Download

Anyone tried latest AMD hardware with UDK?

Have anyone tried udk performance on those amd ryzen chips and and/or rx gpus?
Everyone says that a ryzen cpu mixed with a nvidia gpu gives less fps.Managed to see some tests of unreal engine 3 games like batman and rocket league(dx9)

So i supose that a ryzen shoud be paired with a amd gpu.Since its a new hardware and udk is old im not sure what to expect?(but somehow i doubt it shoud be a problem as udk runes on old amd cpus and gpus too)?

Its hard to ignore those tempting prices.

Prices yeah they’re tempting.

I’ve have 2 AMD cards fail on me. But I have several outdated Nvidia cards lying around, none of which ever failed. Maybe AMD has changed, but I’m twice bitten…

I have mainly had amd cards and never had a bad one,and always had amd cpu.

currently on a gtx680.went with it for PhysX only to find the older amd I took out was just a fast in software.

My Setup:
AMD 1800XT RyZen
64GB PC3200 RAM
ASRock Taichi board
1080GTX EKWB liquid cool series
EVO PCiE 3.0 2TB M.2 SSD (only used for primary drive OS, not used for Unreal due to I/O limitations on SSD)
Barracuda 4TB SATA hard drives (nothing special)

When doing, well, anything, the combo of it all rips UE4 and has 0 issues. There is actually already a thread with real testing done and all the numbers so you can see the difference. My computer boots faster then my monitors can turn on and display the Windows screen (Win 10 Home). I barely notice any “loading”/“pausing” when using UE4. Imports and building shaders isn’t even a notice on my end anymore.

Tho, i do want to go out and say the build was about $5,000. So, i mean, its not an average build at all. In my project i went from ~30FPS to 90+ (didnt update the card at all, last video card was a 1080GTX STRIX)

EDIT:
AMD announced quad-GPU video cards coming out 2018? maybe it was this summer with the thread-ripper chip. 16 core / 32 thread CPU … big jumps coming.

This is good to know.I actually ment performance with new AMD hardware with udk as this is the ue3 subforum hehe.No problem :stuck_out_tongue:

I think i will wait a few weeks to see that amd thread ripper everyone talks about.Its tempting to see 16 cores for 850$ vs the intel 16 cores for 1700$.
Shoud be a monster to crunching lightmass at production level.(and other apps too).

Still if a ue4 user has new ryzen cpu and or a amd rx gpu and by some chance decide to take a spin with udk to give us a heads up on the performance>there are download links above this thread.

If you want me to do some testing on my build to see if there is a worthy performance boost, i will be more then happy to do so with whatever you want to throw my way.

Very kind of you.Im interested in the general performance of the ryzen with udk as its a new chip.If you want you can download the 2013-07 udk build

It comes with a few maps.To test the cpu you coud open the necropolis map that comes with udk>select all lights and move them a cm in order to break the lightning and bake with light mass in high or production quality.I suspect all eight cores shoud boost while baking(see the task manager).You coud see how much time it will take to rebake the scene(with a cpu like that i suspect 10min?).Maybe after that, play the level a litle to see if there is any stutering,rendering errors?Everything by deffault will be running in dx9.

And if its not too late it loud be cool to know your cpu temps after the test.
Cheers :wink:

temps wont be an issue, i had a full EKWB custom loop made for my desk build.
I will do this tonight and post up some data

I am currently building. I forgot i have my own project open but, ummm, well, F it … im going for it while using UE4

It wont let me add another pic, but, im getting 0 jitter, 0 lag, no studder, 0 render issues @ full screen @ 100FPS

My “system” screencap:

