Review: Ryzen 7 1700 processor

A lot of things look wrong with CPU utilization in this video. Specifically Ryzen seems to be almost stuck in 1 core usage.

 

So, for Arma 3 you basically need an i7, since i5 stutters and Ryzen has really low fps.

 

There was stutter or two on i5 but overall gaming performance right up there with an i7, smooth.

If he was testing 6600k or 7600k performance gap would be even closer.

 

Not the first game I've seen where Ryzen has half of it's threads doing pretty much nothing.

 

Undying, is this release date real and confirmed?

 

That 1600 could be a winner, $219 for a 6c12t CPU? A solid middle ground for people who don't outright need an 8c16t processor but don't want to invest in another quad.

 

For me the 1600 and 1600X seem both winners. I belong to the group that doesn´t need an 8 core cpu and doesn´t want another quad... The problem is that i´m itching to build my new rig and i don´t want to wait anymore...

 

I can wait the month OK, something to look forward to

 

Same here.
But...

Between the troubled 6/12 (ie. 4/8 + 2/4) asking all kinds of attention, and the screeching fast 4/8 just ready to go, the later is more likely the better option.

 

Why would it be 4/8 + 2/4 and not 3/6 + 3/6?

 

This was actually posted on guru3d last night but has seen been removed for some reason, was on the facebook page as well

 

Sorry if this has been addressed already but, what are those two SATA3 ports on the ASUS Crosshair VI Hero motherboard?

Thanks!

 

I'm a little confused by your question.

The ASUS Crosshair VI Hero has 8 Sata port on the back of the motherboard, which two are you talking about?

 

That game isn't even using the processor correctly, so i'm not sure what you're trying to get at?

I mean seriously you got the i5 and the i7 which have a fairly even load on all cores (except for the hyperthreads), then you got the Ryzen, which has the first core lit up, and then all of the other cores are at 0%, 5%, and some between 10-20%

It doesn't look like this game really likes to use two threads per core either, but it seems like it simply does not use them on the AMD platform

as an example:

https://youtu.be/6yh9doMXt3I?t=90

at this exact time, and it's basically the same throughout the entire video

i5
Core 1: 46%
Core 2: 38%
Core 3: 52%
Core 4: 60%
Total: 49%

i7
Core 1: 60%
Core 2: 11%
Core 3: 44%
Core 4: 9%
Core 5: 35%
Core 6: 10%
Core 7: 32%
Core 8: 15%
Total: 27%

Ryzen 7
Core 1: 91%
Core 2: 3%
Core 3: 16%
Core 4: 0%
Core 5: 19%
Core 6: 0%
Core 7: 14%
Core 8: 3%
Core 9: 11%
Core 10: 3%
Core 11: 22%
Core 12: 5%
Core 13: 23%
Core 14: 5%
Core 15: 17%
Core 16: 16%
Total 16%

It would be interesting to see what this game would put out with SMT on both Intel and AMD turned off, as this game clearly does not like SMT, and pretty clearly evens out the workload on the non SMT cores on Intels system, yet not on AMDs, so it'd be interesting to see, without SMT, what the total usage of the CPU is in this game, as obviously if the game is not using a CPU to its fullest extent, it will not perform the same, or better, but worse.

The reason i bring this up is yes, for this game, the ryzen does not perform as well. That's a fact. But you're also trying to base findings off of gimping the Ryzen platform by using something that will not use (unless the game gets updated or windows gets updated or bios gets updated, whatever be the case) the full potential of the CPU, and saying it's not a good CPU because of it, whereas there are programs that know how to use the ryzen CPU to its full potential and they destroy both of those CPUs in that test.

It's the same arguement that a single-threaded limited program will be faster on a high Mhz low-threaded CPU then a lower Mhz high-threaded CPU. It's not exactly a fair comparison, as you are only gimping the CPU to prove the competitor is better. Now, if your life runs on programs that are low threaded, sure that's important, i get that, but that doesn't mean that CPU is "better", it means that CPU is better for your specific needs. Just like how i don't need a car that goes 250 mph in a straight away, but can't navigate roads because it's designed to go straight, not around corners and etc. Some people might need it, but not me.

 

I've been playing Arma3 since it was in the alpha stage and from the beginning endless complaining was going on about the way it's using the CPU (or rather, isn't using). It has become a bit better along the way, but it was a rocky road for sure. It's of course bad news for AMD. Who knows if and when they might bother to fix it for Ryzen. AMD should try to coax them directly. Hopefully they do.

 

ARMA 3 is really bad coded. Online is even worse because the FPS pretty much depends of the server and mods, usually I barely get more than 50FPS on towns, the CPU usage is less than 40% and also the GPU downclocks due the low usage.
The Physhics are calculated server side and at the best is 30FPS.

 

I'm quoting the article:

 

New BF1 DLC

http://www.pcgameshardware.de/Battl...eld-1-They-Shall-not-Pass-Benchmarks-1223170/

 

Not sure what you're trying to show us. Ryzen loses to 4ghz 6900 and beats a 2ghz 6900 and some 6 year old intels. Round of applause?

The chart is missing gaming mainstreem (gaming) i5/i7 cpus. It's a bit of a joke, seeing as this is a gaming bench chart.

It seems that in BF1, Ryzen is just short of Broadwell-E IPC. Clocked the same, Broadwell-E just beats Ryzen.

The chart is clearly made by amateurs anyway, one minute the 6950x has 10 cores, the next minute it has 8 cores, which one is it?

 

The charts are not done yet.. And if you think that a 329 1700 beating the 1.7 k 6950 x when OC not a good result then you need to get your head examined immediately and if you thought that zen will have IPC of kabylake and not just below broadwell E then examine you head for the second time.. Read the damn article before spewing nonsense as you usually fo eclap