AMD launches Zen+ 12nm Ryzen and X470 motherboards

Topweasel · Apr 20, 2018

positivedoppler said:
I think we might be right back to cooling. The one used by anandtech has 1/2 the flow rate of the wraith.

Taking it one step farther, lost of sites only used the Wraith in testing Ryzen and used AIO's for 8700k testing. It's entirely possible that Anandtech is getting all its boosts properly on Ryzen with the wraith (which might be rare) and everyone else is getting the best boost performance out of their 8700k's?

PingSpike · Apr 20, 2018

Topweasel said:
Taking it one step farther, lost of sites only used the Wraith in testing Ryzen and used AIO's for 8700k testing. It's entirely possible that Anandtech is getting all its boosts properly on Ryzen with the wraith (which might be rare) and everyone else is getting the best boost performance out of their 8700k's?

Maybe. That's an interesting result on its own if cooling makes such a difference. Didn't Intel remove boost clocks from their spec or something recently?

Its even hard to say on the spectre meltdown patches since there have been so many of them.

PotatoWithEarsOnSide · Apr 20, 2018

Topweasel said:
Taking it one step farther, lost of sites only used the Wraith in testing Ryzen and used AIO's for 8700k testing. It's entirely possible that Anandtech is getting all its boosts properly on Ryzen with the wraith (which might be rare) and everyone else is getting the best boost performance out of their 8700k's?

It would not surprise me if everyone apart from Anandtech (plus one other reviewer) were using Windows Ryzen Balanced Mode, which is effective for gen 1 but not for gen 2; gen 2 should be using the bogstandard Windows Balanced Mode.

positivedoppler · Apr 20, 2018

Topweasel said:
Taking it one step farther, lost of sites only used the Wraith in testing Ryzen and used AIO's for 8700k testing. It's entirely possible that Anandtech is getting all its boosts properly on Ryzen with the wraith (which might be rare) and everyone else is getting the best boost performance out of their 8700k's?

This makes me really appreciate that there are many review sites each doing things a little bit differently. If not, things like cooler differences will never come to light. I won't be overclocking, and the difference in price between the 8700K and 2700 is small enough to make it a difficult decision between the two. But if the 2700X pulls ahead on almost everything when both are at stock and I don't plan on overclocking then it makes it an easier choice if I need to spend another $100 on cooling for the 8700K to keep up.

csbin · Apr 20, 2018

Ryzen 2000 Application Performance Meta Overview: ~1570 benchmarks from 26 launch reviews compiled

https://www.3dcenter.org/news/ryzen...tate-zur-anwendungs-performance-im-ueberblick

Atari2600 · Apr 20, 2018

positivedoppler said:
I think we might be right back to cooling. The one used by anandtech has 1/2 the flow rate of the wraith.

Which is still more than the cooler Intel provide for the price. [I wonder how well the 8700K would run in comparison with just the naked heatspreader....]

If others are going to add in expensive coolers to the 8700K, then include the price of it when they do their price/performance comparisons.

Paratus · Apr 20, 2018

Topweasel said:
Taking it one step farther, lost of sites only used the Wraith in testing Ryzen and used AIO's for 8700k testing. It's entirely possible that Anandtech is getting all its boosts properly on Ryzen with the wraith (which might be rare) and everyone else is getting the best boost performance out of their 8700k's?

If AT ran HWINFO64 in the background of each test we could probably get to the bottom of this.

It records everything.

Temps - CPU, System, GPU, SSDs
Frequency - CPU, GPU, RAM
Voltages - Per Core, Vcore, GPU, MEM
Power, PSU, CPU per Core
Core loading

I wonder if they have more data to play with.

Based on the performance of my 1900X @ 4.15G (CB R15 Single 17X, Multi 18XX) the 2700X is performing where it should.

If we could see the 8700K, temps, frequency, Vcore and VID we could probably narrow down what’s going on.

Well @Ryan Smith, does Ian have more data laying around?

Fanatical Meat · Apr 20, 2018

With the risk of bringing a P&N discussion into this thread, am I crazy for preferring intel because the majority of their cpu are not made in China?

Whitestar127 · Apr 20, 2018

csbin said:
Ryzen 2000 Application Performance Meta Overview: ~1570 benchmarks from 26 launch reviews compiled

https://www.3dcenter.org/news/ryzen...tate-zur-anwendungs-performance-im-ueberblick

Nice! I don't suppose they have a separate one for just the game benchmarks?

jpiniero · Apr 20, 2018

Fanatical Meat said:
With the risk of bringing a P&N discussion into this thread, am I crazy for preferring intel because the majority of their cpu are not made in China?

Uhh, I thought most of the CPUs were being fabbed at GloFo's plant in upper NY.

positivedoppler · Apr 20, 2018

Fanatical Meat said:
With the risk of bringing a P&N discussion into this thread, am I crazy for preferring intel because the majority of their cpu are not made in China?

I like made in USA too, but most US electronics companies depends in FABs outside of the US. Intel would love for us to go USA fab only which would make them almost the only game in town. Good bye xilinx, amd, apple, qualcomm, ibm, nvida, stc, etc...

Fanatical Meat · Apr 20, 2018

jpiniero said:
Uhh, I thought most of the CPUs were being fabbed at GloFo's plant in upper NY.

Maybe I’m wrong, reviews from yesterday seemed to display China on the chips. I know some are made in Malaysia.

guskline · Apr 20, 2018

I just updated the BIOS on my Asus Crosshair VI Hero to 6004 which is the newest released BIOS on the Asus website.

My 2700x is listed on the FedEx website as on the truck for delivery today.