just to show WHILE building …
b48abe1c411be0505770dda5e5a8baec84429aeb.png

==LOG AFTER BUILD==

7:08:13 PM: [Interface:TryOpenConnection] Local connection established
7:08:13 PM: [Job] Accepted Job 9D1A5010-42132AAC-A1BB93B0-24C198AC
7:08:15 PM: [Job] Launched Job UnrealLightmass_2013-07-01_17-34-35_984384-64bit.exe
7:08:15 PM: [Job] PID is 7460
7:08:15 PM: [Job] GUID is “9D1A5010-42132AAC-A1BB93B0-24C198AC”
7:08:15 PM: Log file created: UnrealLightmass_DESKTOP-M35ORHU_FA0AA7C94ECE1478041A9A910A61DB4E.log
7:08:15 PM: Lightmass WIN64 started on: DESKTOP-M35ORHU. Command-line: “D:\UDK\UDK-2013-07\Binaries\SwarmCache\Jobs\Job-9D1A5010-42132AAC-A1BB93B0-24C198AC\UnrealLightmass.exe” 9D1A501042132AACA1BB93B024C198AC -trisperleaf 4
7:08:15 PM: Processing scene GUID: 9D1A501042132AACA1BB93B024C198AC with 16 threads
7:08:15 PM: Building static lighting…
7:08:15 PM: [Job] Found a parent connection for PID 7460
7:08:15 PM: [Job] 1CD5417D → 30F592AA
7:08:15 PM: [Interface:TryOpenConnection] Local connection established
7:08:18 PM: Measured CPU frequency: 3.60 GHz
7:08:18 PM: FStaticLightingSystem started using GKDOPMaxTrisPerLeaf: 4
7:08:18 PM: Number of texture mappings: 3621
7:08:18 PM: Number of vertex mappings: 0
7:08:18 PM: Number of terrain mappings: 8
7:08:18 PM: Number of fluid mappings: 0
7:08:18 PM: Number of landscape mappings: 0
7:08:18 PM: Number of SpeedTree mappings: 0
7:08:18 PM: Number of BSP mappings: 1231
7:08:18 PM: Number of static mesh instance mappings: 3621
7:08:18 PM: Number of SpeedTree mesh instance mappings: 0
7:08:23 PM: Reserving memory for 4860 meshes, 1789034 vertices, 2117196 triangles
7:08:24 PM: Scene surface area calculated at 108255.148 million units (8.396% of the estimated 1289334.250 million units)
7:08:24 PM: Importance volume surface area calculated at 2762.409 million units (94.536% of the estimated 2922.084 million units)
7:08:24 PM: Clamped the number of indirect photon paths to 20000.
7:08:25 PM: Building kDOP took 1.33 seconds.
7:08:25 PM: Static lighting kDOP: 1394014 nodes, 697008 leaves, 2788032 triangles, 1767123 vertices
7:08:25 PM: Static lighting kDOP: 25.154% wasted space in leaves
7:08:25 PM: kDopTree.Nodes : 148.9Mb
7:08:25 PM: kDopTree.SOATriangles : 191.4Mb
7:08:25 PM: kDOPTriangles : 0.0Mb
7:08:25 PM: TrianglePayloads : 63.7Mb
7:08:25 PM: MeshInfos : 0.0Mb
7:08:25 PM: Vertices : 27.3Mb
7:08:25 PM: UVs : 13.6Mb
7:08:25 PM: LightmapUVs : 13.6Mb
7:08:25 PM: Static lighting kDOP: 1394014 nodes, 697008 leaves, 2788032 triangles, 1767123 vertices, 458.7 Mb
7:08:25 PM: Processing…
7:08:30 PM: EmitDirectPhotons complete, 4.388 million photons emitted in 4.6 seconds
7:08:36 PM: EmitIndirectPhotons complete, 13.260 million photons emitted in 5.4 seconds
7:08:36 PM: Marking Irradiance Photons complete, 0.805 million photons marked in 0.3 seconds
7:08:41 PM: Caching Irradiance Photons complete, 1.938 million cache samples in 4.6 seconds
7:08:43 PM: Calculate Irradiance Photons complete, 0.294 million irradiance calculations in 1.8 seconds
7:09:57 PM: Lighting 10.0%
7:11:04 PM: Lighting 20.0%
7:12:16 PM: Lighting 30.0%
7:51:14 PM: Lighting 40.0%
7:53:00 PM: Lighting 50.0%
7:54:39 PM: Lighting 60.0%
7:56:22 PM: Lighting 70.0%
7:58:14 PM: Lighting 80.0%
8:00:10 PM: Lighting 90.0%
8:02:36 PM: [Job] Job is a success!
8:02:37 PM:
8:02:37 PM:
8:02:37 PM: Total Static Lighting time: 3259.06 seconds, 16 threads
8:02:37 PM: 0.2% 6.8s Scene setup
8:02:37 PM: 0.2% 4.9s Mesh Area Light setup
8:02:37 PM: 0.1% 4.6s Emit Direct Photons
8:02:37 PM: 0.0% 0.7s Cache Indirect Photon Paths
8:02:37 PM: 0.2% 5.4s Emit Indirect Photons
8:02:37 PM: 0.0% 0.3s Mark 0.805 million Irradiance Photons
8:02:37 PM: 0.1% 4.6s Cache 1.938 million Irradiance Photon Samples on surfaces
8:02:37 PM: 0.1% 1.8s Calculate 0.294 million Irradiance Photons
8:02:37 PM: 99.2% 3234.4s Lighting
8:02:37 PM: 0.0% 0.5s Unaccounted
8:02:37 PM:
8:02:37 PM:
8:02:37 PM:
8:02:37 PM: Total Lighting thread seconds: 51747.49
8:02:37 PM: 0.3% 143.4s Texel and vertex setup
8:02:37 PM: 8.7% 4500.6s Direct lighting
8:02:37 PM: 8.3% 4288.2s Area shadows with 1372.927 million rays
8:02:37 PM: 0.1% 26.0s Signed distance field source sparse sampling
8:02:37 PM: 0.3% 174.9s Signed distance field source refining sampling
8:02:37 PM: 0.0% 10.5s Signed distance field transition searching
