Can Ryzen walk the workstation walk?

So, is Ryzen gg for workstations?

Not only that, but it can do it at a previously unimaginable price. Consider that you can get an R7 1700 and a B350 based motherboard like the ASUS Prime B350 for around $430. If overclocked, it can get dangerously close to the performance of a stock i7 6900k, which [combined with a mobo] is in the $1300-1500 range.

That’s THREE to almost FOUR TIMES cheaper. Even a top of the line 1800x + X370 combo is half the price of the Intel equivalent.

When compared to Xeon based systems the price/performance ratio shifts even more to AMD’s side, thanks to Intel’s ridiculous pricing. Besides, you basically need two Xeons to get to a good level of rendering performance, and the motherboards are all expensive as hell. Now explaining all the possible Xeon combos to compare is well… impossible. There’s loads of them, but rest assured none are budget champions.

Also, my personal experience with server motherboards has literally been a pain in the ass. They often miss modern features [like audio, for example] and BIOS updates are as common as a dodo bird. [Which is great fun when you build a system in the age of windows 8 and then win10 comes out… Hello BSODs, random boot errors and most convoluted BIOS update methods ever.]

That being said, Xeons with ECC RAM are practically unbreakable. I’ve never had a system crash ever. Once it booted, that is. It’s like a PC version of a VW Golf II. [It won’t start every time, but it’ll outlive you in the end.]

So if I had to choose between an i7 or Ryzen, I’d choose Ryzen every time. The difference in price can get you a LVL-UP on most other parts in your build.

With the Xeons it’s not such an easy decision. It’s a trade-off between price, performance and stability. Xeon based systems are more stable than regular desktop PCs. How much more stable? I wish I had a quantifiable answer to that. I can only speak from experience and say it certainly feels like it. However, let us not forget that Ryzen does support ECC RAM in theory. Perhaps some dedicated workstation boards will be coming our way in the future. For now, I’d say that if you do a lot of rendering and you have lots moar moneyz than what you’d pay for a Ryzen system, get some beefy Xeons. If you’re mostly into modeling or texturing, a desktop CPU will usually perform better in such tasks anyway due to higher clock speeds, so get yourself a Ryzen.

Then again, if you are in the GPU camp, the AM4 ecosystem, as it is now, is definitely not for you – SLI and CrossFire support is limited to 2 GPUs and that only on X370 chipsets. That means you won’t be able to fit your super power-efficient and affordable quad titan x tesla grid something with puny little memoriez inside. [now now, don’t get all fired up, it’s not that I hate GPU rendering per se, I just find it severely over-hyped but YMMV]

If you need more that 64Gb of RAM, you should also go look elsewhere. [And talk to your bank first, I suppose.]

All in all, I hope I helped you get some semblance of a grasp on how these 16 threads of goodness perform for stuff we usually need them to do. Yes, the whole ecosystem is still somewhat rough around the edges, but all that will be ironed out soon enough.

XFR shenanigans

Just in case I haven’t bored you to death already, here’s some additional info on XFR and Turbo speeds on Ryzen you should know about before you click ‘buy now.’

In case you are not a hardware guy person, you might easily misunderstand some things about how turbos actually work. The 1800X was mostly advertised as having a turbo speed of 4.0 GHz going up to 4.1 with proper cooling. And that’s true. However, what is also true is that the 4.0/4.1 GHz speed only works for up to two cores. And that only when the other cores aren’t pushed to their limits. In case of a full 16 thread load, the turbo caps on 3.7 on all cores. I just wish that was clearly stated while the hype train was still driving at full speed, but from what I’ve seen on all the AMD presentations, it was mostly presented in the form of my first sentence above. Hell, if you look at the stores, it’s not stated anywhere.

To be fair, Intel’s Turbo boost and Turbo Boost Max work in more or less the same way. But usually Intel advertises their CPUs with base clocks only, so there’s less chance of you having unrealistic expectations.

You’ll never see the 3.7 speed mentioned on product pages in any store but realistically, that’s the speed you will most likely be seeing. The XFR 4.1 Ghz is something I’ve rarely be able to see on my build despite having pretty low temperatures with the Kraken X52.

Definitely not a deal breaker or anything, but something that feels less disappointing if you know in advance, and I’ll use this chance to mention I’m not a fan of such practice.

And there you have it. Next up, as soon as I catch some time I’ll do some render tests with actual production scenes in both V-ray and Corona, just to get a fully realistic speed comparison.