Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

[imported_jjj]

You can just run Fritz and see how it loads the cores with "8 processors" http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...and-discussion.2499879/page-195#post-38803783

 

[beginner99]

Hardware.fr already covered this in their review
http://www.hardware.fr/articles/956-7/impact-smt-ht.html

So the answer is yes for game (they also did apps where both Intel and AMD only gained from HT/SMT), but Ryzen suffers more. The worst title for Ryzen was BF1 where it lost nearly 14%, while for the Intel the worst was Watchdogs 2 with a ~7% difference.

But for BF1 the charts clearly show that it's not SMT alone that causes the issue since if you turn off core-parking, it's actually better than SMT off. So the scheduler/powerplan also plays a role here,

 

[agouraki]

Download that patch directly here: https://www.windowslatest.com/2017/03/15/direct-download-links-kb3213986-2/
(the KP4013429)

does it really work? caus i got the latest insider update in my windows and fritz is still all over the place when i put 8 threads.

 

[imported_jjj]

does it really work? caus i got the latest insider update in my windows and fritz is still all over the place when i put 8 threads.

Try to disable perf mode

 

[agouraki]

Try to disable perf mode

didnt work but i have done some tinkering with PPM program so i think i might have FUDGED up everything,

it seems my core parking isnt SMT aware on balanced mode but it doesnt migrate a single core thread on the other ccx tho so thats an improvement for single thread application.

Remember, we are a family oriented forum. So watch the language.

Moderator
Smoke

 

[Ajay]

Yes, 1T. Mine is on my $89 motherboard at 2933. I still have not tried 3200, but the others I saw said they had no problem @ 3200.

Oh, and thats @100% load for 2 days now. (WCG distributed computing)

Dayum, that's pretty sweet! So much for those who have been freakin out over RAM settings. Thanks for the reply. Oh, and figured you were running some D.C. project on it

 

[imported_jjj]

Found this on Dota 2 before and after the game update but more reliable data would be nice

https://www.chiphell.com/thread-1717769-1-1.html

 

[innociv]

Nice. 25% increase. But it was really low, so don't know how it compared to the 7600k or 7700k in their test methodology

 

[Ozzyrulez]

I was playing Metro 2033 today and had HWiNFO64 running to capture some temp and load info. My Ryzen system seems to now prefer using physical cores over logical. While no core was maxed the main cores were 60%+ while logical were in the teens.

 

[Malogeek]

I was playing Metro 2033 today and had HWiNFO64 running to capture some temp and load info. My Ryzen system seems to now prefer using physical cores over logical. While no core was maxed the main cores were 60%+ while logical were in the teens.

Balanced or High Performance power profile?

 

[Ozzyrulez]

Balanced or High Performance power profile?

High Performance. I wonder if Balanced would make a difference.

 

[Bouowmx]

For those interested in BOINC projects, a user tested AMD Ryzen 7 1700 3.3 GHz (not sure if boosted to 3.7 GHz) in single-thread OGR (yoyo@home) and RC5-72 (Moo! Wrapper): https://steemit.com/technology/@moi...-net-benchmark-on-amd-ryzen-7-1700-at-3-30ghz
"core" number represents a code path.

My results on Intel Core i5-7200U 3.1 GHz: https://imgur.com/a/Cpbqb

The most striking difference is RC5-72 AVX2: 23.2 vs 49.6 MKey/s

Try the bench yourself: http://www.distributed.net/Download_clients

 

[Rngwn]

For those interested in BOINC projects, a user tested AMD Ryzen 7 1700 3.3 GHz (not sure if boosted to 3.7 GHz) in single-thread OGR (yoyo@home) and RC5-72 (Moo! Wrapper): https://steemit.com/technology/@moi...-net-benchmark-on-amd-ryzen-7-1700-at-3-30ghz
"core" number represents a code path.

*snip*

Try the bench yourself: http://www.distributed.net/Download_clients

The AVX2 performance is halved here, "as expected". The tested Ryzen system was using single-channel RAM, pinch of salt must be taken here.

 

[Suspicious-Teach8788]

Wow I'm totally out of date with CPUs but I keep just enough tabs to know about Intel's releases. I update my PC Part Picker once a year and then last week my i7 930 started dying. It's no longer hitting previous stable OC settings after 7 years and even after clocking it down 200 mhz I have extreme instability issues. I figure after 7 years its time to pull the trigger on a new system so in total YOLO fashion I bought an i7 7700k. Luckily I had read up on motherboards (I rarely do this) during the Z270 launch and so I was prepared with a motherboard too.