8:02:37 PM: 0.0% 0.9s Unaccounted
8:02:37 PM: 21.6% 11172.1s Indirect lighting
8:02:37 PM: 2.8% 1446.9s ImportancePhotonGatherTime
8:02:37 PM: 1.0% 497.7s CalculateImportanceSampleTime
8:02:37 PM: 15.9% 8215.3s FirstBounceRayTraceTime for 2555.892 million rays
8:02:37 PM: 1.3% 680.8s CalculateExitantRadiance
8:02:37 PM: 0.0% 7.8s CausticPhotonGatherTime
8:02:37 PM: 0.6% 323.6s Unaccounted
8:02:37 PM: 0.1% 65.0s Final IrradianceCache Interpolation
8:02:37 PM: 0.0% 6.1s Dominant shadow map
8:02:37 PM: 69.1% 35770.4s Precomputed Visibility
8:02:37 PM: 11.1% 5740.4s Sample generation
8:02:37 PM: 53.3% 27591.4s Uniform tracing
8:02:37 PM: 4.5% 2346.0s Importance sampling
8:02:37 PM: 0.1% 66.5s Volume Samples
8:02:37 PM: 0.0% 23.5s Unaccounted
8:02:37 PM:
8:02:37 PM:
8:02:37 PM: Lighting Threaded processing efficiency 100.0%, 16.0 speedup with 16 threads
8:02:37 PM: Traced 3987.799 million first hit visibility rays for a total of 12442.6 thread seconds (0.320 million per thread second)
8:02:37 PM: Traced 22696.334 million boolean visibility rays for a total of 26867.5 thread seconds (0.845 million per thread second)
8:02:37 PM: Scene radius 320315.4, Importance bounds radius 15249.0
8:02:37 PM: 4860 Mappings, 11.895 million Texels, 8.726 million mapped texels, 0.000 million Vertices, 0.000 million Vertex samples processed
8:02:37 PM:
8:02:37 PM: 30.7% of Total Lighting thread seconds on Texture Mappings, 0.0% on Vertex Mappings, 0% on Volume Samples, 69% on Visibility, 0.1% Unaccounted
8:02:37 PM: 653 Lights total, 157.3 Shadow rays per light sample on average
8:02:37 PM: 68 Emissive meshes, 384 Mesh area lights, 9142 simplified mesh area light primitives, 345784 original primitives
8:02:37 PM: Signed distance field shadows: 12.1 average upsample factor, 8.726 million sparse source rays, 46.410 million refining source rays, 2.321 transition search scatters
8:02:37 PM: 21200 Volume lighting samples, 98.5% placed on surfaces, 1.5% placed in the volume
8:02:37 PM: Precomputed Visibility 9667 Cells (2.0% from camera tracks, 9863 processed on this agent), 3763 Meshes, 22689.672 million rays, 4530.6Kb data
8:02:37 PM: 37.114 million queries, 20.7% visible, (3.6% trivially visible, 83.0% explicit sampling, 13.4% importance sampling)
8:02:37 PM:
8:02:37 PM: 4.388 million first pass Photons Emitted (out of 4.388 million requested) to deposit 0.629 million Direct Photons and 20000 Indirect Photon Paths, efficiency of 14.34%
8:02:37 PM: 13.260 million second pass Photons Emitted (out of 13.260 million requested) to deposit 1.405 million Indirect Photons and 0.000 million CausticPhotons, efficiency of 10.59%
8:02:37 PM: 16.102 million Photon Gathers, 1.685 million Irradiance Photon Gathers
8:02:37 PM: 142.321 million Importance Photons found, 2405.199 million Importance Photon PDF calculations
8:02:37 PM: 61.8% Bounce 1 Irradiance cache miss rate (8.726 million lookups, 5.396 million misses)
8:02:37 PM: 1683.9 Mb Peak Working Set
8:02:37 PM:
8:02:37 PM:
8:02:37 PM: Lightmass on DESKTOP-M35ORHU: 54:22 min total, 3.30 sec importing, 6.80 sec setup, 17.9 sec photons, 53:54 min processing, 0 ms extra exporting [3947/4860 mappings]. Threads: 15:-53084:21 hours total, 4:-14136:44 hours processing.
8:02:37 PM: Lighting complete [Startup = 3.30 sec, Lighting = 54:19 min]
8:02:38 PM: [CloseConnection] Closing connection 30F592AA using handle 30F592AA
8:02:38 PM: [CloseConnection] Connection confirmed for disconnection 30F592AA
8:02:38 PM: [CloseConnection] Connection disconnected 30F592AA
8:02:38 PM: [GetMessage] Safely returning to 30F592AA with no message
8:02:38 PM: [CloseConnection] Closing connection 1CD5417D using handle 1CD5417D
8:02:38 PM: [CloseConnection] Connection confirmed for disconnection 1CD5417D
8:02:38 PM: [CloseConnection] Closing orphaned Job (9D1A5010-42132AAC-A1BB93B0-24C198AC)
8:02:38 PM: [CloseConnection] Connection disconnected 1CD5417D
8:02:38 PM: [GetMessage] Safely returning to 1CD5417D with no message
8:02:38 PM: [MaintainConnections] Local connection has closed (1CD5417D)
8:02:38 PM: [MaintainConnections] Local connection has closed (30F592AA)
8:02:38 PM: [MaintainConnections] Removed connection 1CD5417D
8:02:38 PM: [MaintainConnections] Removed connection 30F592AA
8:02:38 PM: [MaintainConnections] All connections have closed

