think that it was done to almost match the 9900k with the reasoning being that people will KNOW it's not the finished article and will be better still come the launch time, thus generating chat about "just how good will this be when it's finalised with proper BIOS"

Agreed. AMD need people talking positively about themy. They need mind and market share. I think they're well on the way and bringing decent value performance to the new to pc customer will help that a lot. They need to engage the younger crowd who don't remember Athlon or the dirty tricks and show there is an alternative to Intel/Nvidia.

 

[4K8KW10]

I'd like to know more about the rumour of a 29% IPC uplift in certain workloads. Specifically, what workload that applies to, and how different that workload is to a CB 15 MT bench.

Edit: my suspicion is it being AVX workloads, where AMD is a long way behind anyway.

The application is DKERN +RSA, it is not AVX.

In DKERN +RSA AMD's Zen 2 architecture offered a staggering 4.53 IPC score, while their "Zen 1" generation CPU achieved 3.5 IPC, giving Zen 2 a boost of 29.4%. Some websites have been fixated on this 29.4% number, forgetting its context, but let's have a better discussion as to this data's relevance.

As AMD's footnotes confirm, this data refers to IPC (Instructions Per Cycle), which means that clock speeds should not play a major role here, making this 29% boost down to pure architectural advancement, not an increase in clock speed alone. It is also worth noting that this benchmarking data includes both floating point and integer benchmarks, which is another factor that is worth considering.

https://www.overclock3d.net/news/cp..._ipc_is_29_4_higher_than_zen_1_in_dkern_rsa/1

 

[LePhuronn]

Agreed. AMD need people talking positively about themy. They need mind and market share. I think they're well on the way and bringing decent value performance to the new to pc customer will help that a lot. They need to engage the younger crowd who don't remember Athlon or the dirty tricks and show there is an alternative to Intel/Nvidia.

Then AMD should really be shouting to the heavens that the PlayStation and XBox are both powered by AMD.

 

[d_brennen]

I'd like to know more about the rumour of a 29% IPC uplift in certain workloads. Specifically, what workload that applies to, and how different that workload is to a CB 15 MT bench.

Edit: my suspicion is it being AVX workloads, where AMD is a long way behind anyway.

Lucky AVX isn't the be all and end all of consumer or even enterprise server really... Epyc 2 usues 256bit wide AVX units doesn't it? Not many will lose sleep over it

 

[beany_bot]

People are talking about AMD needing mindshare and such... hell maybe 5 years ago they did. Those days are over. I'm telling you Intel are the underdogs now. We should be "supporting" them LOL. Zen2 will just further destroy intel.


** Image removed - Censored maybe but it's still a recognised competitor **

 

[d_brennen]

^^ Box CPUs make up a small percent of CPU sales, the vast majority of computer users have never seen a naked CPU.

 

[4K8KW10]

People are talking about AMD needing mindshare and such... hell maybe 5 years ago they did. Those days are over. I'm telling you Intel are the underdogs now. We should be "supporting" them LOL. Zen2 will just further destroy intel.

Except that you need to take into account that: https://store.steampowered.com/hwsurvey/processormfg/
When this changes in AMD's favour, we can speak on the topic again.

 

[Hagar]

My opinion of him has been pretty low for a while but i didn't know him back in those days, for me he's been shilling for both nVidia and Intel for years, every time nVidia had something about AMD be it true like stuttering on the 7950 or utter BS like Ghosting on Free-Sync Ryan Shrout was there with all the specialised equipment and all the knowledge of how to prove nVidia are right.

He's a slime-ball

Off topic, but he built a big house and a decent business off of AMDMB.com 15 years and more ago, an internet success.

I would tend to agree

 

[beany_bot]

Except that you need to take into account that: https://store.steampowered.com/hwsurvey/processormfg/
When this changes in AMD's favour, we can speak on the topic again.

Steam hardware survey takes into account Chinese servers and all sorts of legacy crap with a steam account. Isnt the most popular CPU on that like a Pentium lol?

 

[d_brennen]

^^ Gamers also make up a small minority of CPU purchases

Honestly, forms exist in their own little bubble, insulated from reality. Company revenue and gross margin are all that matter to Intel and AMD

 

[Journey]

^^ Gamers also make up a small minority of CPU purchases

Indeed, the lack of wide ranging APU's from AMD is going to harm them in the business client machine market, they need to address that as quickly as possible. Obviously if they can push out a 4c/8t CPU with a low power 7nm GPU stuck inside at 30w and as fast as an i5/i7 it'll no doubt help market share especially from the OEM's.