Jump to content

eGPU bandwidth is invalid.

Hi-Tech
Go to solution Solved by xAcid9,
On 11/1/2019 at 8:08 PM, Hi-Tech said:

However, the RTX2080 score is too low. Do you think the figure at 18,455 is normal?

Found this 

He got similar graphic score as yours so i guess it just the limitation of the eGPU?

When I load eGPU, bandwidth is dramatically reduced.

As you can see, when the eGPU is not in use, the bandwidth represents a normal figure. However, when I play 3D mark or games, the eGPU load rate represents 100% and bandwidth is dramatically reduced. at the sametime GPU core performance is also dramatically reduced.

I connected the eGPU using the Nvme M.2 slot. The interface is R43SG. Then, the eGPU was connected to the external monitor to measure performance. The following picture shows the 3D mark test score and bandwidth and core performance.

M.2 eGPU (Externaldisplay).PNG

Time spy (Externaldisplay).PNG

eGPU bandwidth(Idle state).PNG

eGPU bandwidth(full load state).PNG

Link to comment
Share on other sites

Link to post
Share on other sites

That is normal... you're effectively hamstringing the bandwidth available by using a m.2 slot going from PCIE Gen 3 x16 to PCIE gen 3 x4... you do lose quite a bit of performance as a result. 

Community Standards | Fan Control Software

Please make sure to Quote me or @ me to see your reply!

Just because I am a Moderator does not mean I am always right. Please fact check me and verify my answer. 

 

"Black Out"

Ryzen 9 5900x | Full Custom Water Loop | Asus Crosshair VIII Hero (Wi-Fi) | RTX 3090 Founders | Ballistix 32gb 16-18-18-36 3600mhz 

1tb Samsung 970 Evo | 2x 2tb Crucial MX500 SSD | Fractal Design Meshify S2 | Corsair HX1200 PSU

 

Dedicated Streaming Rig

 Ryzen 7 3700x | Asus B450-F Strix | 16gb Gskill Flare X 3200mhz | Corsair RM550x PSU | Asus Strix GTX1070 | 250gb 860 Evo m.2

Phanteks P300A |  Elgato HD60 Pro | Avermedia Live Gamer Duo | Avermedia 4k GC573 Capture Card

 

Link to comment
Share on other sites

Link to post
Share on other sites

6 hours ago, Hi-Tech said:

@Skiiwee29

However, the RTX2080 score is too low. Do you think the figure at 18,455 is normal?

https://www.techpowerup.com/review/nvidia-geforce-rtx-2080-ti-pci-express-scaling/

CPURyzen 7 5800X Cooler: Arctic Liquid Freezer II 120mm AIO with push-pull Arctic P12 PWM fans RAM: G.Skill Ripjaws V 4x8GB 3600 16-16-16-30

MotherboardASRock X570M Pro4 GPUASRock RX 5700 XT Reference with Eiswolf GPX-Pro 240 AIO Case: Antec P5 PSU: Rosewill Capstone 750M

Monitor: ASUS ROG Strix XG32VC Case Fans: 2x Arctic P12 PWM Storage: HP EX950 1TB NVMe, Mushkin Pilot-E 1TB NVMe, 2x Constellation ES 2TB in RAID1

https://hwbot.org/submission/4497882_btgbullseye_gpupi_v3.3___32b_radeon_rx_5700_xt_13min_37sec_848ms

Link to comment
Share on other sites

Link to post
Share on other sites

@BTGbullseye

According to the information you sent my performance is abnormal. I had degradation of 30 % instead of 9 percent. What can be inferred as the reason for the degradation of performance being more than three times greater than expected? Even those who went through the same procedure as me who connected to Nvme M.2 using the R43SG adapter suffered a 5% degradation. But i'm 30%.

Link to comment
Share on other sites

Link to post
Share on other sites

16 hours ago, Hi-Tech said:

@BTGbullseye

According to the information you sent my performance is abnormal. I had degradation of 30 % instead of 9 percent. What can be inferred as the reason for the degradation of performance being more than three times greater than expected? Even those who went through the same procedure as me who connected to Nvme M.2 using the R43SG adapter suffered a 5% degradation. But i'm 30%.

Run GPU-Z and see if it might be running in PCIe 2.0 mode...

CPURyzen 7 5800X Cooler: Arctic Liquid Freezer II 120mm AIO with push-pull Arctic P12 PWM fans RAM: G.Skill Ripjaws V 4x8GB 3600 16-16-16-30

MotherboardASRock X570M Pro4 GPUASRock RX 5700 XT Reference with Eiswolf GPX-Pro 240 AIO Case: Antec P5 PSU: Rosewill Capstone 750M

Monitor: ASUS ROG Strix XG32VC Case Fans: 2x Arctic P12 PWM Storage: HP EX950 1TB NVMe, Mushkin Pilot-E 1TB NVMe, 2x Constellation ES 2TB in RAID1

https://hwbot.org/submission/4497882_btgbullseye_gpupi_v3.3___32b_radeon_rx_5700_xt_13min_37sec_848ms

Link to comment
Share on other sites

Link to post
Share on other sites

9 hours ago, BTGbullseye said:

Run GPU-Z and see if it might be running in PCIe 2.0 mode...

I checked it a long time ago. It's PCIe 3.0

GPU-Z.PNG

Link to comment
Share on other sites

Link to post
Share on other sites

What motherboard/system?

On many the SSD m.2's are connected to the chipset, not directly to the CPU. So some additional latency, bottlenecking and bandwidth sharing is involved.

F@H
Desktop: i9-13900K, ASUS Z790-E, 64GB DDR5-6000 CL36, RTX3080, 2TB MP600 Pro XT, 2TB SX8200Pro, 2x16TB Ironwolf RAID0, Corsair HX1200, Antec Vortex 360 AIO, Thermaltake Versa H25 TG, Samsung 4K curved 49" TV, 23" secondary, Mountain Everest Max

Mobile SFF rig: i9-9900K, Noctua NH-L9i, Asrock Z390 Phantom ITX-AC, 32GB, GTX1070, 2x1TB SX8200Pro RAID0, 2x5TB 2.5" HDD RAID0, Athena 500W Flex (Noctua fan), Custom 4.7l 3D printed case

 

Asus Zenbook UM325UA, Ryzen 7 5700u, 16GB, 1TB, OLED

 

GPD Win 2

Link to comment
Share on other sites

Link to post
Share on other sites

1 hour ago, Kilrah said:

What motherboard/system?

On many the SSD m.2's are connected to the chipset, not directly to the CPU. So some additional latency, bottlenecking and bandwidth sharing is involved.

My chipset is Moblie Intel HM175 Express Chipset. and I use the UEFI system.

 

motherboard.PNG

CPU.PNG

Link to comment
Share on other sites

Link to post
Share on other sites

6 minutes ago, Hi-Tech said:

@Kilrah @BTGbullseye

Does the HDMI cable length affect eGPU performance? My HDMI cable is 3m long.

Not in the slightest.

 

How many USB 3.0 ports do you have? (they are usually connected to a PCIe lane or 4) What exact system are you using? (full specs, and manufacturer)

CPURyzen 7 5800X Cooler: Arctic Liquid Freezer II 120mm AIO with push-pull Arctic P12 PWM fans RAM: G.Skill Ripjaws V 4x8GB 3600 16-16-16-30

MotherboardASRock X570M Pro4 GPUASRock RX 5700 XT Reference with Eiswolf GPX-Pro 240 AIO Case: Antec P5 PSU: Rosewill Capstone 750M

Monitor: ASUS ROG Strix XG32VC Case Fans: 2x Arctic P12 PWM Storage: HP EX950 1TB NVMe, Mushkin Pilot-E 1TB NVMe, 2x Constellation ES 2TB in RAID1

https://hwbot.org/submission/4497882_btgbullseye_gpupi_v3.3___32b_radeon_rx_5700_xt_13min_37sec_848ms

Link to comment
Share on other sites

Link to post
Share on other sites

17 minutes ago, BTGbullseye said:

Not in the slightest.

 

How many USB 3.0 ports do you have? (they are usually connected to a PCIe lane or 4) What exact system are you using? (full specs, and manufacturer)

This is the I/O Port that my laptop has.

 

HDMI 2.0 x 1, mini DP 1.3 x 1, RJ45 x 1, Thunderbolt™ 3(Type-C)*1
USB 3.0 (Type-A)*4, USB 3.1 Gen.2 (Type-C)*1,
Audio Jack : Earphone x 1 (Hi-Fi/With SPDIF out), MIC in x 1
SD Card Reader x 1 (UHS-II), DC Jack x 1

 

I have four USB 3.0 ports. 

Everything about my laptop is on the following link https://us.aorus.com/product-detail.php?p=546

and I took a screenshot to make it easier for you to see the information.

full specs.PNG

Link to comment
Share on other sites

Link to post
Share on other sites

58 minutes ago, BTGbullseye said:

Not in the slightest.

 

How many USB 3.0 ports do you have? (they are usually connected to a PCIe lane or 4) What exact system are you using? (full specs, and manufacturer)

I'll show you HWinfo if you want.

Link to comment
Share on other sites

Link to post
Share on other sites

6 hours ago, Hi-Tech said:

I'll show you HWinfo if you want.

