Xeon 10-core v Xeon 12-core - advice please!

[alexyork]

Performance will be very close. I would go with more cores for less $$. Keep in mind that ECC DDR4 ram is not cheap..

You won't get any NUMA or 64 thread limit problems with 2670v3. I have 14 and 18 core xeons. 14 core work well, with 18 you need workaround because pair of those give you 72 threads. Hopefully Vlado and his team will fix it soon.

I'm pretty sold on this now. Just waiting to pull the trigger

[3LP]

Just wondering,
Would it not be better power for the $$ if you went with some i7-5960X?
I know this could imply a bit of licensing issues.

I'm also in the process of looking into expand my render power and I'm just wondering what the overcost would be between xeons and i7.

Stan

[alexyork]

Just wondering,
Would it not be better power for the $$ if you went with some i7-5960X?
I know this could imply a bit of licensing issues.

I'm also in the process of looking into expand my render power and I'm just wondering what the overcost would be between xeons and i7.

Stan

This is what I've been doing for years - going with the top end i7 machine and using older ones as rendernodes before they're retired. For me I now need maximum single-PC performance, so it makes more sense to go with a dual xeon system. The overhead of running a system over DBR means you lose quite a lot of that machine's native performance as well. It's done me fine for many years but I've just pulled the trigger on a very nice workstation and should have it set up and running on Wednesday hopefully.

I think if you're looking for inexpensive rendernodes in a smaller form factor then 5960X could be a good option. Much cheaper RAM and obviously only 1 processor. But as I said, the performance will not scale as well as a dual CPU machine.

[Joelaff]

Keep in mind that the 5960x will be significantly faster for pretty much everything except the actual act of rendering. Loading scenes, saving scenes, simulating, turbosmooth, displacement, mass fx, animating, viewport playback, all of pflow, etc. are all single threaded. These will all run much slower on the Xeon systems, especially when you overclock the 5960x, which can get a real, serious overclock. The Xeon overclock is almost non-existent. We just used e5-2697v3 (14 real cores), and were able to get a 5.4% overclock, which is a lot compared to what most people get (2-3% seems normal). The Xeon memory runs much slower as well (even with no overclocking on the 5960x).

As someone with both, honestly, I felt like it was kind of a waste as a workstation, and absurdly overpriced for render power, even if you figure the extra 300 or whatever in for the VRay license, and even with another 150 for deadline. We have a good sized farm so I just throw every test on the farm anyway, and continue working on other things. Everything but the render runs faster on the 5960 boxes, and it is no where near close to comparing core count * clock speed vs. core count * clock speed. In the end Megahertz matters so much more than Intel wants you to think. High core counts are really engineered for running multiple virtual machines with its own memory segment, not for all the cores trying to hit the same memory through a relatively slow bus.

[alexyork]

Keep in mind that the 5960x will be significantly faster for pretty much everything except the actual act of rendering. Loading scenes, saving scenes, simulating, turbosmooth, displacement, mass fx, animating, viewport playback, all of pflow, etc. are all single threaded. These will all run much slower on the Xeon systems, especially when you overclock the 5960x, which can get a real, serious overclock. The Xeon overclock is almost non-existent. We just used e5-2697v3 (14 real cores), and were able to get a 5.4% overclock, which is a lot compared to what most people get (2-3% seems normal). The Xeon memory runs much slower as well (even with no overclocking on the 5960x).

As someone with both, honestly, I felt like it was kind of a waste as a workstation, and absurdly overpriced for render power, even if you figure the extra 300 or whatever in for the VRay license, and even with another 150 for deadline. We have a good sized farm so I just throw every test on the farm anyway, and continue working on other things. Everything but the render runs faster on the 5960 boxes, and it is no where near close to comparing core count * clock speed vs. core count * clock speed. In the end Megahertz matters so much more than Intel wants you to think. High core counts are really engineered for running multiple virtual machines with its own memory segment, not for all the cores trying to hit the same memory through a relatively slow bus.