Thanks man :cool:

So let me see if i got this right (havent used windows 10)

-So a ue4 project was opened in the background
-You broke and rebuild the lightning in the (Necropolis?)map in production or high? For a total of 54 minutes.
-All 8 cpus were used(i supose each running at 3.5-3.6ghz) during the bake process.

Did i get it right?

Yes.
Yes, i built on production.
All 8 cores were used
Im not 100% sure what it built, if it built the map or anything else, thats why i posted the data in full. I am actually pretty disappointed if it was a 54min LIGHT only build. I did select “build lighting” from the build menu, so, i can only assume it was lighting only.

Ok,just for kicks im going to do the same as you did on my side and we will compare.
Just in case im going to show a pic here:

With necropolis opened i went in to the content browser>scene tab>in filter text i typed light and shift selected all of the lights.With them selected i moved them a litle up in the editor to break the lightning and started to build wilth production and use lightmass.

https://s13.postimg.org/gihca35iv/Untitled-1.jpg

Im doing this on a first gen i7 950 4 core 3.2 ghz cpu.After is finished i will post the need time it needed and we can compare the %.

Cheers

I hope it finishes as its 31c in my room xd.

Yeah i am interested to see …
I wonder if it wasn’t using all 8 to build lighting tho. I wonder if it was using a single to push lighting. I have no idea. It certainly built a lot faster then i remember UE4 building anything.

Made it!

So if im reading the log right(as i see some strange?) representation of numbers i made it in allmost exactly 3 hours.

I saw that all my 4 cores went up and stayed active and looking at this pic from the task manager(in spanish)

Since is a 4 core > 8 thread all cores seem to take advantage from lightmass building.
I suspect you did used all of your 8 cores but if you are not sure,just click a quick light build (doesnt matter the quality )and see in your task manager if all cores are active and than cancel your bake.No need to do it again all over xd.

