ZEN ES Benchmark from french hardware Magazine

[majord]

Right, I'm starting to think that CB score might be too optimistic and not reflective of overall performance, compared to Intel's quads at least. According to the latest benchmarks 8C/16T Ryzen at 3.2-3.3 GHz all core Turbo would edge stock 7700K with half the cores by only 10-12.5% in applications and still lose badly in CPU limited games - and that's before you overclock the latter to 5 GHz (easily done in many reviews). Go figure.

right... except, Ryzen is launching @ 3.4ghz+ ?

and by CPU limited games, you mean , specifically ones from 2011-2013 or so?

 

[USER8000]

The fact it only performs 11% behind the 6900K in games with a gimped clock paints a good picture. Hopefully the 4C/8T version can OC to 4.3GHz or so on air, it will probably do very well against the 6700K / 4790K.

Plus some here were also trying to paint the games suite as being very multi-threaded. It isn't - the FX8370 and A12-9800 are virtually the same meaning the Ryzen chip tested must have much stronger cores over Piledriver and Excavator.

If you look at the FX8370 which is operating at 4GHZ to 4.3GHZ against the 3.1GHZ to 3.3GHZ Ryzen chip,the latter is 32% faster despite a huge drop in clockspeed.

 

[AtenRa]

Your words sounds familiar...lol...hmm let me think when ive heard them before...oh yeah...bulldozer release lmao...yeah but when software will use more cores bulldozer will stomp everything...that was 5 years ago...and here we are...almost 2017 and an i3 humiliates bulldozer...

Same story repeats himself over an over again...

Yeap completely humiliation for a 14nm 2015 CPU to be slower than 32nm 2011CPU in 2016/2017 games. And we havent touched the FrameTimes yet, but I guess you will be one of those that will recommend the $170 Core i3 in 2017.

 

[USER8000]

If AMD is doing well in games which are traditionally crap on their CPUs which need decent single core performance,then what happens when games like Watch Dogs 2 and Battlefield 1 which do OK on the FX8350 are run on Ryzen??

Anybody who is a enthusiast and gamer wants Ryzen to be good to improve competition,which pushes Intel to release better CPUs and improves pricing on their side too. Its a win-win for the consumer.

I really don't understand why some want this to be a fail - maybe they have just bought a new CPU and are peed off if Ryzen is decent and the second hand value will drop?? Outside that I really don't understand the logic of not wanting more choice in the market.

 

[SlowSpyder]

If these benches are real, not a bad showing for 3.3GHz. An ES may not have a fully functioning turbo, may have lower base clocks, may have some diabled cache, etc. Encouraging signs from AMD.

 

[dark zero]

I hope that those numbers gets near true, that would mean that the 2nd processor brand is back and the x86 market would continue their evolution without problems.

Now after that, I wonder who will be 1st to try to get back to the phone market....

Also... any info about AMD K12? That would be great to watch since they would go after Qualcomm.

 

[lixlax]

Seems that AMD really hit the nail (finally) with Rysen. It doesn't beat Intel 10/8 cores in MT nor 4 or less cores in lightly threaded applications but it seems that AMD haven't been closer to Intel in terms of IPC since Conroe launched in 2006. What have been shown so far the IPC seems to be around Haswell level which is quite a bit better than I was expecting. The 40% gain per clock goal was met or even exceeded.

If AMD can get the clocks 3-400MHz higher as promised (comared to this sample), then they'll have a really competitive lineup if priced correctly.

Edit: and AMD is finally about to beat its K10 in terms of IPC.

 

[inf64]

I went on to compare CPU limited benchmarks from other websites and found a nice CPU limited game selection on another French IT website, hardware.fr (their Broadwell-E review).

http://www.hardware.fr/articles/946-6/performances-jeux-3d.html

Let us see how much faster is 6700K vs 8350 in hardware.fr review :

6700K scored 149.5 pts

FX8350 scored 90.3 pts.

Difference between the two is 1.65 or 65%.

