Jump to content
  • 0

Computer Specifications required to run Secondlife successfully


Chloe Qargen
 Share

You are about to reply to a thread that has been inactive for 326 days.

Please take a moment to consider if this thread is worth bumping.

Question

11 answers to this question

Recommended Posts

  • 0

It's hard to be much more specific than Lindal's response, because we all have different ideas about what "successfully" means. I agree with her that any "gaming" computer sold within the past few years should be fine.  By all means, be sure that it has a graphics card (not a graphics chip on the the motherboard), and opt for a system that doesn't have a postage stamp sized screen.  You will be disappointed by having great graphics on a tiny screen and will be very upset by mediocre graphics there.

  • Like 2
Link to comment
Share on other sites

  • 0
8 hours ago, Chloe Qargen said:

I am looking for the Computer Specifications required to run Second life successfully

These are mine, i am on "ultra" settings 99% of the time, no issues. Vsync disabled, i go from 120 to 280 fps :)

CPU: 11th Gen Intel(R) Core(TM) i9-11900 @ 2.50GHz (2496 MHz)
Memory: 32477 MB
Concurrency: 16
OS Version: Microsoft Windows 11 64-bit (Build 22621.1702)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 3070/PCIe/SSE2
Graphics Card Memory: 8192 MB

  • Like 1
Link to comment
Share on other sites

  • 0
9 hours ago, Chloe Qargen said:

I am looking for the Computer Specifications required to run Second life successfully

The real question is .. are you wanting to run SL ? .. or run SL and a browser with 20 tabs and a copy of photoshop on the side ?

In any case. 16GB Ram, a mid range nvidia card 1060, 1660, 2060 etc etc, the rest is pretty incidental if its a new machine.

  • Like 2
Link to comment
Share on other sites

  • 0

SecondLife has a few generalized bits of hardware qualities it prefers. Namely single core performance, ram quantity (dependent on the viewer) and video card performance can vary wildly dependent on a massive amount of factors. The key one is single core performance though. SL is a very old game that’s on a platform which doesn’t use multiple cores very well, and a lot of its graphical rendering is still cpu bound, not entirely gpu bound. The better your single core performance the better your overall experience.

I recommend using passmark listings to compare cpu choices favoring single core performance. An older 6 or 8 core processor will be outperformed by a much newer 4 core processor when it comes to SL for this reason. Something like the i7 9700 despite being 8c/8t, is outperformed overall by the i3 12100 which is 4c/8t. And in the case of SL it’s a very dramatic difference due to the superior single core performance of the newer i3.

Theres a point of diminishing returns as well depending on cpu generation, the i9 13900 will obviously perform the best, but the difference in its single core performance versus the i5 13400 isn’t really worth it for SecondLife unless you have some other workloads going on at once where you want more cores.

Ram, this will depend mostly on your choice of viewer and how you allocate memory to SL. I use firestorm and let it use whatever it wants, at times SL will use upwards of 10gb of ram that way. I recommend 16gb at minimum for any build these days regardless of purpose. More ram doesn’t hurt (outside of rare scenarios where it does), though if your choices are a much faster 16gb kit with low latency, or a much slower 32gb kit with high latency, go with the fast and low latency option regardless of capacity.

For a video card, basically anything mid tier or better from the last 10 years will play SL fine from a graphical standpoint. SecondLife is still cpu bound, lighting is the big one that’s not handled by the gpu anyway, and the rest of SL’s rendering is a walk in the park for nearly any gpu. You can’t really make a wrong choice here so long as you avoid the low end. A GT 1030 will do fine in SL, but for the price of a GT 1030 you can also just get a used GTX 1070 which will do way better. And really anything of that caliber will do fine. I’ve played SL on dozens of different video cards and found most limitations with graphical performance come down to video memory at higher resolutions more than anything. I have yet to see a scenario really tap out 8gb though. Closest I got was using black dragon I managed to use 6gb of video memory somehow.

If you want a specific type of card recommendation, any hybrid use case card generally plays best with SL in my experience. Nvidia Titan’s, AMD Vega cards, or full workstation style Quadro or FirePro/Radeon pro cards. SL is laid out a lot more like a 3D workspace than a game, and some features of workstation oriented gpu drivers actually do have a minor benefit in game versus the consumer counterparts. This is also why Vega plays well with SL, if you’re on a budget a Vega 56 is dirt cheap and does SL great compared to AMD Polaris or nvidia pascal cards of the same tier. It’s not a super common scenario to come across outside of SecondLife so you don’t hear about that benefit much. I’ve found a similar result with nvidia as well, where I had a first gen GTX Titan punching well above its weight, outperforming a GTX 780 by a decent bit, despite being nearly the same card on paper. Sometimes drivers do matter.