It is a complex level so i suspect if all 8 cores were used, than 54minutes vs my 3 hours is pretty good!That in theory shoud mean that the 16 core amd thread ripper im waiting for shoud be close to half of your time :smiley:

What looks strange is that you got 100fps when you played.With a beast of a gpu like that it shoud say 120fps as udk is set max to that.I have a quadro k4200(gtx 670)and i do get 120fps.
I suspect that your fps is just stronger or it may have something to do with people saying that Ryzens behave strange in 1080p or maybe is the gtx fault.In the start of my thread i do show a link to a test of a unreal engine 3 game and there it shows that a nvidia gpu pared with a ryzen most of the time has a loss of 20-25 fps vs the same ryzen with a amd gpu.I suspect its a corporate sabotage but i really dont care.

Here is my log.

15:48:26: [Interface:TryOpenConnection] Local connection established
15:48:26: [Job] Accepted Job 623639B6-4ECC512F-A338EC82-3073008E
15:49:26: [Job] Launched Job UnrealLightmass_2013-07-01_17-34-35_984384-64bit.exe
15:49:26: [Job] PID is 3572
15:49:26: [Job] GUID is “623639B6-4ECC512F-A338EC82-3073008E”
15:49:31: Log file created: UnrealLightmass_EQUIPO1_9DF4751C4B92775BE1E8F7A744106F51.log
15:49:31: Lightmass WIN64 started on: EQUIPO1. Command-line: “C:\UDK\UDK-2013-07\Binaries\SwarmCache\Jobs\Job-623639B6-4ECC512F-A338EC82-3073008E\UnrealLightmass.exe” 623639B64ECC512FA338EC823073008E -trisperleaf 4
15:49:31: Processing scene GUID: 623639B64ECC512FA338EC823073008E with 8 threads
15:49:31: Building static lighting…
15:49:31: [Job] Found a parent connection for PID 3572
15:49:31: [Job] 148FECE3 → 35F6EC1C
15:49:31: [Interface:TryOpenConnection] Local connection established
15:49:37: Measured CPU frequency: 3.07 GHz
15:49:37: FStaticLightingSystem started using GKDOPMaxTrisPerLeaf: 4
15:49:37: Number of texture mappings: 3621
15:49:37: Number of vertex mappings: 0
15:49:37: Number of terrain mappings: 8
15:49:37: Number of fluid mappings: 0
15:49:37: Number of landscape mappings: 0
15:49:37: Number of SpeedTree mappings: 0
15:49:37: Number of BSP mappings: 1231
15:49:37: Number of static mesh instance mappings: 3621
15:49:37: Number of SpeedTree mesh instance mappings: 0
15:49:43: Reserving memory for 4860 meshes, 1789034 vertices, 2117196 triangles
15:49:43: Scene surface area calculated at 108255.148 million units (8.396% of the estimated 1289334.250 million units)
15:49:43: Importance volume surface area calculated at 2708.165 million units (92.679% of the estimated 2922.084 million units)
15:49:43: Clamped the number of indirect photon paths to 20000.
15:49:46: Building kDOP took 2.22 seconds.
15:49:46: Static lighting kDOP: 1394014 nodes, 697008 leaves, 2788032 triangles, 1767123 vertices
15:49:46: Static lighting kDOP: 25.154% wasted space in leaves
15:49:46: kDopTree.Nodes : 148.9Mb
15:49:46: kDopTree.SOATriangles : 191.4Mb
15:49:46: kDOPTriangles : 0.0Mb
15:49:46: TrianglePayloads : 63.7Mb
15:49:46: MeshInfos : 0.0Mb
15:49:46: Vertices : 27.3Mb
15:49:46: UVs : 13.6Mb
15:49:46: LightmapUVs : 13.6Mb
15:49:46: Static lighting kDOP: 1394014 nodes, 697008 leaves, 2788032 triangles, 1767123 vertices, 458.7 Mb
15:49:46: Processing…
15:50:00: EmitDirectPhotons complete, 4.388 million photons emitted in 14.0 seconds
15:50:15: EmitIndirectPhotons complete, 12.999 million photons emitted in 14.1 seconds
15:50:16: Marking Irradiance Photons complete, 0.589 million photons marked in 0.4 seconds
15:50:26: Caching Irradiance Photons complete, 1.938 million cache samples in 9.9 seconds
15:50:29: Calculate Irradiance Photons complete, 0.237 million irradiance calculations in 3.8 seconds
15:54:34: Lighting 10.0%
15:58:02: Lighting 20.0%
16:02:09: Lighting 30.0%
18:11:17: Lighting 40.0%
18:17:22: Lighting 50.0%
18:22:54: Lighting 60.0%
18:28:44: Lighting 70.0%
18:35:17: Lighting 80.0%
18:41:50: Lighting 90.0%
18:50:06: [Job] Job is a success!
18:50:08:
18:50:08:
18:50:08: Total Static Lighting time: 10831.34 seconds, 8 threads
18:50:08: 0.1% 8.7s Scene setup
18:50:08: 0.1% 5.8s Mesh Area Light setup
18:50:08: 0.1% 14.0s Emit Direct Photons
18:50:08: 0.0% 1.5s Cache Indirect Photon Paths
18:50:08: 0.1% 14.1s Emit Indirect Photons
18:50:08: 0.0% 0.4s Mark 0.589 million Irradiance Photons
18:50:08: 0.1% 9.9s Cache 1.938 million Irradiance Photon Samples on surfaces
18:50:08: 0.0% 3.8s Calculate 0.237 million Irradiance Photons
18:50:08: 99.5% 10778.2s Lighting
18:50:08: 0.0% 0.7s Unaccounted
18:50:08:
18:50:08:
18:50:08:
18:50:08: Total Lighting thread seconds: 86225.51
18:50:08: 0.3% 277.2s Texel and vertex setup