Now results from CanardPC’s Ryzen preview:

6700K scored 118.2%

FX8370 scored 73.6%.

FX8370 is 4.3Ghz max boost Vs 4.2Ghz max boost which is around 2.3% difference.

6700K is 1.60 or 60% faster than FX8370 which fits perfectly in line with hardware.fr’s result of being 65% faster than FX8350 in CPU limited games. When one accounts for slight clock difference and a bit of different game mix you end up at 3% point difference in results between 2 reviews.

Now, from the same hardware.fr review we have FX 9590 vs FX 8350 results in game benchmark suite.

As can be seen from hardware.fr’s FX 9590 review , Turbo clock makes no difference in games they tested, results are basically the same. In FX 9590 review this chip is 12.5% faster than 8350, while due to different game selection in 6900K review, the difference is 9%. As can be seen, even on the FX series the performance jumped with clockspeed significantly, not 1:1 but clearly at least 70-80% of what was expected.

Conclusion : Ryzen clocking via Turbo boost to ~3.8Ghz (for example) will likely raise the CPU limited scores by solid 8% making it almost 4790K level (in this chart it should score 129pts Vs 6900K’s 132pts and 149pts for 6700K). On top of that comes manual OCing which should negate some of the advantage of higher out of the box clock of 4C/8T intel parts like 4790K and 6700K.



On to the productivity benchmarks comparison between CanardPC’s preview and hardware.fr.

From the same 6900K review on hardware.fr we have this kind of difference between 6700K and 8350 in prod. benchmarks:

6700K scored 181.1 pts

FX8350 scored 138.3 pts

According to this review 6700K is 1.31x or 31% faster than FX8350 in these kind of workloads.


Compare with CanardPC’s article:

6700K scored 137.3%

FX8370 scored 105.2%.

According to CanardPC, 6700K is 1.3x or 30% faster than FX8370 (which in turn is a percentage point or two faster than FX8350 as we can see perfectly demonstrated on hardware.fr).

Conclusion: If AMD ships Ryzen with just 10% better clocks (3.4Ghz base/3.6Ghz all core Turbo/3.8Ghz ST Turbo), this sort of Ryzen SKU should score around 245pts on hardware.fr performance scale. That is right in between 5960x and 6900K , 3.2% faster than 5960x and 4.2% slower than 6900K.



So we have a perfect match in results, in both games and productivity between two French websites.

Scaling the Ryzen to likely release clockspeeds (3.4Ghz base + Turbo) we get something of ~4-5% lower than 6900K productivity performance and 2-3% lower than 6900K gaming performance. Stock vs stock.

 

[TennesseeTony]

You put a lot of work into that post inf64, good info, and thanks.

 

[Charlie22911]

I'm very curious what Ryzen will overclock to, I've got two personal x99 systems and have had a total of 4 Broadwell-E chips (6900k, 3 6850k)... I can tell you from personal experience that you are very lucky if you can hit 4.5GHz with 4.2GHz being what most people can probably reasonably expect to achieve with reasonable voltage.

Being that Broadwell-E is an overclocking turd it could potentially even give the total performance advantage to AMD. Very exciting, even if my expectations are a bit lofty.

 

[USER8000]

I'm very curious what Ryzen will overclock to, I've got two personal x99 systems and have had a total of 4 Broadwell-E chips (6900k, 3 6850k)... I can tell you from personal experience that you are very lucky if you can hit 4.5GHz with 4.2GHz being what most people can probably reasonably expect to achieve with reasonable voltage.

Being that Broadwell-E is an overclocking turd it could potentially even give the total performance advantage to AMD. Very exciting, even if my expectations are a bit lofty.

Well AMD just needs to get close enough to the Core i7 6900K,and they have the advantage of using a much cheaper platform using only dual channel memory. They also probably will bundle the Wraith cooler which should do the job perfectly fine if you are not overclocking.

