multi core speed increases - new Ryzen

I did a thread on using multi core cpus to increase speed of filterforge. I noticed that you get a significant speed up from 1 to 4 cores ( 8 threads in an i7) - yet I was surprised that when I used a 6 core i7 - it was the exact same speed as the 4 core, which made me worry that filterforge was clamping their algorithm to 4 cores. Getting the speed faster is critical for my work - including buying computers that are best at in, The standard3d graphics or Ai world , you spend money on GPUS for fantastic speed up for the price ( $550 for a geforce can give 50 x speed up). Does anyone know if more than i7 4 core gives you speed ups. The new AMD Ryzen are out - they are CPU that go up to 8 cores ( 16 threads) - do they really speed up filter forge well? Has anyone tried them or get high cores with success? What would by nice for the 1000s of power users using filterforge is what speed up do you really get for what level of cpu- so at least an understand for clients and projects cost / benefit. Ryzen brings this issue to the current discussion because they are cost effective but very high core cpu. So would they be great for speeding up filter forge or a waste of money.

Great to hear that > 4 cores i7 works well. My test two versions back showed no speed up on 6 core. So all fixed now. If you get a chance, test & document your 8 core against a 4 core i7. As you know currently 8 core i7 CPUs have quite a high price delta versus 4 core. So knowing the benefit speed ups you would get for that coin would help. Again why this thread as well about AMD Ryzen super multicore at less price.

Just a bump for anyone able to do speed compare tests on comparing 4 core vrs > 4 core CPUs and/or the new AMD CPUs. I am willing to pay for speed but need to know if it pays off.

Kind of crazy that FIlterForge folks can't give us a good answer / metric on this - the question has been asked over a year ago.

stevedipaola, Compare BENCHMARKS not FilterForge render speeds. Benchmarks are way more accurate and since the tests are carried out exactly the same per hardware, you'll get a better picture and a clearer comparison between different types of hardware. You should always use benchmarks to see which hardware you want to purchase!

There are tons of benchmark scores available online. I personally use Cinebench scores for graphical performance.

I don't think its crazy FF doesn't have an answer. It's honestly a lot crazier to see a company perform strict benchmark tests across tons of possible hardware configurations. Many companies don't do this and probably don't disclose specific numbers per hardware. I've seen some companies disclose which hardwares they've ran beta tests on but for the most part won't disclose anything more. Do understand that conditions may vary between users and all sorts of factors may lead to different render speeds. Nobody wants to make empty promises of faster render speeds just because you're using x processor.

Darn! I'm sorry I must have misread your post last couple nights ago. My mind's numb fr om a ton of work at my office. Either way, I'm really not sure of the issue, it appears to be something you got going than something I've observed here.

I work with FilterForge quite a lot and I've used it on multiple machines just as usual. The speed differences I observe between multiple high end and lower end machines are pretty significant and while I haven't done scientific benchmarks I always feel the difference with render speeds.

This is just to be clear and to just really make sure we're at least talking on the same page:
- Make sure you have multi-threading enabled in your options.
- Do you have the basic or standard versions? Those versions have a lim it on the number of CPUs you can use at once.

Again that's just clarification on the matter. I wouldn't know your setup here. If otherwise I'd throw my hands up. I wish I knew what the issue was but the tests I do here in my art office kinda confirm that my pro version of FF's running perfectly ok and that I do see that speed difference between 8 and 16 threads.