About 2 hours later I stumble onto Anandtech just to check up on news and I read about Ryzen. Now in retrospect I'm a bit saddened I didn't do more research first. I have a special place for AMD in my heart given my first ever build was an Athlon, and I had a Venice and Opteron system in college, both of which ended up being my Bitcoin rigs (I still have them sitting in a closet somewhere collecting dust). I'm impressed and very happy that AMD has pulled this launch off and I love the feeling when a company is back in the game.

Given I don't upgrade my hardware all that quickly anymore I guess I'll just have to see how the competition is a few years later.

 

[rvborgh]

here it is for my quad Opteron 61xx (K10 cores). i set all the K10 cores to run at 3.3 GHz.

curious how the old school K10 compares with Ryzen...

dnetc v2.9112-521-CTR-16020314 for Win64 (WindowsNT 6.1).

[Mar 22 07:57:45 UTC] Automatic processor type detection found
an AMD Opteron 6xxx processor.
[Mar 22 07:57:45 UTC] OGR-NG: using core #0 (FLEGE-64 2.0).
[Mar 22 07:58:05 UTC] OGR-NG: Benchmark for core #0 (FLEGE-64 2.0)
0.00:00:16.72 [42,522,710 nodes/sec]
[Mar 22 07:58:05 UTC] OGR-NG: using core #1 (cj-asm-generic).
[Mar 22 07:58:24 UTC] OGR-NG: Benchmark for core #1 (cj-asm-generic)
0.00:00:16.45 [48,876,314 nodes/sec]
[Mar 22 07:58:24 UTC] OGR-NG: using core #2 (cj-asm-sse2).
[Mar 22 07:58:43 UTC] OGR-NG: Benchmark for core #2 (cj-asm-sse2)
0.00:00:16.17 [52,250,115 nodes/sec]
[Mar 22 07:58:43 UTC] OGR-NG: using core #3 (cj-asm-sse2-lzcnt).
[Mar 22 07:59:02 UTC] OGR-NG: Benchmark for core #3 (cj-asm-sse2-lzcnt)
0.00:00:16.30 [57,883,562 nodes/sec]
[Mar 22 07:59:02 UTC] OGR-NG benchmark summary :
Default core : #3 (cj-asm-sse2-lzcnt) 57,883,562 nodes/sec
Fastest core : #3 (cj-asm-sse2-lzcnt) 57,883,562 nodes/sec
[Mar 22 07:59:02 UTC] RC5-72: using core #0 (SNJL 3-pipe).
[Mar 22 07:59:21 UTC] RC5-72: Benchmark for core #0 (SNJL 3-pipe)
0.00:00:16.08 [12,421,499 keys/sec]
[Mar 22 07:59:21 UTC] RC5-72: using core #1 (KBE-64 3-pipe).
[Mar 22 07:59:39 UTC] RC5-72: Benchmark for core #1 (KBE-64 3-pipe)
0.00:00:16.19 [13,246,388 keys/sec]
[Mar 22 07:59:39 UTC] RC5-72: using core #2 (GO 2-pipe c).
[Mar 22 07:59:58 UTC] RC5-72: Benchmark for core #2 (GO 2-pipe c)
0.00:00:16.16 [15,341,312 keys/sec]
[Mar 22 07:59:58 UTC] RC5-72: using core #3 (GO 2-pipe d).
[Mar 22 08:00:16 UTC] RC5-72: Benchmark for core #3 (GO 2-pipe d)
0.00:00:16.17 [16,327,013 keys/sec]
[Mar 22 08:00:16 UTC] RC5-72 benchmark summary :
Default core : #3 (GO 2-pipe d) 16,327,013 keys/sec
Fastest core : #3 (GO 2-pipe d) 16,327,013 keys/sec
[Mar 22 08:00:16 UTC] Compare and share your rates in the speeds database at
http://www.distributed.net/speed/
(benchmark rates are for a single processor core)

 

[Bouowmx]

Tested distributed.net on same Intel Core i5-7200U, this time with 1-channel memory: same results.

While AMD Ryzen tanks RC5-72 AVX2, a BOINC user would likely not use a CPU for a GPU project; an AMD Redwood (Radeon HD 5670/FirePro V4800) 930 MHz achieves higher rate (~520 MKey/s) with less power.

