Speculation: Ryzen 3000 series

Page 121 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DrMrLordX

Lifer
Apr 27, 2000
22,021
11,594
136
Are we talking about someone 'who joined yesterday' and has already posted a lot of controversial stuff, including antagonizing and trashing other people's opinion?

Wouldn't be the first time, won't be the last. If a paid viral marketer shows up and says stuff that is opinionated and/or based in facts then there isn't much mods can do about it, short of monitor for activity from the same IP (multiple accounts).

Sure. Let's ignore all the other not so nice things he said and concentrate on one theoretical weakness of the Intel platform. Unless, of course, you can show me a real world example of such a weakness being exploited.

Someone else already pasted evidence of the 9900k being slowed down by mitigations intended to stop said vulnerabilities. That's the real problem, though I do not think this is the thread to discuss such things in isolation.

How many "theoretical" weaknesses do you suppose there are in Matisse?

Other people also posted opinions and the results were different.

Feel free to complain in the appropriate subforum if you think users aren't being treated fairly.
 

PotatoWithEarsOnSide

Senior member
Feb 23, 2017
664
701
106
The 9900K's "theoretical" weakness is that it has 8 fewer cores than it's AMD top-tier competitor. Of course, on Monday that stops being a theoretical weakness and becomes a factual weakness.
 
Feb 4, 2009
35,238
16,707
136
Looks like some of Intel fans are so much vulnerable as their CPUs./s

I know I don’t get it. Just buy what you want/like/can afford.

I prefer intel generally, I feel they typically offer a better product however this has changed with Ryzen 2000 but not totally maybe it will change more with Ryzen 3000. Intel historically has had better reliability and better standards for their partners, better compatibility.
Intel has become lazy and hasn’t offered anything compelling in years, seems like every new release is a refinement from a years old design.

AMD is hungry and doing something different, is priced well and appears to have substantially upped their quality, even AMDs partners appear to be more quality focused.

Bottom line if you like AMD for reasons that is fine as long as you are honest and not pooping on every intel discussion
If you prefer intel because of stability and overall performance great stick with intel just don’t poop on every AMD discussion.
 

.vodka

Golden Member
Dec 5, 2014
1,203
1,537
136
So you agree with his statements about Intel products being "hot garbage" and "trash heaps."? That doesn't need addressing? Just ask the mods to get rid of the report button if it serves no purpose. If it's okay too, then let's be consistent.

In light of recent events (2018 and later)... they are.

First, the hot part.

What do you think is the result of Intel rehashing Skylake on improved 14nm iterations all these past years due to the 10nm debacle? The 9900k is a furnace because of only being able to push clock speeds higher to increase performance. 5GHz ST turbo is nice, but if you want that juicy MCE all core 5GHz turbo or a manual 5GHz all core OC, then you'd better be prepared to cool that thing down. You get FX9590 power consumption, and the performance to go along with it.

Now, let's go to the trash part which is about the glaring hardware vulnerabilities.

If you've been paying attention since the start of 2018, Intel's products all the way back to P6/Pentium Pro are designed in a way that can be exploited (save for the most secure modern x86 design, 1st gen in-order Atom Bonnell). They are in the process of dying from a thousand cuts, mitigations. Once you apply all of them for all known (stress in KNOWN as there will probably be even more ways to exploit the hardware since it's the problem) you're already taking a ~ -15% performance hit. If you need to go as far as disabling HT for your security needs, then you get another significant hit on top.

https://www.phoronix.com/scan.php?page=article&item=mds-zombieload-mit&num=1

If looking at the geometric mean for the tests run today, the Intel systems all saw about 16% lower performance out-of-the-box now with these default mitigations and obviously even lower if disabling Hyper Threading for maximum security. The two AMD systems tested saw a 3% performance hit with the default mitigations. While there are minor differences between the systems to consider, the mitigation impact is enough to draw the Core i7 8700K much closer to the Ryzen 7 2700X and the Core i9 7980XE to the Threadripper 2990WX.

