Hello.
I can’t find the answer so far:
Will Sketchup make use of 2 processors at onces (means on one motherboard)? I mean while buidling models, scene. Or is it useful only while rendering, like by V-Ray or Kyrkythea?
One more question is about Win 7, does it work with 2 CPUs and so any programme under this sistem?
Thanks in advance
SketchUp will only use a single core no matter how many are available.
Thanks a lot,
Would it be used while rendering (on Win 7 or Win 10)?
Many renderer use multiple cores.
While talking about SU modeling, should I understand, that if only one core then also only one CPU?
You could have multiple CPUs each with multiple cores. But SU will only use one core of one CPU for modelling.
What John wrote.
Is there any plan to switch to multi GPU/CPU/GPGU/distributed on network engine for Sketchup ? Not that we are in 2017 and multicore and hybrid GPU/CPU rendering exists for years…
I cannot answer about SketchUp. Intel released their first Pentium processors in 1993 together with the first multi-CPU motherboards. Immediately after that, developers promised an imminent release of multiprocessor-supporting versions of, for instance, AutoCad and 3D Studio. It hasn’t happened yet, almost 25 years later. Non-realtime rendering generally supports multicore today, but it seems that multicore modelling might be impossible.
Anssi
That’s true, and we can think that they deliberately cripple multiprocessor capabilities to make us buy more powerful hardware (AMD and Intel) or overprice re-branded “pro” hardware (Nvidia and ATI/AMD), whith Autodesk being the first cripple company I know. Game engines can render using several cores and GPUs, and they are showing 3D polygons, but Autocad can’t render thousands of vectors in 2D without a Xeon ?!
Please don’t raise a conspiracy theory without any facts!
If you read other topics on this forum about multi-threading as well as the general computer-science literature, you will find that making an interactive geometry editor use multiple cores is a very difficult problem that nobody has yet solved after years of trying. An editor is fundamentally different from a renderer in that its work can’t (so far as anyone has yet figured out) be divided into spatial areas and each area assigned to a separate CPU. It has always to cope with the consequences of the user editing something that otherwise might have been sent off for processing by a different thread/CPU.
I get your point… However, when a simple line in a config file for some Adobe products can make the renderer use your GPU whereas it was not supposed to be “compatible”, I know that something is fishy… Nevertheless you’re right, I don’t have proof for Autodesk and the manufacturers.
When the editor is in “read only” or view mode, doesn’t he act like a simple renderer ? Like when you scroll or pan the mouse, you just render a different view of the model, no modification can happen before you release the mouse, what is preventing the use of several threads in this case ?
I think you’re confusing two fundamentally different things: Modeling and Rendering.
Modeling (in this context) is the act of changing the geometry of the model. Rendering (in this context) is the act of taking a model and displaying a highly refined version.
To the best of this forum’s collective knowledge, nobody has been able to implement modeling that uses more than one core.
SketchUp is a modeling program. It does include some very basic rendering in order to display the model, but relies on 3rd party plug-ins and services for high detail rendering.
SketchUp already uses your GPU to manage the view of the model, including orbiting, panning, zooming and SU’s limited sort of rendering. Most modern GPUs are themselves multi-core and manage spreading the work over those cores automatically. I won’t claim to know that the SU code couldn’t use the GPU more effectively since I don’t have access to the source and I’m not an OpenGL expert, but rework of that aspect was a major part of SU 2017. Try turning off Hardware Acceleration in an older version of SketchUp (no longer an option in SU 2017) and check out the performance on any non-trivial model!
This whole multi-thread/multi-core thing comes up again so often that I considered writing an essay about it for posting here. But since nobody seems to search and read the previous posts, that might be a waste of effort (Ack! Am I negative today or what?)
Since multi-threading isn’t really supported, would overclocking the CPU be the only real way to increase the speed of modeling?
Would the GPU bottleneck the speed of modeling at all or does it really only pertain to rendering speeds?