18:50:08: 10.6% 9166.5s Direct lighting
18:50:08: 10.1% 8739.1s Area shadows with 1506.696 million rays
18:50:08: 0.1% 72.9s Signed distance field source sparse sampling
18:50:08: 0.4% 332.2s Signed distance field source refining sampling
18:50:08: 0.0% 21.5s Signed distance field transition searching
18:50:08: 0.0% 0.8s Unaccounted
18:50:08: 19.7% 17022.0s Indirect lighting
18:50:08: 2.3% 1958.9s ImportancePhotonGatherTime
18:50:08: 0.8% 731.2s CalculateImportanceSampleTime
18:50:08: 15.2% 13090.8s FirstBounceRayTraceTime for 2553.629 million rays
18:50:08: 0.9% 799.7s CalculateExitantRadiance
18:50:08: 0.0% 7.9s CausticPhotonGatherTime
18:50:08: 0.5% 433.5s Unaccounted
18:50:08: 0.1% 100.9s Final IrradianceCache Interpolation
18:50:08: 0.0% 13.9s Dominant shadow map
18:50:08: 68.1% 58700.2s Precomputed Visibility
18:50:08: 10.0% 8641.4s Sample generation
18:50:08: 53.5% 46144.5s Uniform tracing
18:50:08: 4.5% 3882.8s Importance sampling
18:50:08: 0.1% 102.9s Volume Samples
18:50:08: 1.0% 841.9s Unaccounted
18:50:08:
18:50:08:
18:50:08: Lighting Threaded processing efficiency 100.0%, 8.0 speedup with 8 threads
18:50:08: Traced 4119.309 million first hit visibility rays for a total of 21790.3 thread seconds (0.189 million per thread second)
18:50:08: Traced 22695.221 million boolean visibility rays for a total of 44672.6 thread seconds (0.508 million per thread second)
18:50:08: Scene radius 320315.4, Importance bounds radius 15249.0
18:50:08: 4860 Mappings, 11.895 million Texels, 8.726 million mapped texels, 0.000 million Vertices, 0.000 million Vertex samples processed
18:50:08:
18:50:08: 30.8% of Total Lighting thread seconds on Texture Mappings, 0.0% on Vertex Mappings, 0% on Volume Samples, 68% on Visibility, 1.0% Unaccounted
18:50:08: 653 Lights total, 172.7 Shadow rays per light sample on average
18:50:08: 68 Emissive meshes, 384 Mesh area lights, 9142 simplified mesh area light primitives, 345784 original primitives
18:50:08: Signed distance field shadows: 12.1 average upsample factor, 8.726 million sparse source rays, 46.415 million refining source rays, 2.320 transition search scatters
18:50:08: 21181 Volume lighting samples, 98.5% placed on surfaces, 1.5% placed in the volume
18:50:08: Precomputed Visibility 9667 Cells (2.0% from camera tracks, 9863 processed on this agent), 3763 Meshes, 22689.674 million rays, 4530.6Kb data
18:50:08: 37.114 million queries, 20.7% visible, (3.6% trivially visible, 83.0% explicit sampling, 13.4% importance sampling)
18:50:08:
18:50:08: 4.388 million first pass Photons Emitted (out of 4.388 million requested) to deposit 0.419 million Direct Photons and 20000 Indirect Photon Paths, efficiency of 9.55%
18:50:08: 12.999 million second pass Photons Emitted (out of 12.999 million requested) to deposit 1.357 million Indirect Photons and 0.000 million CausticPhotons, efficiency of 10.44%
18:50:08: 15.660 million Photon Gathers, 1.683 million Irradiance Photon Gathers
18:50:08: 141.936 million Importance Photons found, 2401.353 million Importance Photon PDF calculations
18:50:08: 61.8% Bounce 1 Irradiance cache miss rate (8.726 million lookups, 5.396 million misses)
18:50:08: 1648.6 Mb Peak Working Set
18:50:08:
18:50:08:
18:50:08: Lightmass on EQUIPO1: 3:-10620:37 hours total, 5.81 sec importing, 8.74 sec setup, 44.3 sec photons, 2:-7021:38 hours processing, 0 ms extra exporting [3947/4860 mappings]. Threads: 26:-91984:43 hours total, 7:-24758:50 hours processing.
18:50:08: Lighting complete [Startup = 5.81 sec, Lighting = 3:-10620:31 hours]
18:50:08: [CloseConnection] Closing connection 35F6EC1C using handle 35F6EC1C
18:50:08: [CloseConnection] Connection confirmed for disconnection 35F6EC1C
18:50:08: [MaintainConnections] Connection 35F6EC1C is DISCONNECTING
18:50:08: [CloseConnection] Connection disconnected 35F6EC1C
18:50:08: [GetMessage] Safely returning to 35F6EC1C with no message
18:50:08: [CloseConnection] Closing connection 148FECE3 using handle 148FECE3
18:50:08: [CloseConnection] Connection confirmed for disconnection 148FECE3
18:50:08: [CloseConnection] Closing orphaned Job (623639B6-4ECC512F-A338EC82-3073008E)
18:50:08: [CloseConnection] Connection disconnected 148FECE3
18:50:08: [GetMessage] Safely returning to 148FECE3 with no message
18:50:08: [MaintainConnections] Local connection has closed (148FECE3)
18:50:08: [MaintainConnections] Local connection has closed (35F6EC1C)
18:50:08: [MaintainConnections] Removed connection 148FECE3
18:50:08: [MaintainConnections] Removed connection 35F6EC1C
18:50:08: [MaintainConnections] All connections have closed