Link to comment
Share on other sites

  • 0
On 5/28/2023 at 11:07 PM, gwynchisholm said:

I’ve found a similar result with nvidia as well, where I had a first gen GTX Titan punching well above its weight, outperforming a GTX 780 by a decent bit, despite being nearly the same card on paper. Sometimes drivers do matter.

The Titan has twice the VRAM of the 780. That's what makes the difference, not the drivers.

Link to comment
Share on other sites

  • 0
3 hours ago, Lyssa Greymoon said:

The Titan has twice the VRAM of the 780. That's what makes the difference, not the drivers.

I probably should’ve specified, a 6gb 780, I forget most are 3gb, Regardless it didn’t top 2gb of memory usage in 1080p anyway.

The cards in question for reference, my 780 is an ASL 6gb King Kong GTX 780

IMG_4576.jpeg.042b548380d188dd7e6a641ce8

My first gen Titan is an EVGA SC model, but both cards are clocked not the same but similarly. Under most circumstances the Titan should outperform the 780 and be closer to a 780ti (which usually outperforms the Titan), but the difference here was way out of the margin i was expecting, upwards of a 10fps difference over the 780.

And this can be replicated with a variety of other cards in my collection. A Quadro M4000 outperforming a GTX 970, a Tesla C2050 outperforming a GTX 570, a Radeon pro WX4100 outperforming an RX460, etc. Its usually nothing super massive, but it’s this outlier of performing above conventional expectation as if you were comparing performance in conventional games.

I am not an expert on the technologies that make professional cards and drivers what they are, but I suspect this has to do with how they’re optimized with OpenGL in mind, versus consumer gpus where directX and Vulkan would be a bigger focus.

Link to comment
Share on other sites

  • 0

Here is my "main system" I can go to around 90FPS in Firestorm, bit higher in other viewers. I have gotten SL Viewers to run on my Steam Deck at around 60FPS. TBH though I frame limit because of the refresh rate of the monitors my system is attached to. I have a secondary system my GF is using that is running a AMD RX580 8GB card same amount of RAM and a bit of an older Intel CPU and Windows 10 with about 60-70 FPS. It all depends on your whole setup what resolution and refresh rate your monitors run at. Also, what you find acceptable for your graphics settings vs. your budget.

CPU: AMD Ryzen 5 3600 6-Core Processor (2200 MHz)
Memory: 32048 MB
Concurrency: 12
OS Version: Linux 5.15.0-73-generic #80-Ubuntu SMP Mon May 15 15:18:26 UTC 2023 x86_64
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 3060 Ti/PCIe/SSE2
Graphics Card Memory: 8192 MB
OpenGL Version: 4.6.0 NVIDIA 530.41.03

Link to comment
Share on other sites

  • 0

What makes the biggest difference is what is being rendered.  My i9-13900K and RTX 4080 will render better than my i9-9900K and 1070 GTX, but saying X FPS is useless because you don't know what they are rendering.  I could have the i9-13900K and RTX 4080 bogged down rendering 5120 x 1440 in a crowd in a built-up area at dusk with shadows enabled and say 48 FPS and I could have the i9-9900K and 1070 GTX rendering 1920 x 1200 when I am alone in Voice Echo Canyon at mid-day with shadows disabled and say 180 FPS.

I wish we had a test scene accessible to all via a method that makes it consistent for every test.

  • Like 1
Link to comment
Share on other sites

  • 0
On 5/31/2023 at 10:17 AM, Ardy Lay said:

I wish we had a test scene accessible to all via a method that makes it consistent for every test.

Going up above 1000m in a region with no other avatars in it and out of sight of any skyboxes in the region should give you a pretty consistent reading.

Link to comment
Share on other sites

  • 0
44 minutes ago, Lindal Kidd said:

Going up above 1000m in a region with no other avatars in it and out of sight of any skyboxes in the region should give you a pretty consistent reading.

But that would be a pretty consistently useless reading of rendering pretty much nothing.

Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 326 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share

×
×
  • Create New...