I have my 1800x under custom water so hopefully the switch will be painless.

I'll keep you all posted.

BTW, for the poster who said we should all imitate Moonbogg, does that include the requirement that we eat catfood????. Say what you want, but Moonbogg is a man of his word!

zinfamous · Apr 20, 2018

csbin said:
Ryzen 2000 Application Performance Meta Overview: ~1570 benchmarks from 26 launch reviews compiled

https://www.3dcenter.org/news/ryzen...tate-zur-anwendungs-performance-im-ueberblick

Nice. seems like relative performance to 8700k, then, is a pretty wide spread across all reviewers and AT isn't any kind of outlier, really. Consistent point here is big performance improvement over 1800X and it is tracking to be better than 8700K--between 1% and 20% faster, lol. I guess with a little maturity with the new boards and BIOS and memory this will become a more consistent picture.

positivedoppler · Apr 20, 2018

While we're on the subject of made in the USA. I'm Taiwanese ethnicity. I recall 25 years ago when my cousins visit me here in the US, they would go hunt for that made in the USA tag while shopping because they recognize that made in the USA means high quality premium product. Compare AMD's wraith cooler and then Intel's cheap heat spreader and cheap stock cooler, I think one brand is starting to win that long term brand recognition of delivering premium component while another is establishing themselves as being cheap.

Dasa2 · Apr 20, 2018

if cooling was the problem cinbench would be different but its the same between most reviews (except techradar and when you look at how there 8700k is using 68w load vs anandtech 130w its fairly obvious there cpu is throttling for some reason)
one reason for this could be that cinbench doesn't care about ram performance or gpu performance which means that the reason for the performance difference in other programs probably has nothing to do with the cpu

coercitiv · Apr 20, 2018

Fanatical Meat said:
Maybe I’m wrong, reviews from yesterday seemed to display China on the chips. I know some are made in Malaysia.

Diffused in USA.
Made in China.

guskline said:
does that include the requirement that we eat catfood????. Say what you want, but Moonbogg is a man of his word!

Did someone say... catfood?!

VirtualLarry · Apr 20, 2018

AFAIK, both AMD and Intel both have packaging plants in China, as well as elsewhere. But they are fabbed in the USA. (AMD has fabs in Dresden and Ireland, but AFAIK, all Ryzen production is done in New York.)

The Stilt · Apr 20, 2018

moonbogg said:
Anyone here can test the 8700K in cinebench R15 for single thread. Just make sure you have the latest windows patch, right? Also, if anyone wants to imitate moonbogg for a day, you must understand, very deeply, the meaning of the word hyperbole. You must then realize that this word is BS and no longer exists in your vocabulary, because your insane passion just got real, bihtce.

In my test, a stock i7-8700K scores 201 in CB15, when the system is fully GPZ patched.

Core i7-8700K
ASUS PRIME Z370-A (0613 w/ µCode 0x84)
G.Skill FlareX 3200C14 2x8GB, 2666MHz CL14-14-14-32
Windows 10 Enterprise x64 16299.251 / 16299.334
CFL Notes: Power limits were manually enforced and verified (95W PL1, 119W PL2, 1 second Tau). 1C = 4.7GHz, 2C = 4.6GHz, 3C = 4.5GHz, 4-5C = 4.4GHz, 6C = 4.3GHz

Fanatical Meat · Apr 20, 2018

VirtualLarry said:
AFAIK, both AMD and Intel both have packaging plants in China, as well as elsewhere. But they are fabbed in the USA. (AMD has fabs in Dresden and Ireland, but AFAIK, all Ryzen production is done in New York.)

From the Anand tech review

Again I don’t want to turn this into P&N, I’m moderately concerned about Chinese chips in machines. I know it’s nearly unavoidable but I’d like to mitigate it as much as reasonably possible.
What does “diffused” mean in chip manufacturing?

Dasa2 · Apr 20, 2018

https://www.computerbase.de/2018-04..._benchmarks_mit_ddr43466_und_scharfen_timings

Chicken76 · Apr 20, 2018

Has anyone tested the 2000-series for the linux compilation bug of the 1000-series?

The Stilt · Apr 20, 2018

Fanatical Meat said:
From the Anand tech review

Again I don’t want to turn this into P&N, I’m moderately concerned about Chinese chips in machines. I know it’s nearly unavoidable but I’d like to mitigate it as much as reasonably possible.
What does “diffused” mean in chip manufacturing?

Each and every piece of silicon used in Ryzen CPUs comes from Fab 8 (Saratoga County, NY).
Only the final (physical) assembly steps are made outside the US.

Fanatical Meat · Apr 20, 2018

The Stilt said:
Each and every piece of silicon used in Ryzen CPUs comes from Fab 8 (Saratoga County, NY).
Only the final (physical) assembly steps are made outside the US.

Thank you

sandorski · Apr 20, 2018

The CPU results and mini-drama of AT's results has been largely hashed out, but what about Motherboards? Not just reviews of 470's, but how does 470 compare to 370?

One quick review(link below) of my last question seems to indicate little difference, performance wise, but the review was quite limited. Further testing of this comparison seems warranted.

370v470

AMD launches Zen+ 12nm Ryzen and X470 motherboards

Diamond Member

Lifer

Senior member

Golden Member

Senior member

Golden Member

Lifer

Lifer

Senior member

Lifer

Golden Member

Lifer

Diamond Member

No Lifer

Golden Member

Senior member

Diamond Member

No Lifer

Golden Member

Lifer

Senior member

Senior member

Golden Member

Lifer

No Lifer