AMD Ryzen DDR4 Memory Scaling Tests On Linux

[darkbasic]

the competition has quad memory channels, so AMD has no way to compete there anyway.

No, it doesn't... Intel has quad channel only on their high end socket, which AMD will have as well with their 16 cores processors lineup.

[efikkan]

No, it doesn't... Intel has quad channel only on their high end socket, which AMD will have as well with their 16 cores processors lineup.

i7-6800K cost less than Ryzen 7 1800X, and offers double memory bandwidth.

[leipero]

As i tought in the very beggining, it's all about the code, recent tests with Ashes of Singularity proves that.

efikkan
That simply isn't true, does i7-6800k have better memory BW than i7-6950x? If so, then your statement could be true. Given same frequency, 6950x will have better memory BW, but not even close to double. From G3D (read MB/s):
i7-6950x DDR 3200MHz = 62152
R7 1800x DDR 3600MHz = 56189
R7 1800x DDR 3000MHz = 46922

So, even with lower frequency, Ryzen does not have 2x less mem.BW. You can find article and read it for the rest.

[efikkan]

As i tought in the very beggining, it's all about the code, recent tests with Ashes of Singularity proves that.

efikkan
That simply isn't true, does i7-6800k have better memory BW than i7-6950x? If so, then your statement could be true. Given same frequency, 6950x will have better memory BW, but not even close to double. From G3D (read MB/s):
i7-6950x DDR 3200MHz = 62152
R7 1800x DDR 3600MHz = 56189
R7 1800x DDR 3000MHz = 46922

So, even with lower frequency, Ryzen does not have 2x less mem.BW. You can find article and read it for the rest.

i7-6800K is the same socket 2011-3 as i7-6950X and has the same memory bandwidth.
It will have higher bandwidth at 2133 MHz than R7 1800X has at 3600 MHz.

[leipero]

i7-6800K is the same socket 2011-3 as i7-6950X and has the same memory bandwidth.
It will have higher bandwidth at 2133 MHz than R7 1800X has at 3600 MHz.

Equal, but it is tripple channel, so comparison is not as good tho, there are otehr factors, it is 6 core 12 thread CPU, 1800x have 8/16, and 1700x is much cheaper with little performance degradation, so... , it's about product design and purpose .

[efikkan]

Equal, but it is tripple channel, so comparison is not as good tho, there are otehr factors, it is 6 core 12 thread CPU, 1800x have 8/16, and 1700x is much cheaper with little performance degradation, so... , it's about product design and purpose .

What? The x99 has four memory channels, it even clearly says so in the documentation from Intel. As an owner of the previous x79 platform, I can assure you it has four channels. i7-6800K has double memory bandwidth over Ryzen 7 and i5-7600K/i7-7700K/etc., but that doesn't mean it will get twice the score in every memory intensive benchmark.

i7-6800K will allow you to have more memory bandwidth cheaper, greater performance in games, Photoshop, web browsing and most other tasks. Ryzen 7 1800X will only outperform it in specific workloads, so it's only the better choice if these apply to your

[bridgman]

i7-6800K will allow you to have more memory bandwidth cheaper, greater performance in games, Photoshop, web browsing and most other tasks. Ryzen 7 1800X will only outperform it in specific workloads, so it's only the better choice if these apply to your

Just curious, where are you seeing these benchmarks ? Most of the tests I have seen (even the early ones with slow memory) show Ryzen outperforming the 6800K in most of the scenarios you listed. The exceptions were in gaming, but AFAICS with faster memory that has changed too.

I haven't seen any direct 7700K vs Ryzen with faster memory comparisons yet, but looking across multiple benchmarks it seems that the games most affected by memory speed are also the ones where there was the most performance deficit initially.

My impression is that Ryzen/6900K need something between dual- and quad-channel (at 2133) to reach the point where more bandwidth doesn't make a big difference in performance. Dual channel fast DDR4 seems to be able to get up to roughly that level, which makes things really interesting.

[leipero]

What? The x99 has four memory channels, it even clearly says so in the documentation from Intel.

That makes it even worse, I tought it was tripple-channel sicne it goes in line with benchmarks, but yeah #of channels =/= same # times performance. I'm not going to argue about what would make better choice, for me, it is clear that Ryzen 1700/1800 are much faster CPU's than 6800k with 40W+ less power consumption, and that's only natural, since it is new generation CPU. As for games, they are not benchmark, even less CPU benchmarks..., with better code (games made for Ryzen CPU's), it's not even question of 8C/16T would outpreform 6C/12T at same IPC and similar frequencies if that power is required, if not, it is meaningless to even talk about it.

Considering the fact that web browsing experience is mainly connected with internet speed, Core2Duo can do that job just fine for me.

[efikkan]

That makes it even worse, I tought it was tripple-channel sicne it goes in line with benchmarks, but yeah #of channels =/= same # times performance.

That makes no sense what so ever.
Just look up AIDA64 and SiSoftware Sandra in benchmarks, and you'll see X99 can deliver roughly double memory bandwidth. With the recent DRAM price increases, quad memory channels will give you great memory bandwidth without buying very expensive memory. It's also great for workstations with ECC, since ECC memory don't support XMP.

I'm not going to argue about what would make better choice, for me, it is clear that Ryzen 1700/1800 are much faster CPU's than 6800k with 40W+ less power consumption, and that's only natural, since it is new generation CPU.

Ryzen 7 1800X is only faster in certain benchmarks, if these aren't your primary workloads, then i7-6800K will offer much better overall performance. You'll have to do a lot of rendering in Blender etc. to justify choosing Ryzen 7 1800X over i7-6800K.

As for games, they are not benchmark, even less CPU benchmarks..., with better code (games made for Ryzen CPU's), it's not even question of 8C/16T would outpreform 6C/12T at same IPC and similar frequencies if that power is required, if not, it is meaningless to even talk about it.

What?
The only thing that matters is real world performance, and that's where Intel excel.
Ryzen is a more superscalar CPU than Intel's architecture, but lacks a decent prefetcher, which makes it suffer in a number of workloads including gaming. Having more computational power doesn't matter if the front-end of the CPU can't feed it properly. That's why you see all the Intel CPUs are fast enough for gaming with turbo at ~4 GHz, and Ryzen lagging behind. Just as with Bulldozer, it's not just a matter of waiting for the software to "optimize" for it.

[Luke_Wolf]

The only thing that matters is real world performance,

Absolutely Correct,

and that's where Intel excel.
Ryzen is a more superscalar CPU than Intel's architecture, but lacks a decent prefetcher, which makes it suffer in a number of workloads including gaming. Having more computational power doesn't matter if the front-end of the CPU can't feed it properly. That's why you see all the Intel CPUs are fast enough for gaming with turbo at ~4 GHz, and Ryzen lagging behind. Just as with Bulldozer, it's not just a matter of waiting for the software to "optimize" for it.

But not quite. In real world performance, Ryzen is more than Good Enough(TM) for gaming, just as those Intel parts are Good Enough(TM) for gaming. Does the Octocore Ryzen run behind the Quad Core Intel parts in terms of sheer FPS? oh absolutely in a lot of cases, but in real world performance does that mean anything at all? Nope. Because after 60FPS (and in some special cases up to 144FPS if you've happen to got a lot of money to throw at displays) framerate for real world performance loses all meaning and other utilization figures become much more useful, such as... How heavily loaded are the cores? If they're running near maxed out, you're going to get microstutters in terms of that Real World Performance(TM). Ryzen performance isn't bad in any sense in the real world, it's just in the current state of software, it's not as fast at shoving out frames compared to the Intel product.