Some vulnerabilities like SPOILER can't be mitigated by software. Yeah...

Mom and pop browsing facebook won't notice the slowdown, but everyone else doing productive work (from individuals to datacenters) with their machines is bound to be... angry, by the result. That number is for Skylake... if you go back to Sandy/Ivy well, it gets ugly since they lack the INVPCID instruction that Haswell/Broadwell and above have to speed up one of the mitigations. Nehalem and earlier, even uglier.

The effective IPC/performance per clock gap after mitigations between Zen+ and Skylake has disappeared or outright turned into AMD's favor... and that brings us to AMD's handling of these vulnerabilites. Out of all known so far, AMD only gets hit by parts of Spectre and it's been patched.

No, the hit piece done by CTS labs on AMD doesn't count as these "vulnerabilities" required root access on the system to exploit. Yeah, nice try. On the other hand, if you ask Intel hardware nicely for the keys to the kingdom, it'll hand them over.



So, Intel's only way to power through the performance hit right now is thanks to clock speeds, and that's alright... but that doesn't cover all the locked parts out there and previous, still completely viable generations that got a significant downgrade.

I'm pretty sure whatever they have in the pipeline ready to come out (Icelake) thanks to the 10nm train wreck is a nice product and probably has fixed many or all of these vulnerabilities without the associated performance loss... but that is still far, far away, and Zen2 launches or is revealed on Monday. Willow Cove with its redesigned cache and security features (complete Spectre fix?) is even further away.



To put things in perspective, if AMD still were competing with faildozer derivatives (or even somehow not gone bankrupt), then it wouldn't be fair to call Intel's product hot garbage even in this mess. When you have competition as good as what Ryzen is right now and will be shortly, also a product that is clearly not vulnerable to the same attacks, well, yeah. You get the point.

If the +~10-15% IPC increase over Zen+ is real (seems to be according to leaks) apart from the confirmed improved SIMD performance AND that performance is gained without any hardware vulnerabilities (why would they start cutting corners now?) AND the rumors of 4.5-5GHz operation are confirmed too, then we can completely dismiss Intel parts as complete, total hot garbage until Icelake comes out.
 
Last edited:

RaV666

Member
Jan 26, 2004
76
34
91
I know I don’t get it. Just buy what you want/like/can afford.

I prefer intel generally, I feel they typically offer a better product however this has changed with Ryzen 2000 but not totally maybe it will change more with Ryzen 3000. Intel historically has had better reliability and better standards for their partners, better compatibility.
Intel has become lazy and hasn’t offered anything compelling in years, seems like every new release is a refinement from a years old design.

AMD is hungry and doing something different, is priced well and appears to have substantially upped their quality, even AMDs partners appear to be more quality focused.

Bottom line if you like AMD for reasons that is fine as long as you are honest and not pooping on every intel discussion
If you prefer intel because of stability and overall performance great stick with intel just don’t poop on every AMD discussion.

Intel had a great products, but they just got worse and worse with their consumer approach.
First, block overclocking.Then sell K units.
Then add crappy paste to the mix.Keep 4 cores forever. Then INCREASE prices (broadwell) at this point i got fed up.Then constant socket changes, still price increases and products like KF , so remove igp , KEEP THE PRICE.
Argh.
I bought x99 5820K, and at that time i was somewhat happy with it, 6 cores, not much price increase from 4 core, 4 channels, i thought that when broadwell launches im gonna get 8 core.But no. even tho cpus were much smaller they increased prices :-[ .After that i knew i would wait until either intel changes or amd does someting decent, yes i bought 5960x used, but i already knew my next platform is gonna be ryzen.
And now im on the mindset that i would rather pick a bit worse amd product even at the same price because of the constant abuse they did to the DIY market.
And with the impending 10 core skylake on a yet another socket i guess they learned nothing.
 

jpiniero

Lifer
Oct 1, 2010
15,154
5,686
136
So, Intel's only way to power through the performance hit right now is thanks to clock speeds, and that's alright... but that doesn't cover all the locked parts out there and previous, still completely viable generations that got a significant downgrade.