I hope it’s not an issue quoting here guys from other forum;
"> Re: Dual Xeons v. Core i7
SkellObissis wrote:
Blaw789 wrote:
Hello all,
I have read several threads about new rendering machines and it seems the Core i7 is universally loved, but I was wondering if anyone has built a new dual Xeon machine? I am getting a new rendering workstation and am trying to decide between a Core i7 920 build and a dual Xeon E5520 (Neahalem). The Xeon machine is about 400 dollars more expensive but has a PassMark CPU benchmark almost twice as high as the Core i7. It seems like it might be worth it to go for the extra CPU. Also dont know how many more non-server CPUs Intel is making on the 1366 socket… a Core i7 machine may not be upgradable? Thanks for your input!
_I myself looked into this, but from what Ive read, the operating system would not be able to assign both CPUs to one task, you wouldn’t be able to utilize all 8 cores on a single render. you could run two renders with 4 cores each simultaneously… _
i may be wrong but thats what i recall. at least with windows.
-Tony
_This is not correct. The whole point of a dual socket board and Xeon chips is so they can communicate between one another. The OS will be fine. Maybe you were reading about trying to use (2) Core i7 920s. _
There is a reason Xeons have a price premium…they are higher binned silicon and they can run with multiple CPUs."
If Xeon stations are made so that they can communicate betweeen one another and so “the OS will be fine” - wouldn’t modeling/rendering/CAD programms use it’s double potential as well?
As explained above, modelling ‘no’, rendering ‘yes, if the renderer supports it’.
As was noted by @sjdorst above, one has to be very careful to distinguish between interactive model building/editing and rendering.
Rendering involves two crucial facts: First, it does not modify the items being rendered; the model database must be frozen during the render. Second, the effects of illumination are inherently additive and order-free. That means that you can calculate how any particular part of the model appears by dealing separately and in any order with each thing that casts light on it, and can simply add up them up. These facts make it possible to break rendering up into a collection of independent calculations that can all run at the same time, that is, to multi-thread the rendering application. And, in fact, many renderers are multi-threaded and will use multiple CPU cores if available.
Neither of these facts applies to interactive model building and editing! The whole idea is for the user to modify the model’s contents. The effects of modifications are not additive and are order-dependent. The user relies on a correct presentation of the model on the view to guide his edits; no other thread can be manipulating the model at the same time else the user will experience complete chaos!
There are a few tasks, such as saving a file, that could potentially be split off into separate threads. That might avoid freezing SU while a save is taking place. But to do so the app would still have to freeze while making a copy of the model database in memory because the save can’t happen if the user might modify something that has already been saved! The copy would temporarily double SU’s model memory requirement, which of itself might cause problems. Also, because modern OS’s all cache data and run separate OS threads for disk writes, it is unclear how much of an improvement there would be. That is, how does the time for copying in application memory and the again copying to the OS memory compare with just copying to the OS memory? I don’t know the answer, I only know it isn’t a simple question.
Yes, a faster CPU, whether inherently faster or via overclocking, will improve SketchUp’s performance up to a point. There is still a limit based on passing data to the GPU for presentation, so the improvement might be less than 1-for-1.
OK. I see, double CPU issue is clear to me now, thanks a lot.
How do I recognize then which CPu is better for faster MODELING then?
Is it just clock or cache too (is 12MB a lot?). I read that CPU of newer generation are faster though are not getting more GHz but due to getting more cores. Since we don’t need more cores here, what makes newer generation CPUs better (like i-7) from older ones but with the same Hz?
Then GPU limiting the process: is it about bus interface like PCI? How do I choose the best onethen? All my architect fellows advice me Quadro 2000 (I will also work in CAD programms). It’s GDDR5 1024MB, memory clock 2600 MHz and OpenGL 11. But here in this forum I read that Ge Force GTX are better for modeling in SU. Actually why?