 

[vissarix]

Snip

After your previous very reliable predictions in the past...which happened to be all wrong

We should take this one with just a little grain of salt, like a truck load one

 

[unseenmorbidity]

Your words sounds familiar...lol...hmm let me think when ive heard them before...oh yeah...bulldozer release lmao...yeah but when software will use more cores bulldozer will stomp everything...that was 5 years ago...and here we are...almost 2017 and an i3 humiliates bulldozer...

Same story repeats himself over an over again...

That is not sound logic. Sure, some people long ago were wrong, but that doesn't make this person wrong.

Have you even looked at recent benchmarks?! The freaking 8350, which is crap, is actually starting to compete with i5 kabylakes!

I5's are no longer good enough for anything gaming, except budget/upgrade machines. If you want a futureproof CPU, and why wouldn't you, seeing as it will last the better part of a decade, then you might as well make it a good one. That means you will need at least 8+ threads going into the future, and ideally 12+.

 

[inf64]

The panic mode has been activated in some users it seems. Nice
I find it amusing that people are trying to feverishly push 400$ 4C/8T pre-OCed parts with even links to deals on other sites. Isn't that against the forum rules(don't know, just asking)?

Exciting 2-3 weeks ahead, especially now that cat is out of the bag

 

[unseenmorbidity]

The panic mode has been activated in some users it seems. Nice
I find it amusing that people are trying to feverishly push 400$ 4C/8T pre-OCed parts with even links to deals on other sites. Isn't that against the forum rules(don't know, just asking)?

Exciting 2-3 weeks ahead, especially now that cat is out of the bag

Yah, it's actually kind of entertaining to watch the trolls flail about with ridiculous hail mary fallacious arguments.




You just got off a a vaction and you're back to this?


esquared
Anandtech Forum Director

 

[lolfail9001]

I find it amusing that people are trying to feverishly push 400$ 4C/8T pre-OCed parts with even links to deals on other sites. Isn't that against the forum rules(don't know, just asking)?

I find it more amusing you are pushing for purchasing part that does not actually exist yet: $400 Ryzen 8-core, namely.

 

[inf64]

I find it more amusing you are pushing for purchasing part that does not actually exist yet: $400 Ryzen 8-core, namely.

I have made zero recommendations to buy Ryzen. Unlike some others who recommend those overpriced OCed quad cores, all with links to online shops .
Also I have no idea how much will any of the parts cost, all we have are some wild speculations on the internet.

PS Anyone is free to buy anything they like. I bought i5 4690K because it provided best perf/$ for example,so I have really no dilemma what is best for my wallet

 

[railroadmaster]

I think some people are overreacting. This processsor should be fine for games, more to the point Processor performance isn't a substantial bottleneck games. Heck a AMD phenom ii or Intel Core 2 Quad can run most games no problem. If what you do with a computer is gaming and not much else, this Processor won't be bad it will be overkill. Moreover games in the future will be able to better utilize multi threaded processors and AMD processors thanks to the fact that most games are also on consoles which use AMD cpus+gpus.

While I do sometimes like to play games, computers aren't just for video games. Processors also do other tasks. This processor will be great for video editing, professional audio production, photo editing, video conversion, running virtual machines, Computer aided design/3d rendering or programming/software development. This is also good for multi tasking. Compiling or building applications from source or through an SDK is super CPU intensive. Linux will be able to properly utilize those RYZEN cores to their full potential.

I am sure a lot gamers will buy this but thr architecture will have its biggest impact on workstations.

 

[Abwx]

The panic mode has been activated in some users it seems. Nice

....

 

[.vodka]

Thanks inf64 for the writeup and somewhat of a confirmation of the results. Everything looks in order so far for all the parts in the comparison, and these numbers for Ryzen are similar to what we've been given at new horizon.

Only two weeks to CES.