yeah the numbers look ok when seeing side-by-side
The threadripper, in my guess, would easily be -50% time for you. My Zen 1800xt is a 8core with 16 threads, ripper is a 16 core with 32 threads. I would imagine, tho, with the boost in Ghz, it should hit -60% less time to build. Even at -50% thats really nice.

As for the GTX1080, yea, sitting at 100FPS static in game (played in editor) does seem low, but, it didnt move but maybe 1 or 2 FPS up or down. I havent noticed anything in anoy other games for drops of FPS. Just feel more static now is all.

Yup,most probably it will still show you 100 fps in 2k or even 4k so its a driver thing.I will probably go with a amd gpu this time since i will use the amd cpu.One of those new radeon frontier editions or the new cheaper radeon pro duo sound interesting.With the duo i wont take advantage of sli in ue3 but with that gpu you can dedicate one of the two wx7100 to work in udk and the other to gpu render in maya at the same time without losing frames.
Will see when they come out.

Meanwhile if someone is planning on upgrading his cpu ,feel free to compare your render time scores here so that you know +/- what you will get.

Thank you AP_Studios for the favour :wink:

It was my pleasure. Anything i can do to help people out before they spend/waste money.
I think i may move to the AMD quad-GPU once they come out … in 2018? 1080 should hold me up fine until then.