 

[Puffnstuff]

Wow I'm totally out of date with CPUs but I keep just enough tabs to know about Intel's releases. I update my PC Part Picker once a year and then last week my i7 930 started dying. It's no longer hitting previous stable OC settings after 7 years and even after clocking it down 200 mhz I have extreme instability issues. I figure after 7 years its time to pull the trigger on a new system so in total YOLO fashion I bought an i7 7700k. Luckily I had read up on motherboards (I rarely do this) during the Z270 launch and so I was prepared with a motherboard too.

About 2 hours later I stumble onto Anandtech just to check up on news and I read about Ryzen. Now in retrospect I'm a bit saddened I didn't do more research first. I have a special place for AMD in my heart given my first ever build was an Athlon, and I had a Venice and Opteron system in college, both of which ended up being my Bitcoin rigs (I still have them sitting in a closet somewhere collecting dust). I'm impressed and very happy that AMD has pulled this launch off and I love the feeling when a company is back in the game.

Given I don't upgrade my hardware all that quickly anymore I guess I'll just have to see how the competition is a few years later.

You didn't make a bad choice in buying the 7700k as currently Ryzen is cuttings its teeth in the real world. Later on it will become more stable and supported which will make moving to that platform painless for people who are looking to replace their systems. I'm disappointed in that AMD kept touting superior performance from this new cpu and for the most part it didn't materialize. What it was for them was a huge performance leap over previous offerings and that's a good thing.

I watched a video this morning from Jayztwocents and he tested two systems, 1800x vs a 5960x both at equal clocks then shows the difference when the 5960x is oc'd and the Ryzen was several minutes behind on rending but at half the price. Depending on what is important to you this might be important but most companies operating under TVM principles would opt for an Intel system. Perhaps if AMD could bring us a 16 core system for around the same price as Intel's 10 core system we'd really have something and they could take the performance crown for a little while.

 

[zinfamous]

You didn't make a bad choice in buying the 7700k as currently Ryzen is cuttings its teeth in the real world. Later on it will become more stable and supported which will make moving to that platform painless for people who are looking to replace their systems. I'm disappointed in that AMD kept touting superior performance from this new cpu and for the most part it didn't materialize. What it was for them was a huge performance leap over previous offerings and that's a good thing.

I watched a video this morning from Jayztwocents and he tested two systems, 1800x vs a 5960x both at equal clocks then shows the difference when the 5960x is oc'd and the Ryzen was several minutes behind on rending but at half the price. Depending on what is important to you this might be important but most companies operating under TVM principles would opt for an Intel system. Perhaps if AMD could bring us a 16 core system for around the same price as Intel's 10 core system we'd really have something and they could take the performance crown for a little while.

You're extrapolating a large company decision based on what happens with an OC'd processor compared to a processor at default clocks. This is a fundamental error on your part.

OC performance is irrelevant, here. Companies aren't going to buy boxes of these things and individually OC them on every system (let alone a small production studio). That is ridiculous.

 

[IEC]

You didn't make a bad choice in buying the 7700k as currently Ryzen is cuttings its teeth in the real world. Later on it will become more stable and supported which will make moving to that platform painless for people who are looking to replace their systems. I'm disappointed in that AMD kept touting superior performance from this new cpu and for the most part it didn't materialize. What it was for them was a huge performance leap over previous offerings and that's a good thing.

I watched a video this morning from Jayztwocents and he tested two systems, 1800x vs a 5960x both at equal clocks then shows the difference when the 5960x is oc'd and the Ryzen was several minutes behind on rendering but at half the price. Depending on what is important to you this might be important but most companies operating under TVM principles would opt for an Intel system. Perhaps if AMD could bring us a 16 core system for around the same price as Intel's 10 core system we'd really have something and they could take the performance crown for a little while.

2m30s behind on a render that took ~23m on the OC'd 5960X. For individual usage, OC is a fair comparison.

However, 99%+ of companies don't OC their rigs at all, much less individually so a fairer comparison would be stock chips, which I'd suspect would give Ryzen the edge with its 600MHz higher base clock vs the 5960X.

It's likely that AMD's X399 HEDT platform will have 16c/32t chips that will make Ryzen look wimpy in properly multithreaded tasks. I'd expect some 16c/32t chips around 150/180W TDPs at the $1K mark which would be a very attractive value proposition for content creation.