-------------------------------------

https://www.reddit.com/r/AMD_Stock/comments/5k31sr/i_bought_canard_pc_there_you_go/

Better pictures of the publication, these are actually legible

Also, it seems there's more to read in the magazine, lots of interesting information. As provided by this user:

They said they didn't have time to bench Kaby Lake against Zen in the benchmarks, but there is another section in which they bench Kaby Lake, I posted them in the last picture.

Anyway it's 100% legit

If someone could explain the table with all the values where they compare Bristol Ridge, Summit Ridge and Kaby Lake, I don't understand anything

Also they think it's unlikely Zen will be available in January given the current state of the platform, even a March mass availability is optimistic.

Info I got from reading their file on Intel:

1. They have internal sources at Intel and the climate there is currently awful apparently, employees are very discouraged and some feel like "they have nothing left to lose". They are making big profits now by totally screwing up the future. According to the magazine, Intel is "probably in the most delicate situation it has had to face to this day". The CEO is reducing cost so much that he's managing engineers like "supermarket cashiers", he doesn't care about taking the time to train them.
   
   
2. Krzanich is very impatient and eager and keeps changing his mind about projects. If a new architecture isn't created in like a couple weeks, he gives up and cancels the project... he keeps sending contradictory instructions to the teams.
   
   
3. "Fab Hell": Intel is likely going to have a 6 month delay on 10nm. Worse, even Cannon Lake is not expected to feature any significant architectural improvement. Basically Intel was just hoping AMD would keep not competing with them.
   
   
4. Krzanich is apparently a disaster, and he won't be able to stay CEO for long. Apparently some people have heard him yelling from the next building when he was angry. But he seems unaware of him being perceived so negatively. Employees at Intel hope Murthy Renduchintala will replace him ASAP, and he seems much more capable and is slowly refocusing Intel in the right path, but basically R&D is redactedatm and there will be a huge empty space until about 2019. Apparently, Krzanich completely underestimated the possibility of an AMD comeback.
   
   
5. (I didn't really understand that point I'm not expert enough) but apparently x86 is going to disappear sooner than expected, it'll be replaced by ARM and Intel is panicking about that.
   
   
6. Intel is currently working on a "multichip package" (MCM) integrating an Intel CPU and an AMD GPU. So this is confirmed guys.
   
   
7. In the picture before the last picture I posted, it is said that Kaby Lake has exactly the same IPC as Skylake. No improvement as to perf/watt but there seems to be more room for overclocking.

Their conclusion is that if AMD doesn't mess up the launch, "Zen obviously has the necessary capacities to shake up the market..."

If so, it looks like it's an even better situation than 2003 + Athlon 64 for AMD... since Intel won't be doing any of their anti competitive crap anymore.

Intel's roadmap after Skylake always seemed a bit... fuzzy... Coffee lake added at the last minute until Icelake... this could explain it.

 

[f2bnp]

Let's just wait and see. The leaked tests are looking way better than ever expected to be honest. We'll know in a little more than a week's time for sure.

I really love seeing people trying to spin the results though.

 

[railroadmaster]

With regards to the FX Processors, I mean they aren't particularly bad especially if you can find a deal on them. Sure they aren't as good as Intel, but they are older and haven't beem updated since 2013/2014 if I am not mistaken. AMD haters like to show them as proof of AMD's incompetence, but to be honest they've held up very well despite not having an update in a long time. Given that the FX have held up as long as it has, Zen as well Zen plus will hold up even longer. The only thing AMD processors have been really bad at is power consumption, which Zen more than addresses.

 

[StinkyPinky]

It all comes down to clock speeds. 8 core is great, but a 4.4GHz quad core compared to a 3.3GHz octo-core....most people are going to choose the former.

Of couse their hex and quad core cpus will also be great options. Exciting times ahead.

 

[DerpEinstein]

It all comes down to clock speeds. 8 core is great, but a 4.4GHz quad core compared to a 3.3GHz octo-core....most people are going to choose the former.

Of couse their hex and quad core cpus will also be great options. Exciting times ahead.

Yah, well most people aren't very intelligent.

 

[AtenRa]

Just to cheer up a little bit