Actually, it's looking more like a software or driver issue right now. What all was running in the background when the benchmark was running?

CPURyzen 7 5800X Cooler: Arctic Liquid Freezer II 120mm AIO with push-pull Arctic P12 PWM fans RAM: G.Skill Ripjaws V 4x8GB 3600 16-16-16-30

MotherboardASRock X570M Pro4 GPUASRock RX 5700 XT Reference with Eiswolf GPX-Pro 240 AIO Case: Antec P5 PSU: Rosewill Capstone 750M

Monitor: ASUS ROG Strix XG32VC Case Fans: 2x Arctic P12 PWM Storage: HP EX950 1TB NVMe, Mushkin Pilot-E 1TB NVMe, 2x Constellation ES 2TB in RAID1

https://hwbot.org/submission/4497882_btgbullseye_gpupi_v3.3___32b_radeon_rx_5700_xt_13min_37sec_848ms

Link to comment
Share on other sites

Link to post
Share on other sites

7 hours ago, BTGbullseye said:

Actually, it's looking more like a software or driver issue right now. What all was running in the background when the benchmark was running?

Steam

AhnLab Safe Transaction

AhnLab V3 Lite

Intel XTU

MSI Afterburner

RivaTunerStatisticsServer

Killer Control Center

Nvidia Setup

HWMonitor

etc...(There are too many background processes, so I'll put up screenshots.)

Background process1.PNG

Background process2.PNG

Background process3.PNG

Background process4.PNG

Background process5.PNG

Background process6.PNG

Link to comment
Share on other sites

Link to post
Share on other sites

Chrome is a huge detriment to system performance in benchmarks if it's running in the background, as are many antivirus programs. (like the AhnLabs and nProtect stuff) Definitely turn them off when benchmarking.

 

You want to make sure the Nvidia installer isn't actually running when you're benchmarking either.

 

Closing MSI Afterburner would be a good idea too, since RTSS is notorious for causing issues with DX11 and DX12 tests. (usually crashes, but performance drops are also fairly common) ASDF, dtyWork, FloEFD, GSMG, HNC Update, INISafe, Java Update Scheduler, OC Gauge, PerfTune, SMV4, SoftEther, System Guard, Wizvera, Xtu, and all the update services for them should probably also be closed. (and several could probably be entirely removed from your system, but that's an entirely different subject)

 

Try running the benchmark with those off, and see if it helps.

CPURyzen 7 5800X Cooler: Arctic Liquid Freezer II 120mm AIO with push-pull Arctic P12 PWM fans RAM: G.Skill Ripjaws V 4x8GB 3600 16-16-16-30

MotherboardASRock X570M Pro4 GPUASRock RX 5700 XT Reference with Eiswolf GPX-Pro 240 AIO Case: Antec P5 PSU: Rosewill Capstone 750M

Monitor: ASUS ROG Strix XG32VC Case Fans: 2x Arctic P12 PWM Storage: HP EX950 1TB NVMe, Mushkin Pilot-E 1TB NVMe, 2x Constellation ES 2TB in RAID1

https://hwbot.org/submission/4497882_btgbullseye_gpupi_v3.3___32b_radeon_rx_5700_xt_13min_37sec_848ms

Link to comment
Share on other sites

Link to post
Share on other sites

On 11/1/2019 at 8:08 PM, Hi-Tech said:

However, the RTX2080 score is too low. Do you think the figure at 18,455 is normal?

Found this 

He got similar graphic score as yours so i guess it just the limitation of the eGPU?

| Intel i7-3770@4.2Ghz | Asus Z77-V | Zotac 980 Ti Amp! Omega | DDR3 1800mhz 4GB x4 | 300GB Intel DC S3500 SSD | 512GB Plextor M5 Pro | 2x 1TB WD Blue HDD |
 | Enermax NAXN82+ 650W 80Plus Bronze | Fiio E07K | Grado SR80i | Cooler Master XB HAF EVO | Logitech G27 | Logitech G600 | CM Storm Quickfire TK | DualShock 4 |

Link to comment
Share on other sites

Link to post
Share on other sites

3 hours ago, xAcid9 said:

Found this 

He got similar graphic score as yours so i guess it just the limitation of the eGPU?

I think you're right. I have a pretty similar performance. Since the RTX 2080 is the latest graphic series, it may not be ready for eGPU to digest its power. You find a suitable example and perhaps I may be normal. Thank you for your answers. 

 

and I also found the same case with me on egpu.io

https://egpu.io/forums/builds/2018-15-razer-blade-15-rtx-208032gbps-tb3-asus-xg-station-pro-win10-1803-itsage/#post-43867

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×