I think you hit the nail on the head when you mentioned that the 5960X is probably a better option when you've got a farm behind you to throw your test renders and final renders on. In that instance it would be best to have better single-core performance and save the cash. But if, like me, you don't have a farm behind you, just a few older machines that you're using in DBR, then you really will benefit enormously from having better rendering performance from the dual xeon setup in a single workstation. Furthermore, these high-end xeons are actually not too far behind the top-end i7s (overclocked or not) for single-core performance in applications like max, since they're quite terribly optimised anyway. I would also contest that the CPU makes much of a difference when it comes to Nitrous performance, but perhaps you can cite some sources for this? In my experience that will all be down to the GPU. And I would also mention that the Xeon ECC RAM is really not slower than whatever is usable with i7s, and if it was it would be a negligible real-world render speed difference.

I would also add that Autodesk are apparently in the process of trying to multithread more of max, including the items you mentioned, so when that happens (hopefully sooner rather than later), there will be even more reason to use the dual xeon setup. Bang for buck is a delicate thing. There are other factors like power consumption, space, cooling (air con), maintenance of multiple machines etc. Again, it's a balance.

[Joelaff]

I think you hit the nail on the head when you mentioned that the 5960X is probably a better option when you've got a farm behind you to throw your test renders and final renders on. In that instance it would be best to have better single-core performance and save the cash. But if, like me, you don't have a farm behind you, just a few older machines that you're using in DBR, then you really will benefit enormously from having better rendering performance from the dual xeon setup in a single workstation. Furthermore, these high-end xeons are actually not too far behind the top-end i7s (overclocked or not) for single-core performance in applications like max, since they're quite terribly optimised anyway. I would also contest that the CPU makes much of a difference when it comes to Nitrous performance, but perhaps you can cite some sources for this? In my experience that will all be down to the GPU. And I would also mention that the Xeon ECC RAM is really not slower than whatever is usable with i7s, and if it was it would be a negligible real-world render speed difference.

I would also add that Autodesk are apparently in the process of trying to multithread more of max, including the items you mentioned, so when that happens (hopefully sooner rather than later), there will be even more reason to use the dual xeon setup. Bang for buck is a delicate thing. There are other factors like power consumption, space, cooling (air con), maintenance of multiple machines etc. Again, it's a balance.

Without a farm, sure, I would rather have a Xeon box. If ALL I did was 3d (no 2d comping, etc.) then I would also want the Xeon box. If I also did a lot of 2d comping (motion, mainly, Photoshop is so fast on any hardware now that it doesn't matter much) then I might go with two or three 5960x, which you could likely build for the same price as one dual Xeon (maybe even four vs. a pre-built xeon machine).

I think this machine will likely work out great for you. I wasn't trying to instill buyer's remorse...

With regard to viewport performance I was referring to playback, which involves animating things with the CPU, as well as the GPU. Complex rigs start to slow down, or many objects on paths, etc. In reference to RAM speed, the x99 motherboards let you use DDR4-3200, vs. the DDR4-2133 of the Xeon. This doesn't make a huge difference in the real world, unless you are doing operations that move a lot of data from main memory to the cpu and back (compositing would be doing this more than 3d rendering, since rendering managers to get a lot to stay in the cache, though render can certainly take advantage of the faster memory, but I would guess a lot of dynamic memory items like proxies would benefit some as well.). You also have to remember that 28 cores (or 24, or 36, whatever) all accessing the same memory does cause bottlenecks. More cores is not a linear thing. Dual processors are even less linear than multi-core processors (though on some loads this can be outweighed by having more cache). It is really hard to make a good dual processor motherboard.

I really hope Autodesk optimizes more of Max! There are plenty of things that simply can't be parallelized, though. I really wish Intel would work on clock speed instead of core count.

I am sure you will love your new box.

[alexyork]

Without a farm, sure, I would rather have a Xeon box. If ALL I did was 3d (no 2d comping, etc.) then I would also want the Xeon box. If I also did a lot of 2d comping (motion, mainly, Photoshop is so fast on any hardware now that it doesn't matter much) then I might go with two or three 5960x, which you could likely build for the same price as one dual Xeon (maybe even four vs. a pre-built xeon machine).

I think this machine will likely work out great for you. I wasn't trying to instill buyer's remorse...

