Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

Page 62 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

arandomguy

Senior member
Sep 3, 2013
556
183
116
Could you point me in the direction of suitable reviews please?

Not singling you out, but it'd be nice to have an agreed list of programs (between all on the forum) before the Ryzen reviews come out - just to avoid anyone accusing anyone else of bias etc etc.

Are you saying what would be suitable tests or examples of existing suitable reviews?

If it's the latter that is my gripe in that on a whole its very rare due to how reviewers review. Granted some of this is understandable, they want to fast and repeatable tests, but those end up being built short run built in benchmarks or scripted SP sequences that tend to that tend to be low impact on the CPU and things like memory.

And because there isn't as much investigation into this issue we don't fully know how many situations result in the CPU performance mattering either. I don't mind being pointed out as I already try to bring this problem up as much as possible to bring light to the issue and hopefully have more data come out.

If we want to look at relevant results to gamers as examples the two biggest new MP games by far are Overwatch and BF1. BF1 should be tested in full 64 player conquest. No site should be testing it in SP. Overwatch has competitive gamers looking to drive as close to 300 fps as possible on lower graphics settings while still at 1080p or 1440p, this also again should only be tested in full 6v6 MP and not running around an empty level. Right here in these two cases we also have a possible fast 4c/8t vs more slower cores/threads comparison as well.

Other popular games? Fallout 4 is mega popular and we know it struggles in certain locations like Corvega and Diamond City, the latter being a pretty important hub area so you are going to visit it. How about those Watch Dog 2 numbers that came out from sites like Gamegpu? Would be interesting to see how sites also investigate.

But maybe we should at least stop reusing the same tests that either show no difference in CPUs or all run well into the hundreds for non competitive SP games?

Really this issue also applies to GPUs. Personally I now tend to look at specific tests scenarios that are relevant to me as oppose to relying on any single reviews sites overall results. I don't nor have I been interested in playing AoTS and certainly wouldn't ever play or care about the AoTS benchmark like most people, so why would I care about results at reviews that use that?

In terms of Ryzen I'm rather interested in it at the moment but I do have some questions related to its comparative 4c/8th workload performance and memory due to relevant games for me like Overwatch and Fallout 4 that I will be researching on before I make a final decision. But I have a feeling there is might be lacking data in most test reviews.
 

unseenmorbidity

Golden Member
Nov 27, 2016
1,395
967
96
If you can get 4ghz on air with a 1700...why in the world by an 1800X? I'll wait for something more reliable than WCCFFAKE lol....If so though, that is really good. I just don't understand the marketing strategy.

Because most people don't overclock. People are really looking down on the 1700, but I think they are in for a surprise.
 
  • Like
Reactions: french toast

AtenRa

Lifer
Feb 2, 2009
14,001
3,357
136
16 lanes from CPU go directly to the board on Z270 platform, the only x4 lanes bottlenecked stuff is everything on the chipset. So, all the block devices, audio, NICs and I/O. In comparison, on Ryzen it means you can have some Gen 1 ports and 1 NVMe x4 drive not bottlenecked by the chipset. OTOH on Z270 you can have 2 NVMe drives that may not bottleneck each other unless they are used together, but on Ryzen one of them has to be on 2.0, cutting the speed in half or more.

Not with X300
 
  • Like
Reactions: Drazick

wahdangun

Golden Member
Feb 3, 2011
1,007
148
106
Why didn't you read the rest of my post. You could easily find I was being facetious because the next line says that the results show only 5-7% gains at 1080p.



Useful in an academic sense, but no one with those CPUs games at 720p so completely pointless in the real world.

Who cares if those conditions exist when no one in the real world is going to run into them.

Also, the [H] article I linked also tested at 1440 and found a truly bizarre result where more cores actually resulted in slightly lower performance despite being an advantage at lower resolutions. They said no one at AMD, NV, or Intel could really explain why either. My guess is drivers or some buggy game code, but even if the results weren't that way, you'd see almost no performance gain at the higher resolutions where most people game.