 

[french toast]

X399 threadripper @999-1200$ would compete well against SKY X, hope they can get the clocks up, dont care if it hits 200w TBH.

 

[Bouowmx]

However, 99%+ of companies don't OC their rigs at all, much less individually so a fairer comparison would be stock chips, which I'd suspect would give Ryzen the edge with its 600MHz higher base clock vs the 5960X.

AMD and Intel prominently list "base" and maximum frequency, but never the frequency all cores can operate: gets me slightly annoyed >>
AMD Ryzen 7 1800X all-core Boost frequency: 3.7 GHz
Intel Core i7-5960X: 3.3 GHz

 

[amenx]

I made an impulse buy of a 7700k recently too and feel I should have gone Ryzen. The gaming benches where Ryzen is behind the 7700k are very misleading. They only apply to GTX 1080 or better (Ti). If you have a GTX 1070 or lesser, the benches are much closer @ 1080p. If a GTX 1060 where the GPU is more a bottleneck, Ryzen/Intel are virtually equal. At 1440p it would be a total non-issue.

Some benches demonstrating this (incl simulated benches for Ryzen 1600x, 1500x):

https://www.youtube.com/watch?v=O5cqOtWz5sU

 

[Despoiler]

I'm disappointed in that AMD kept touting superior performance from this new cpu and for the most part it didn't materialize. What it was for them was a huge performance leap over previous offerings and that's a good thing.

I watched a video this morning from Jayztwocents and he tested two systems, 1800x vs a 5960x both at equal clocks then shows the difference when the 5960x is oc'd and the Ryzen was several minutes behind on rending but at half the price. Depending on what is important to you this might be important but most companies operating under TVM principles would opt for an Intel system. Perhaps if AMD could bring us a 16 core system for around the same price as Intel's 10 core system we'd really have something and they could take the performance crown for a little while.

AMD has superior performance in just about all productivity workloads even besting Intel's 10 core in some tests. It seems you find isolated cases and then use them to make broad generalizations that are just not true. As far as this video, it can't be used to conclude anything. The systems aren't as identical as they could be. Most importantly, most professionals do not overclock their systems. They run them at stock for stability. That wasn't tested here. Not only that, Premiere has zero optimization for Ryzen currently.

This is an actual review using Premiere and there are links to other Adobe product tests.
https://www.pugetsystems.com/labs/a...2017-AMD-Ryzen-7-1700X-1800X-Performance-909/

TVM? Do you even know what you are talking about when you put subjects into a sentence or do you just throw out acronyms to attempt to appear learned? Lets see here, Ryzen is 1/2 the price of the Intel CPU. Top flight Ryzen motherboards are ~$100 less than comparable Intel. Ram is more expensive on the Intel platform. Ryzen has a 95w TDP vs 140w for Intel. Somehow in your mind Intel clearly wins? You seem to just be using T. Really?

 

[Insert_Nickname]

However, 99%+ of companies don't OC their rigs at all, much less individually so a fairer comparison would be stock chips, which I'd suspect would give Ryzen the edge with its 600MHz higher base clock vs the 5960X.

I can almost guarantee it is closer to 99.9999% of companies. You do not overclock a production machine. Period. No if's, but's and maybe's. You just don't.

 

[Bouowmx]

Top flight Ryzen motherboards are ~$100 less than comparable Intel.

Depends on comparable: what is comparable to ex. GIGABYTE GA-X99P-SLI? Comparable by RAM and PCI Express slots or by USB connectivity and Wi-Fi?

Ram is more expensive on the Intel platform.

I am not sure how. With same total capacity (ex. 32 GB), a kit with 2 modules and another with 4 are not radically different in price. Yet still, Intel HEDT can still operate with 2 modules.

Ryzen has a 95w TDP vs 140w for Intel.

45-W difference in TDP does not necessarily manifest to a 45-W difference in power; same as comparing to Intel Core i5-7600K, i7-7700K: 4-W difference in TDP is not necessarily a 4-W difference in power.
I remember in old Tom's Hardware reviews, author put "energy consumed", which determines what people pay, per task. There may be cases in which a higher-power processor completes task sooner such that energy consumed is lower.

Still, arguing with difference in pricing of AMD Ryzen 7 1700 and 1800X, and Intel Core i7-5960X and 6900K is hard.