AMD launches Zen+ 12nm Ryzen and X470 motherboards

How are these so called leaked reviews valid when they are using an x370 motherboard without proper bios support to show what a 2700x is capable of? My takeaway is that it works out of the box but needs proper motherboard support to see what the chip is truly capable of....more like a preview of what is coming. I will wait for proper reviews from (H) and others before passing judgement.
 
Not to debate your decision to wait, but how is the new system going to cost the same or more importantly, LESS?

At this time, a 1600X costs $219.99, with no cooling. If you factor in getting a usable cooling unit, you're going to add at least 30 bucks to the cost.

If the 2600X price of $229 is valid, then it would certainly be the less expensive option, since it comes with a Wraith Spire unit.
 
At this time, a 1600X costs $219.99, with no cooling. If you factor in getting a usable cooling unit, you're going to add at least 30 bucks to the cost.

If the 2600X price of $229 is valid, then it would certainly be the less expensive option, since it comes with a Wraith Spire unit.

Agreed. The amount of times I've quoted a 1600x system for a user and found that a 1700 system with stock cooler would be mere tens of dollars more..
 
I don't think I'd be comfortable running my cpu at 1.476V 24/7. That seems way too high for my liking.

It's at the high end of what is safe but comfort level is definitely up to your own personal tastes.
 
How are these so called leaked reviews valid when they are using an x370 motherboard without proper bios support to show what a 2700x is capable of? My takeaway is that it works out of the box but needs proper motherboard support to see what the chip is truly capable of....more like a preview of what is coming. I will wait for proper reviews from (H) and others before passing judgement.

the x370's have had bios support for zen+ for a couple months now. there will probably be another bios to fine tune a few settings around release day but i doubt there will be a drastic difference between them. but ultimately it's just a starting point to compare with proper reviews.. either way i'm more interested in the 400 series changes then anything else to see if they're worth grabbing.
 
https://www.chiphell.com/thread-1841382-1-1.html

2600X+DDR4 3660 CL14

5acab2bea49fe.png


Compare to core i7-6700K :

http://www.legitreviews.com/ddr4-me...nding-the-best-ddr4-memory-kit-speed_170340/3

3866 17-21-21-38 : 40.9ns

so far 15ns behind Intel but massive improve on Ryzen system,with New Bios on X470 maybe up to 5ns.
 
So according to El chapuzas, the 2700X is 8% faster than 1700X on Wprime MT, 12.8% faster in CB15, and 11.6% faster in x264. The average is about 11%.

According to the CPC review the 2700X was 13.6% faster than the 1700X in MT.

13.6% > 11%

And el chapuzas used a top X370 mobo. :D

Blah blah and you ignore the weird latency, memory speed, driver issues and other erroneous results (which clearly indicate there is more on the table) to again cherry pick and push your toxic attempts to steer conversation to 'muh AMDEE SUXX'.
You have a bad name for yourself across the internet. If I didn't know better I'd say you were a paid AMD hack, in order to make Intel fanboys look like a bunch of rabid idiots, because that's the end result of your attempts to make Intel look perfect, especially on a site like [H].
 
https://www.chiphell.com/thread-1841382-1-1.html

2600X+DDR4 3660 CL14

View attachment 65394

Compare to core i7-6700K :

http://www.legitreviews.com/ddr4-me...nding-the-best-ddr4-memory-kit-speed_170340/3

3866 17-21-21-38 : 40.9ns

so far 15ns behind Intel but massive improve on Ryzen system,with New Bios on X470 maybe up to 5ns.

Case in point. This is exactly what I have been waiting for, the IMC tweaks will be the biggest changes for gaming, probably on par with the expected 300-400MHz core bump.
 
Blah blah and you ignore the weird latency, memory speed, driver issues and other erroneous results (which clearly indicate there is more on the table) to again cherry pick and push your toxic attempts to steer conversation to 'muh AMDEE SUXX'.
You have a bad name for yourself across the internet. If I didn't know better I'd say you were a paid AMD hack, in order to make Intel fanboys look like a bunch of rabid idiots, because that's the end result of your attempts to make Intel look perfect, especially on a site like [H].

Amusing that you talk about "toxic attempts" then start another round of personal attacks. Didn't you read what I said to you in the other thread? I don't know how to say that I laugh at what some people has to say about me. Maybe I would change my signature to make my point more clear. And don't miss that I am comparing AMD to AMD. It is you guys who are continuously introducing Intel in the discussion again and again, I am not.