Yeah, they had to go all the way down to 720p with a Titan X to achieve that. Let me go make a poll in the VC&G section to ask how many Titan owners game at 720p.

Once again, useful in academic sense, but of no practical real world value right now. I suppose 5% gains at 1080p are 5% gains, but you have to understand that the percent price difference for AMDs 8C/16T CPU is going to be almost as bad as Intel's once Ryzen R3 processors are on the market, especially if the price leaks continue to be correct or close enough.

its stupid to test CPU with GPU bottleneck scenario because lets be real, GPU advancement now is magnitude higher than CPU, and every new generation of GPU will shift the bottleneck more and more to the CPU and it worsened by the fact that every Next gen PC games engine will increase drawcall because it will more complex.
 
  • Like
Reactions: Rifter and Karnak

WhoBeDaPlaya

Diamond Member
Sep 15, 2000
7,414
401
126
If you can get 4ghz on air with a 1700...why in the world by an 1800X? I'll wait for something more reliable than WCCFFAKE lol....If so though, that is really good. I just don't understand the marketing strategy.
^ This. If the 8c/16t parts max out at ~4.2GHz or so anyway, the $170 saved on the R7-1700 would be like getting a free mid-range motherboard.
You also get a free HSF with it, for whatever that's worth.

Learned my lesson from the PII-400 days. It barely OCed, and was matched by a MUCH cheaper PII-333. Let's not even talked about the dick punch called the Celeron 300A that came out just a few months after and OCed to 450MHz past both of them :(
 
Last edited:

inf64

Diamond Member
Mar 11, 2011
3,698
4,018
136
Ryzen 1800X is in Passmark database :

www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+7+1800X&id=2966
:)

CPU score : 15945

Single Thread Rating: 2129
Samples: 1*

Credit to http://www.mykancolle.com/?post=1362

Overall score has to be affected by Prime and Physics because otherwise it is crushing 5960x and still is ranked lower than it. Exceptional performance in Passmark.

http://imgur.com/a/H8ywc

O5t3y9g.jpg


BRfoGTi.jpg


5OOwY3s.jpg


yehm5lv.jpg

WT7qTp6.jpg

Z8hHhcB.jpg


CTvaXjO.jpg


xDTlZOr.jpg


PlUvz7n.jpg

7myJTQK.jpg


It was assumed that 1800X has XFR enabled in ^^ results above and that it boosted to 4.1GHZ (ST) and 3.7GHZ(MT). This is not confirmed.
 
  • Like
Reactions: Drazick

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
That would seem to tie in with other indications that XFR is only giving an extra 100mhz...?
 

CatMerc

Golden Member
Jul 16, 2016
1,114
1,149
136
Damn, guess that poor physics and prime number performance wasn't a one off thing from a poor platform.
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
14,330
4,918
136
Damn, guess that poor physics and prime number performance wasn't a one off thing from a poor platform.

The data source is the same as the original Chinese forum "I don't know if this is real, but here's something I got second-hand" post. Marvin just made graphs out of the data IIRC.
 

Teizo

Golden Member
Oct 28, 2010
1,271
31
91
It was assumed that 1800X has XFR enabled in ^^ results above and that it boosted to 4.1GHZ (ST) and 3.7GHZ(MT). This is not confirmed.

So the clocks go down as the number of cores used goes up?
 

inf64

Diamond Member
Mar 11, 2011
3,698
4,018
136
So the clocks go down as the number of cores used goes up?
Base clock is 3.6Ghz and ST turbo is 4Ghz. These are the specs without XFR. XFR adds 100Mhz on each of these. We don't know if the results were with XFR, that was the assumption of the creator of the graphs.

Note that 1800X used 2133Mhz if Passmark's system info is to be trusted (CL15). It had lower latency than any previous Ryzen passmark entry (67ns Vs 75ns on 2400Mhz from previous Ryzen passmark leak).
 
  • Like
Reactions: Drazick