Intel does have an updated stepping (R0) of Coffee Lake Refresh which has the hardware fixes similar to what Cascade Lake has. It's not available yet but will be by 7/7.

Which brings me to my point - figure that Intel will source updated chips for reviewers in time for the 3000 series reviews. I wonder if any reviewers will disclose that...
 
Feb 4, 2009
35,238
16,707
136
Intel had a great products, but they just got worse and worse with their consumer approach.
First, block overclocking.Then sell K units.
Then add crappy paste to the mix.Keep 4 cores forever. Then INCREASE prices (broadwell) at this point i got fed up.Then constant socket changes, still price increases and products like KF , so remove igp , KEEP THE PRICE.
Argh.
I bought x99 5820K, and at that time i was somewhat happy with it, 6 cores, not much price increase from 4 core, 4 channels, i thought that when broadwell launches im gonna get 8 core.But no. even tho cpus were much smaller they increased prices :-[ .After that i knew i would wait until either intel changes or amd does someting decent, yes i bought 5960x used, but i already knew my next platform is gonna be ryzen.
And now im on the mindset that i would rather pick a bit worse amd product even at the same price because of the constant abuse they did to the DIY market.
And with the impending 10 core skylake on a yet another socket i guess they learned nothing.

Similar but different for me.
My nearly 10 year old Q9659 machine has been trouble free until about 2 years ago, small failures then bigger mysterious failures. That’s a different subject but I feel that system has delivered outstanding value.
I build my wife a Facebook web machine and wanted to keep it cheap so I went AMD, 3 mother board swaps due to failure or something not working out of the box. Thank god I purchased microcenters warranty for them. I finally give up on the gigabyte(?) board and pay a little more for an Asus board, this board looked pretty good had 1080P video on the board and I thought awesome. Took it home, opened the box and it is 1080P analog connection. The board had no DVI or hdmi. I’m disappointed I got snookered. Get an adapter from bestbuy to make it work, get audio cable so audio works.
Fast forward three months, video dies on this board too. Go to micro center again, swap out the board and give up on on board video get a new card for my machine and move my old gtx 260 to this machine.
While swapping out the board I fumble the cpu and bend a pin, try to straighten it but fail, off to micro center again get another cpu.
I’ll take complete responsibility for
a) building a system for someone else
b) going cheap on components
c) fumbling the cpu

I won’t accept the overall poor quality all the boards had, I won’t accept a board being sold in 2013 with on board video not having at minimal DVI. I will say it’s highly unlikely I’d have so many failures with an intel board, even a cheap one. I wouldn’t have bent the cpu pin if I wasn’t removing in 5 times within 18 months.

However looks like the quality is back with Ryzen, only snookering that appears to happen is memory support & memory speed which hopefully gets addressed with Ryzen 3000.
 

exquisitechar

Senior member
Apr 18, 2017
683
940
136

DooKey

Golden Member
Nov 9, 2005
1,811
458
136
Think I'll be making a switch back to AMD.

I run both at the moment. I'll continue to use Intel for gaming until AMD has equal IPC for that task. For Distributed Computing I'll stick to AMD because it's the best bang for the buck with that kind of task. Win win.
 
Reactions: lightmanek

.vodka

Golden Member
Dec 5, 2014
1,203
1,537
136
What was the extra parameter to pass in the URL so the geekbench browser returns all the detailed data? Damn, I can't remember...

I'd like to check the clockspeeds during the entire run part of that data.
 