With regard to viewport performance I was referring to playback, which involves animating things with the CPU, as well as the GPU. Complex rigs start to slow down, or many objects on paths, etc. In reference to RAM speed, the x99 motherboards let you use DDR4-3200, vs. the DDR4-2133 of the Xeon. This doesn't make a huge difference in the real world, unless you are doing operations that move a lot of data from main memory to the cpu and back (compositing would be doing this more than 3d rendering, since rendering managers to get a lot to stay in the cache, though render can certainly take advantage of the faster memory, but I would guess a lot of dynamic memory items like proxies would benefit some as well.). You also have to remember that 28 cores (or 24, or 36, whatever) all accessing the same memory does cause bottlenecks. More cores is not a linear thing. Dual processors are even less linear than multi-core processors (though on some loads this can be outweighed by having more cache). It is really hard to make a good dual processor motherboard.

I really hope Autodesk optimizes more of Max! There are plenty of things that simply can't be parallelized, though. I really wish Intel would work on clock speed instead of core count.

I am sure you will love your new box.

Let's hope so - it's been many years since there was any real innovation in max and that would certainly be a huge leap forward.

Yes, for me the vast majority of my time is spent doing test renders, and I almost never animate or do any kind of simulations, so this setup will work for me.

[alexyork]

Just an update to say thanks to everyone for the advice. My system arrived a week or so ago and is absolutely screamingly fast, as I'd hoped. I'm doing full-res production renders in the time I was doing low-res test renders before... quite something!

[Resolution01]

Alex

So what exact spec did you go for in the end?

[peterguthrie]

We have nothing but dual xeon machines here. Being able to do super quick test renders without resorting to DR speeds up my work sooooo much. Most of what I do seems to be sitting waiting for renders to finish so spending £6000 on a machine is well worth it.

Enjoy your new machine Alex!

[alexyork]

Cheers Peter. It's made a whopping difference to the amount of time I'm spending on test renders.

Resolution: I went for the following:



3XS GW-HTX30 Scan Base BuildASUS/Z10PE-D8/WS
Intel Xeon E5-2670 V3 s2011-3 x 2
Corsair H80 with SP120 Quiet x 2
32GB Cruc DDR4 Server (4x8GB) x 2
256GB SSUNG 850 PRO SSD
LG BH16NS40AUAU10B 16x BLU OEM
MS Win Pro 8.1 64Bit
FD-CA-DEF-XL-R2-BLack Case
3XS RM1000W GOLD CP-9020062-UKTotal was around £3800+VAT.I did spec up the slightly faster processors but the price difference was massive and they came with less cache. For the faster processors with the same amount of cache the price was enormously more. I think this config represented a really good bang-for-buck. Compared to my i7 3930k @ 4ghz this thing is stupidly faster. Having so many threads means that huge chunks of the image can be rendered even when other buckets get stuck on complex areas. Frees up so much time!Very happy with it so far. Service from Scan was very so-so, however, and there are issues with some dead USB ports I need to deal with. They seem totally disinterested in fixing them.

[Resolution01]

I have always been a dual Xeon user but I was thinking of going for the overclocked i7 route but I don't think it is going to be much faster then what I have now and that it is worth it to go the dual Xeon route again. Are you going to overclock them and get that extra 10/15%? How noise is the setup at full load?

It has been an interesting thread, thanks guys.

Renato

[alexyork]

I have always been a dual Xeon user but I was thinking of going for the overclocked i7 route but I don't think it is going to be much faster then what I have now and that it is worth it to go the dual Xeon route again. Are you going to overclock them and get that extra 10/15%? How noise is the setup at full load?

It has been an interesting thread, thanks guys.

Renato

These new xeons are not overclockable. The whole system is very, very quiet even under full load. Quieter than my 3930k@4ghz with an H60.

[Resolution01]

I read some where you can increase the BCLK and it gives you between 10/15% increase in performance. This is probably nonsense.

[alexyork]

I read some where you can increase the BCLK and it gives you between 10/15% increase in performance. This is probably nonsense.

My warranty would be void if I overclocked this machine anyway so I'm steering clear!