The performance measured by the CPC review of Zen+ agrees with the performance measured by AMD (as I already demonstrated). El chapuzas changed the mobo and the results are similar. Latencies and drivers is another excuse. I did heard similar claims during Zen launch and how new drivers and new BIOS/AGESAs with latency improvements would change things radically for Zen. Guess what? It didn't happen. The performance of Zen after the improved AGESA with latency reduction was virtually the same

getgraphimg.png


There is no magic drivers/AGESA for Zen+. Performance is what it is.
 
Last edited:
They did.I couldn't bring my Mobo to DDR4 3200 Cl 15 but only DDR4 2993 Cl16. That french reviewer Tested with same memory kits in both AGESA 1.0.0.2 and 1.0.0.4a .so don't say any shit about improve since you don't have Ryzen system. , even some members of OCN were able to reach 3600 CL 14 with modded bios on ASUS prime X370 pro which they couldn't do it before.
 
Last edited:
Review: AMD Ryzen 5 2600 (Chipset X370)
BY IVÁN MARTÍNEZ
04/09/2018
CPUS & APUS
AMD-Ryzen-Logo-740x339.jpg


Following the leaks before the launch date, with us today we have the new AMD Ryzen 5 2600 , the younger brother of the Ryzen 7 2700X that we analyzed last week. Of course, like the Ryzen 5 1600 of the previous generation, offers 6 cores and 12 threads , which increase their frequencies to the base of 3.40 GHz , and can reach 3.90 GHz in Boost mode.

Let's see first its complete technical specifications :

AMD Ryzen 5 2600
Socket AM4
Fabrication process 12 nm
Cores / Threads 6 cores / 12 threads
Base Frequency 3.40 GHz
Turbo Frequency 3.90 GHz
Cache L3 2 × 8 MB
Compatible Memory DDR4-2933 Dual-Channel
PCIe lines 16 lines
TDP 65W
Index of contents [Ocultar]

Packaging and Accessories
AMD-Ryzen-5-2600-01-694x600.jpg

AMD-Ryzen-5-2600-02-694x600.jpg

AMD-Ryzen-5-2600-01-1-694x600.jpg

AMD-Ryzen-5-2600-02-1-694x600.jpg

The AMD Ryzen 5 2600 comes in a box similar to the previous generation design, showing the processor from the side and protecting it perfectly inside, as well as the heatsink.

AMD-Ryzen-5-2600-03-740x555.jpg


AMD-Ryzen-5-2600-04-644x600.jpg


In addition to the warranty card and a Ryzen 5 sticker, this model comes with the AMD Wraith Stealth dissipator, a model already seen in the first generation and which offers nothing new.

AMD-Ryzen-5-2600-05-732x600.jpg


AMD-Ryzen-5-2600-06-740x535.jpg


AMD-Ryzen-5-2600-07-732x600.jpg


AMD Ryzen 5 2600
AMD-Ryzen-5-2600-08-740x577.jpg


As we anticipated, the AMD Ryzen 5 2600 offers 6 cores , which in turn offer 12 threads that start from a base frequency of 3.4 GHz to reach 3.9 GHz in Turbo mode thanks to Precision Boost technology. It offers a total of 2 × 8 MB of L3 cache , as well as 6 x 512 KB of L2 cache and 6 x 32 KB + 6 x 64 KB of L1 cache, so it offers a cache distribution identical to that of the Ryzen 5 1600 .

AMD-Ryzen-5-2600-09-740x369.jpg


As we can see in CPU-Z , the set of instructions supported has MMX (+), SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4A, x86-64, AMD-V, AES, AVX, AVX2, FMA3 and SHA, also the same as in the first generation of Ryzen .

AMD-Ryzen-5-2600-10-740x577.jpg


Obviously it is a processor for the AM4 socket and therefore makes use of the " Pin Grip Array " connection system, where the pins are in the processor and not in the motherboard, highlighting the use of a total of 1,331 contacts.

Test Equipment
AMD-Ryzen-5-2600-11-672x600.jpg


To analyze the performance of the AMD Ryzen 5 2600 processor we have assembled the following test equipment:

AMD-Ryzen-5-2600-12-740x493.jpg


Unlike the Ryzen 7 2700X that we already analyzed, this model has not allowed us to configure the memories above 2933 MHz , something that reminds us of what happened in the previous generation, although having no optimized drivers for the chipset was a possibility . In addition, the impossibility of playing on this platform is maintained , something that is sure to be solved at its official launch or with other motherboards.

Tests
Next, we show you the results of the performance tests performed on this and other processors in calculation benchmarks (wPrime), rendering (Cinebench), coding (x264), memories (Aida64) or the most popular focused on games (3D Mark ).

AMD-Ryzen-5-2600-Tests-01.jpg


AMD-Ryzen-5-2600-Tests-02.jpg


AMD-Ryzen-5-2600-Tests-03.jpg


AMD-Ryzen-5-2600-Tests-04.jpg


AMD-Ryzen-5-2600-Tests-05.jpg


AMD-Ryzen-5-2600-Tests-06.jpg


AMD-Ryzen-5-2600-Tests-07.jpg


Temperature, Overclocking and Consumption
Under an ambient temperature of 20ºC, the AMD Ryzen 5 2600 offers temperatures of about 28ºC at rest and about 45ºC in its stress test, a really excellent value for such a compact heatsink. This is also not too noisy, around 40 dB at full performance, although easily improved by more advanced models.

On the consumption , say that it has improved notably remaining around 140W for the complete equipment in the stress test of Aida64 (without discs or GPU), a big difference compared to the 200W that the Ryzen 7 2700X has with the same configuration.

AMD-Ryzen-5-2600-13.jpg


Overclocking is still not the strong point of these processors, because although they have increased their frequencies compared to the previous generation, we have only managed to take them up to 4.09 GHz with 1.292V . The Ryzen Master OC application has not been able to use it for not having it updated to give compatibility to these processors, but without problems from the BIOS.

conclusion
AMD-Ryzen-5-2600-99-740x577.jpg


The AMD Ryzen 5 2600 supposes a revision of frequencies on the Ryzen 5 1600 , something that improves its performance around 10% . We speak of an improvement of the kind that Intel usually offers us, so the price is the one that will mark whether it compensates or not.

For now, the platform seems somewhat green , although we will have to wait to have X470 motherboards to check their performance in games, which is very important for the lower performance they offered in the previous generation against Intel.

However, it offers a very good overclocking capacity , since it exceeds 4 GHz being a model of the simplest, although it is something to be expected when other models allow it and this only seems a limitation of frequencies.

If the prices that we saw in Amazon Germany last week are true , it will go on sale at around 215 euros , competing against the 8th generation i5 , which also offer 6 cores , although not HyperThreading and therefore "only" 6 threads

Main advantages
+ Good performance / price ratio for professional environments
+ Excellent multi-core performance
+ Very good temperatures and consumption
+ 6 cores / 12 threads
+ Good overclocking capacity, for its factory frequencies
+ Compatible with previous generation base plates
Negative aspects
- The platform is still somewhat green
Other Aspects
* We do not know his performance in games
Therefore, from El Chapuzas Informático we awarded the Gold Award to this processor for its performance in tasks of calculation, rendering and coding, in which the brute force of its 12 threads matters, but the test in 3DMark presages that it does not will have the same luck in games .

Galardon-Oro.png




Read more: https://elchapuzasinformatico.com/2018/04/amd-ryzen-5-2600-review/
 
I just want to see if Kyle gets 4.4 out of the 2700X. That's the big one for me. If he can pull that off, I'll go buy one just for kicks. If not... I'll wait around until Zen 2 next year.
 
Honestly, the wraith stealth seems to be pretty darn impressive for a stock cooler. Obviously, you can get better aftermarket coolers, but for something AMD is throwing in, being able to OC at all is actually pretty nice.
 
Honestly, the wraith stealth seems to be pretty darn impressive for a stock cooler. Obviously, you can get better aftermarket coolers, but for something AMD is throwing in, being able to OC at all is actually pretty nice.

Free coolers are great for some. Not for the [H]ard, though.
 
They did.I couldn't bring my Mobo to DDR4 3200 Cl 15 but only DDR4 2993 Cl16. That french reviewer Tested with same memory kits in both AGESA 1.0.0.2 and 1.0.0.4a .so don't say any shit about improve since you don't have Ryzen system. , even some members of OCN were able to reach 3600 CL 14 with modded bios on ASUS prime X370 pro which they couldn't do it before.

They measured performance using the same memory modules to check if the latency reduction introduced by the new AGESA changed things significantly. AMD did claim 6ns reduction at same clocks), so one has to check AMD claims using the same memory configuration, not comparing different overclocking configurations.
 
Yeah, if he's getting 4.1 on the stock cooler, that's really impressive. 1xxx chips were 4.0 on high end air cooling and ~3.8 on stock cooling, so bodes well for the high end 2xxx chips on high end air or water. I'm really interested in seeing if the boost curves are editable, I could see 4.5-4.6 on 2 cores with 4.3 on 8 core (for a 2700x) which would give a pretty big boost in games, especially older ones.
 
Ok, it's a bump in perf. I still say the 2800 is conspicuously missing in the lineup. Early adopters=crying as is borne out by history. Process yield and perf improve over time just as the smallest feature is 12nm and they dial it in. Lots of good words on the 7nm process so far too. I'm figuring just wait for the high bin "12nm" coming. TR 2 may be a killer product. It's an exciting time in the semi world.
 
They measured performance using the same memory modules to check if the latency reduction introduced by the new AGESA changed things significantly. AMD did claim 6ns reduction at same clocks), so one has to check AMD claims using the same memory configuration, not comparing different overclocking configurations.

So You're saying AMD was lying ? Sure AMD did claim and It reduced latency by 6ns but I don't think this do serious impact on Performance.only on latency-sensitive applications/Games : like Hitman
 
Last edited:
New AGESA really did help a lot to many Ryzen Users.some of people were able to reach 4.1GHz which they couldn't do it before, some people like me were able to reach 3200Mhz CL15 from 2993 Cl 16 which they couldn't do it before.But if you want to troll , I really don't think 6ns can impact on Performance or not but AMD can claim last AGESA really did Impact on Performance.I said ability to reach clock or latency like 4.1Ghz or 3466Mhz CL 14 on Asus prime X370 Pro

AMD promises regarding the 1.0.0.4 AGESA update are well-known. So sites as the one that I mentioned in #329 or others as OC3D tested what AMD promised, not what you imagine they would test. If you pay attention, OC3D also used the same memory kits for both AGESA versions, because that is what there is to do for a proper testing. You having memory issues with a former AGESA version is not enough reason for reviews to invent stuff. FYI the "french reviewer" could get 3200 CL16 with the launch AGESA. Now ignore the point again...

################################ EDIT

I see you edited entirely your post, after I replied it. I leave your original post and my answer to it above and I am replying your edit below

So You're saying AMD was lying ? Sure AMD did claim and It reduced latency by 6ns but I don't think this do serious impact on Performance.only on latency-sensitive applications/Games : like Hitman

Check the OC3D review or check this

getgraphimg.png


Using AIDA64, reviewers found a reduction in latency of about 6ns. Now, was that reduction in measured latency accompanied by an improvement in performance? No. As the graph in #329 showed the performance gain was exactly 0.0%. Next I quote from the "french reviewer":

In itself, 6ns is the equivalent of switching from a CL16 memory to a CL12 memory, which is a massive jump that can yield, according to our article, between three and five percent performance in applications limited by memory ( the case for example of 7-Zip) and almost 4% in games.

[...]

We therefore wanted to measure the practical performance gap obtained. For the sake of time (we only got this BIOS last Friday, we thank Asus who provided us with a beta version for our X370 motherboard test), we went on the R7 1700 to make the comparison:

getgraphimg.png


We will spare you the graph of games, the performances are there also strictly identical between the two versions of the AGESA ... In front of this syllogism we carried out multiple checks:
  • Perform the test again on another motherboard (Gigabyte AX370 Gaming 5 in BIOS F5J)
  • Watch performances with two memory modules only in DDR4-2400
  • Watch performances with two memory modules only in DDR4-3200

In any case the result is without appeal: there is strictly no difference in performance. So we contacted AMD who confirmed the thing, there does not seem to be any difference in performance related to this latency gain obtained under Aida64.

So AMD confirmed the findings of the "french reviewer" about the latency improvement measured by Aida64 providing zero performance gains even in memory sensible workloads. So the reviewers asked AMD for a clarification:

Still, on the issue of latency gain, AMD has offered us at the time we write these lines no response. Is this a specific optimization for algorithms used by software like Aida64 to measure latency? A generous interpretation would say that AMD may have reduced values in these programs to values closer to reality. The fact that this change has no real impact on performance, and that AMD sent to the press motherboards equipped with AGESA 1.0.0.4a BIOS configured with the two options mentioned above, makes us doubt our generosity, we admit it!

Take your own conclusions.
 
Last edited:
So AMD confirmed the findings of the "french reviewer" about the latency improvement measured by Aida64 providing zero performance gains even in memory sensible workloads. So the reviewers asked AMD for a clarification:

How come ? Here :

http://i.imgur.com/68n8LJN.png

Done by The Stilt.Yes Huge impact between 3200 CL12 and 3200 Cl 16 which 6s is equal to CL12 to CL16. How can they get zero performance by 6ns ? Perhaps they should test again between DDR4 3200CL21/CL16 or they could show reader which application/games will get benefit from 6ns.
 
the x370's have had bios support for zen+ for a couple months now. there will probably be another bios to fine tune a few settings around release day but i doubt there will be a drastic difference between them. but ultimately it's just a starting point to compare with proper reviews.. either way i'm more interested in the 400 series changes then anything else to see if they're worth grabbing.[/QUOTE/]

Thanks for the info but I was referring to the youtube video where the guy had trouble maintaining an all core overclock of 4.3 and benched at 4.250 at 1.4 volts and inability to benchmark games without crashing and then the 4.5 overclock at 1.679 volts which has since disappeared hence I felt that the motherboard recognized the 2700x but was not optimized for it.
 
How come ? Here :

http://i.imgur.com/68n8LJN.png

Done by The Stilt.Yes Huge impact between 3200 CL12 and 3200 Cl 16 which 6s is equal to CL12 to CL16. How can they get zero performance by 6ns ? Perhaps they should test again between DDR4 3200CL21/CL16 or they could show reader which application/games will get benefit from 6ns.

The answer to your question is in my former message. Read it carefully.
 
I exactly wrote what quote said.if they say :
In itself, 6ns is the equivalent of switching from a CL16 memory to a CL12 memory, which is a massive jump that can yield, according to our article, between three and five percent performance in applications limited by memory ( the case for example of 7-Zip) and almost 4% in games.
then How come they get zero performance but "The Stilt" did test on hitman and got 16% more fps? OC3D only tested AIDA.either AMD used an algorithms to show fake numbers or Bench is wrong.Who knows.

Edit : I found one :

https://thetechaltar.com/amd-ryzen-agesa-1-0-0-4-testing/
 
Last edited:
All this whoo haaa about the new Intel and AMD cpus.. and here I am still running my 8 year old AMD Phenom II X6 1090T rig and playing all the kick ass games with my 2x ASUS VG278HE on NVIDIA GeForce GTX 660 Ti in SLI
Damn 8 year old rig still kicking it.. Maybe I should upgrade to a new rig...
 
All this whoo haaa about the new Intel and AMD cpus.. and here I am still running my 8 year old AMD Phenom II X6 1090T rig and playing all the kick ass games with my 2x ASUS VG278HE on NVIDIA GeForce GTX 660 Ti in SLI
Damn 8 year old rig still kicking it.. Maybe I should upgrade to a new rig...

I know the feeling, went from a release purchased phenom II 940 to my r5 1600. If it wasn't for that board starting to die I'm not sure if I would of upgraded yet.
 
Memory Latency.If someone can find 3666 Cl14 on Skylake.then we can predict how much behind Intel.i think max up to 17ns.
 
Last edited:
So it's either kickin it, or it's not kickin it. make up ur mind. Steal the thread and make it about mining or your next upgrade. we are here for ya.
I'm just saying that my system is 8 years old and still able to play all these games. I might not get all the bells and whistles of the new games today but my system is still performing and stable even with windows 10. SO my internal debat is do I want to spend the money for a new right to run things with all the bells and whistles?
 
I'm just saying that my system is 8 years old and still able to play all these games. I might not get all the bells and whistles of the new games today but my system is still performing and stable even with windows 10. SO my internal debat is do I want to spend the money for a new right to run things with all the bells and whistles?

The answer to the question "to upgrade or not to upgrade" is almost always "upgrade." FYI.
 
The answer to the question "to upgrade or not to upgrade" is almost always "upgrade." FYI.

And I still have a core2quad that is running as my linux based PVR / file server. Anyways I believe on May 19th (when I will be a few miles away from MC instead of ~200) I believe a Ryzen 2700 and X470 board and 16GB of unbuffered ECC DDR4 2666 will replace that..
 
Im almost out of patience for this to drop. Been holding on to this 3570k longer than any other processor Ive ever owned lol
 
I exactly wrote what quote said.if they say :

then How come they get zero performance but "The Stilt" did test on hitman and got 16% more fps?

Once again, the answer to your question is in #342. Read it carefully, because they explain why CL16 --> CL12 RAM improves performance, but AGESA 1.0.0.2 --> 1.0.0.4 does not.
 
Back
Top