BigDaveX

Senior member
Jun 12, 2014
440
216
116
Reactions: beginner99

.vodka

Golden Member
Dec 5, 2014
1,203
1,537
136

tamz_msc

Diamond Member
Jan 5, 2017
3,865
3,729
136
If the systems were actually operating at the listed clock speeds and there was no turboing involved, it means that the Zen 2 sample was running at 3.2GHz, while the Coffee Lake was at 2.9GHz. Better than the original Zen, but still not quite level with Coffee Lake.
No. They're both operating at ~4GHz.
 
Reactions: french toast

PotatoWithEarsOnSide

Senior member
Feb 23, 2017
664
701
106
Similar but different for me.
My nearly 10 year old Q9659 machine has been trouble free until about 2 years ago, small failures then bigger mysterious failures. That’s a different subject but I feel that system has delivered outstanding value.
I build my wife a Facebook web machine and wanted to keep it cheap so I went AMD, 3 mother board swaps due to failure or something not working out of the box. Thank god I purchased microcenters warranty for them. I finally give up on the gigabyte(?) board and pay a little more for an Asus board, this board looked pretty good had 1080P video on the board and I thought awesome. Took it home, opened the box and it is 1080P analog connection. The board had no DVI or hdmi. I’m disappointed I got snookered. Get an adapter from bestbuy to make it work, get audio cable so audio works.
Fast forward three months, video dies on this board too. Go to micro center again, swap out the board and give up on on board video get a new card for my machine and move my old gtx 260 to this machine.
While swapping out the board I fumble the cpu and bend a pin, try to straighten it but fail, off to micro center again get another cpu.
I’ll take complete responsibility for
a) building a system for someone else
b) going cheap on components
c) fumbling the cpu

I won’t accept the overall poor quality all the boards had, I won’t accept a board being sold in 2013 with on board video not having at minimal DVI. I will say it’s highly unlikely I’d have so many failures with an intel board, even a cheap one. I wouldn’t have bent the cpu pin if I wasn’t removing in 5 times within 18 months.

However looks like the quality is back with Ryzen, only snookering that appears to happen is memory support & memory speed which hopefully gets addressed with Ryzen 3000.
Isn't that down to mobo manufacturers for making piss poor boards?
 

Thunder 57

Platinum Member
Aug 19, 2007
2,975
4,544
136
Intel does have an updated stepping (R0) of Coffee Lake Refresh which has the hardware fixes similar to what Cascade Lake has. It's not available yet but will be by 7/7.

Which brings me to my point - figure that Intel will source updated chips for reviewers in time for the 3000 series reviews. I wonder if any reviewers will disclose that...

That's good and all, but how do you know you are buying one of those new R0 steppings? You don't. At least not through your usual Amazon/Newegg types.
 
Feb 4, 2009
35,238
16,707
136
Isn't that down to mobo manufacturers for making piss poor boards?

Yes however AMD can choose to pull their logo from boxes that do not support a minimum feature standard also I’m confident AMD & intel each hold copyrights that the board partner need to make a board. I’m not positive but I am confident.
As I said bare minimum they could say no using our logo on any for your products
 

jpiniero

Lifer
Oct 1, 2010
15,154
5,686
136
That's good and all, but how do you know you are buying one of those new R0 steppings? You don't. At least not through your usual Amazon/Newegg types.

Yep. The recently announced non-K models are all R0, so they could announce something like a 9950K/9750K/9650K if they want.
 
Reactions: ZipSpeed

Thunder 57

Platinum Member
Aug 19, 2007
2,975
4,544
136
Yep. The recently announced non-K models are all R0, so they could announce something like a 9950K/9750K/9650K if they want.

They could, at least then you would know for sure.

As far as the non K models, I'm sure there are older ones that will continue to be sold though. It's not like Amazon would toss a bunch of inventory on an item that hasn't been recalled. Really a new model or whole generation is what Intel needs before this can go away.

It's not like AMD is infallible here either. I'm sure we all remember the TLB bug in Phenom. They did rename the new stepping by adding a 50 like you said Intel could do. You can see the B3 all end in 50. It would be nice if Intel did the same now.
 
Last edited:
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |