[Matteo Villa]

Tested the latest release of Maxwell Render within dual 4090 RTX

Tested with NVIDIA RTX 6000 too.

GPUs not recognized.

Requested an ETA for an update to support new RTX 4000, they don’t have actually any plan for an imminent update/release to support 4000 series.

So the latest viable solution for a gpus rendering workstation are 3090/3090ti or quadro counterpart

[Tea_Bag]

Sadly not much is happening with Maxwell Render at the moment....I hope I am wrong and they are working on a huge update

I have finally switch to Octane Render for MODO to use GPU rendering and so far it is Impressive

[Matteo Villa]

Their response to my ticket about future plan to support Ada Lovelace Gpus, make me wondering the health state of their developing Team.

New workstation will be soon upgraded to Ada Lovelace Quadro.

Other Studios and rendering specialist are upgrading their workstation to RTX 4000.

Even who’s buying new portable workstation will be cut out from using Maxwell Render.

On the long term even Render Farms services upgrading their gpus server would drop eventually the support to Maxwell without an upgrade to make series 4000 usable.

New customers with workstation based upon Ada Lovelace, would never consider Maxwell Render as a viable solution, moving naturally to Octane Render or Vray.

Having multiple workstation based on 3000 series and Threadripper cpus keep me using Maxwell in my workflow, especially since Octane doesn’t have a specific plugin for Archicad 26 or a functional stand alone studio version.

But Twinmotion is on the right way to become a quality and viable solution for architecture and interior design.

I feel sad to loose a great software like Maxwell.

absolutely deserve a V6 and more to come.

[Tea_Bag]

Yup the response is not a positive one.

I think Maxwell Render is still being developed behind the scenes but without care of loosing customers, marketing, feedback etc after all Next Limit is more than just Maxwell Render, other products in their portfolio is probably funding Next Limit.

I assume Maxwell Render is in this dormant state where it might stay for a while and who knows we might be surprised with a V6 one day that is fully featured, Hybrid CPU & GPU Renderer.

Maxwell Render V4 & V5 (GPU) did feel like a Preview release for the GPU Engine. I think series 5 will not see a production ready GPU Engine especially after development & public communication stopped at V5.2....Unfinished , No updates, No Roadmap & Plugins being dropped.

if anything maybe Next Limit drew a line in the sand and said that is it for V5 and are now working on V6 till they have a fully featured GPU Engine....How long? As long as it takes I guess. Next Limit know that they cannot afford to repeat the mistakes of Maxwell Render 4 & 5 series.

Who knows I am just speculating..........it would be fantastic to see a Maxwell Render V6 Production ready GPU Engine plus more.......but only Next Limit know what is going on and where the future lies for Maxwell Render.

[Mark Bell]

Rather than go for the latest, which would no doubt be pricey, why not look at the Benchwell list which shows the PC specs and benchmark achieved. Adding more of the older GPU's should be more affordable so you could create a system with multiple older GPU's vs less higher cost new GPU's? Just a thought.
.

.

[Forester]

It appears that y RTX 3080 + outperforms the RTX 4000 in most of the professional benchmarks. Just a thought.

[Matteo Villa]

https://benchmark.chaos.com/v5/vray-gpu ... only=false

You can clearly check with Vray, Octane, and other rendering software that a single 3080 can’t outperform a single 4090.

Plus the 4090 have have an ECC feature ( like Quadro cards) that’s a life saver on stability within longer rendering project.

More efficient and newer Cuda cores in comparison within 3000 series.

Multi gpus system with more than 4/6 gpus need specific hardware that’s costly upfront and need sometimes a dual PSU configuration.

Maxwell Render didn’t received a major update in the last year.

Considering it doesn’t support Amd gpus, and there is no ETA about a minor release just to integrate the support for Ada Lovelace Gpus, their share in the rendering software market will decrease progressively.

Technology evolve, and if you work in a tech company you must evolve with it.

If we invest in new hardware, we look to newer ones that’s give us an increased performance ratio with other rendering software.

I’m still in love with Maxwell Render, and I would invest more money in the company if a V6 come out.

But actually look like more a death end project.

Clearly I hope to be totally wrong.


https://www.pugetsystems.com/labs/artic ... view-2374/

https://www.pugetsystems.com/labs/artic ... -creation/

[Mark Bell]

A lot of info to go through so I'm sure you've done your homework on best value and performance. Maxwell still appears to have a reasonable sized user base so I don't think it would make sense to cease development? I guess we'll just have to see what gets announced whenever that is.

[Matteo Villa]

Its not about “making homework”, it’s about just know reading a benchmark when you’re investing thousand of € in your hardware.

Twinmotion, Vray, Octane doesn’t wait over one year to release an update.

• There is a lack of communication between the developing team and the still active user base using Maxwell.

• More communication transparency on the website would be welcome.
Like underline for potential new customers that series 4000 ( RTX and Quadros) isn’t supported.

Anyway still waiting to see the V4 and V5 roadmap completed as promised.

[Brian Buxton]

Who's left to develop Maxwell now? They struggle to keep Maxwell cloud functioning, nothing on the website has been updated since the 5.2 release.
I don't think there is anything happening with Realflow either.

[choo-chee]

well as a member using Maxwell for about 15 years I can say that I really don't care too much for GPU rendering as it never had the potential - the GPU cards RAM is never enough for my scenes (arch-viz). And investing 1000's $ for cards that are old after 2-3 years seems wrong. On the other hands, CPU's are much faster now and more stable so I leave PC's working over-night to get a clean image in the morning still on V.4 as there was no point for V.5 and not even the denoiser needed anymore (I'm using AMD CPU's and they are fast!!). I would like to see further improvements and maybe a better dedication to CPU dev so it's being used even faster as most of other renderers are still struggling to get Maxwell quality (unless you're a real expert and know a lot about those...Maxwell is so easy to work with) and that plugin integration will be better (I'm using Maya and it needs some work esp. with the viewport preview). I think they started out very good but after some very bad management decisions they went down a lot. Still dedicating some more time and work could revive their status quite easily I think. Add assets and make sure plugins work as best as possible, and modern CPU's speed will help us power users than cannot render a scene on a card with 12 Gb...

[Brian Buxton]

Keeping the renderer compatible with current hardware is the absolute minimum requirement. If the Maxwell team cannot even do that, then the product is dead.

[Andreas Hopf]

I would like to see further improvements and maybe a better dedication to CPU dev so it's being used even faster as most of other renderers are still struggling to get Maxwell quality (unless you're a real expert and know a lot about those...Maxwell is so easy to work with) and that plugin integration will be better (I'm using Maya and it needs some work esp. with the viewport preview). I think they started out very good but after some very bad management decisions they went down a lot.

I just finished a 3D surface modelling course with 30 technical designers (Alias, Creo, Fusion 360, SolidWorks), and after 3 days, they can do what's shown below. However, Maxwell Studio has not been developed in years, no bugs I listed here years ago were ever ironed out, none of the essential functions needed today were ever implemented. I think the product is at its EOL, and we in the various design professions have to transition to Keyshot.

AXOR Bouroullec faucet and Bisazza mosaic tiles - Maxwell FIRE SL 13 quality 7 10 minutes (beginner test)

[Mark Bell]

I think they're all valid comments. My guess is that with the recent take off of AI rendering on top of the Drag 'n' Drop packages such as Lumion, Enscape, TM, there has been a big shift in customers towards those looking for ease and quickness over high-end true-to-life quality? There has been some improvements this year and last, indicating there's probably a small number of programmers still on Maxwell. At least we don't have to pay annually on a subscription basis unlike the majority of other software. The NL team are also competing with some packages which are given away for free - TM, UE5 which would make it more challenging as a business to turn a profit. I read somewhere that NL did a capital raise a few years ago and were venturing into AI so their resources may be committed until that target has been met?

I think CC has a good approach, and I tend to agree that the CPU works with everything and has been proven over time, whereas GPU still has some gaps to fill. If you get good results that allow you to remain competitive in the sector you're in, why change?

And when you see what Andreas teaches his students to create after only 3 days.....amazing~!

[choo-chee]

all they have to do is merge the weird "roombox" app they did, into maxwell, as an assets manager, and use a bit of AI to know were to put more CPU's horsepower during a render, and they still can move on for years without any need to add all sort of features etc. every year the CPU's market gives you better and faster products. on the other hand, I did try the AI stuff such as Dall-e2 or Midjourney, it's incredible if you need conceptual work but you are way behind if you need it to read your client's drawings... about 25 years ago I did try to create an app that would use AI to understand clients but failed to raise money for it as most of the investors - that were tech guys - said that it's impossible for the computers we had (back then) to solve such complex stuff. maybe today ...