[Tested and Burned] EVGA GeForce GTX 580 SC Review

EVGA GeForce GTX 580 SC Review Index

• 1 – Meet with EVGA’s GeForce GTX 580 SC
• 2 – EVGA GeForce GTX 580 SC Features
• 3 – EVGA GeForce GTX 580 SC OpenGL performances
• 4 – EVGA GeForce GTX 580 SC Direct3D performances
• 5 – EVGA GeForce GTX 580 SC Gaming performances
• 6 – EVGA GeForce GTX 580 SC DirectCompute performances
• 7 – EVGA GeForce GTX 580 SC Power consumption and overclocking
• 8 – EVGA GeForce GTX 580 SC: Misc Links and Reviews

7 – EVGA GeForce GTX 580 SC Power consumption and Overclocking

---

UPDATE: final stable overclocking values: GPU: 870MHz, VDDC: 1.125V
More details in this article: EVGA GeForce GTX 580 Superclocked GPU Overclocking Session

---

For the power consumption and maximum stress test, I used the upcoming FurMark 1.9.0. The graphics workload in the new FurMark 1.9.0 has been slightly increased,
leading to more power consumption (few watts in more). But this extra graphics workload can make the difference with high overclocking settings.
It was the second real test of FurMark 1.9.0 (first test has been done with SAPPHIRE HD 6870).

And to find the max stable GPU core clock, I used the version 1.5.0 of EVGA OC Scanner (version 1.5.0 will be released shortly on EVGA website). This tool is really nice because it allows to detect artifacts that are difficult to see (with a size of one or few pixels).

As you may know dear readers, NVIDIA has added a power draw limiter (or over current protection, OCP) on its GTX 580. The power draw limiter includes a hardware monitoring chip and the graphics drivers. One of the functions of the driver is to detect GPU stress test applications like FurMark (FurMark 1.8.2 is blacklist in NVIDIA drivers) in order to limit the power draw. See here for more details: GeForce GTX 580 Power Monitoring Details.

How to bypass the OCP? Two solutions:

• 1 – using a stress test tool that is not blacklisted YET (like FurMark 1.9.0)
• 2 – using GPU-Z 0.4.9 and up: GPU-Z has a magic function to disable GTX 580 OCP. See here: GeForce GTX 580 Unlocked: 350W Under FurMark Thanks To GPU-Z.

So for my burn-in tests, I disabled the OCP with GPU-Z and used FurMark 1.9.0…

The total power consumption of my testbed in idle is 120W (EVGA GTX 580 SC with default clocks).

With default settings (GPU core: 797MHz, Vcore or VDDC: 1.082V, memory: 1012MHz), the total power consumption of the system is 480W for a max GPU temperature of 90°C (FurMark 1.9.0 settings: 1920×1080, fullscreen, Burn-in mode, no AA, no postfx).

We can the calculate the power consumption of EVGA’s GTX 580 SC with stock clocks. The Corsair AX1200 PSU has an efficiency factor of around 0.9 (see this article, there is a graph of the AX1200 efficiency).
P = (480-120) * 0.9
P = 324 watts

In this overclocking test, I only concentrated on the GPU overclocking. I used EVGA Precision to increase the GPU clock and after each new setting, I used a session of OC Scanner (in 1920×1080 fullscreen) to find artifacts and a session of FurMark 1.9.0 to push the GPU beyond the max 😉

• core: 797MHz, mem: 1012MHz. Total power consumption: 480W, GPU temp: 90°C, GPU current: 107.5A and GPU power draw: 110W, FurMark 1.9.0 score: 2433 points (40FPS)
• core: 831MHz, mem: 1012MHz. Total power consumption: 496W, GPU temp: 91°C, GPU current: 112.0A and GPU power draw: 114W
• core: 846MHz, mem: 1012MHz. Total power consumption: 498W, GPU temp: 91°C, GPU current: 114.0A and GPU power draw: 115W
• core: 861MHz, mem: 1012MHz. Total power consumption: 505W, GPU temp: 91°C, GPU current: 115.0A and GPU power draw: 117W, FurMark 1.9.0 score: 2614 points (43FPS)
• core: 880MHz, mem: 1012MHz. Total power consumption: 512W, GPU temp: 92°C, GPU current: 117.0A and GPU power draw: 119W, FurMark 1.9.0 score: 2677 points (44FPS)
• core: 897MHz: artifacts found by OC Scanner

Result: I managed to get a stable value under FurMark 1.9.0 for a GPU clock of 880MHz and the default GPU voltage.

To reproduce such overclocking test, I recommend you to have some quality hardware, because we exceed the electric specifications of the system. My testbed includes a GIGABYTE A-UD5 motherboard and a high-end PSU: Corsair’s AX1200!

The maximun power cosumption of the card is:
P = (512-120) * 0.9
P = 352 watts

Here is a comparative table of the power consumption of the card ALONE (not the total power consumption of the system):

Power: 352W – EVGA GTX 580 SC (core: 880MHz, Vcore:1.082V)

Power: 330W – ASUS ENGTX580> (core: 871MHz, Vcore:1.088V)

Power: 324W – EVGA GTX 580 SC (core: 797MHz, Vcore:1.082V – default settings)

Power: 290W – ASUS ENGTX580 (default settings)

Power: 272W – EVGA GTX 480

Power: 220W – SAPPHIRE Radeon HD 6870 (core: 1000MHz, Vcore: 1.300V)

Power: 185W – ATI Radeon HD 5870

Power: 155W – SAPPHIRE Radeon HD 6870 (core: 900MHz, Vcore: 1.175V)

Power: 147W – ASUS EAH6870

Power: 135W – MSI N460GTX Cyclone 768D5 OC

The max total power consumption of the EVGA GTX 580 SC is 352W while the GPU one is 119W (according to GPU-Z). The difference, 352-119 = 233W, is the power consumption of the rest of the board (especially the VRM).