AMD

What is the GPU Stress Test?

A GPU stress test is made to get your GPU to work at its best ability. This test makes sure that the GPU uses all its processing and cooling power. Also, the test makes sure to push up to the max temperature the GPU can handle by giving it all the electric power it can take. The point of doing this test is for you to be able to know the limits of your GPU and how to use it to it’s best ability. Also, make sure that your GPU doesn’t crash when you are using it in normal activities. Furthermore, the stress test will help you decide what GPU best suits you.

How to do a GPU Stress Test?

When you are doing a GPU Stress Test it’s important to monitor your temps and GPU usage. One software that allows you to monitor the GPU while stress testing is the MSI Afterburner. As you are doing the stress test the Afterburner will show GPU usage which sometimes reaches 100% while the top temperature going up to 90°C. But, over 90 °C can be harmful to your GPU.
Next, you should find a proper software to conduct the test. For this purpose there is the software you can select from that suits your need.

GPU stress testing tools.

These tools overclock your GPU to find out the stress it can handle.

FurMark.

This is free to use a stress test that is known as the best stress testing tool. The FurMark allows you to push your GPU to limits and do it without being bothered by any problems it faces. Another good thing about this test is it comes with new updates which makes sure it can run on any GPU. Also, the interface makes sure that even a person without extensive technical knowledge can use it. Although, FurMark has many good features it best suits a basic stress test or for a stability test. Therefore, to conduct a large scale stress test you must look into other stress tests.

Unigine Heaven

This is another good stress test, this is best to test old and current GPU. Although the stress test is quite old it is great for stability and to overclock your GPU. But unlike the FurMark you will have to pay for some parts of the tool. This isn’t the best stress testing tool for new GPU tools.

Unigine Superposition

This is the stress testing tool that you need to use for a large scale test. Superposition can handle the best of the GPU in today’s market. Like Heaven, this tool is free to use but some features you must pay for it.

Once you select the stress testing tools you can set up the overclock options. Once you set up the overclocking options close other unnecessary applications running on your computer. Then you must run your stress test for a certain period until your GPU reaches 100%. Make sure the time you run the stress testing is suits the tools and the GPU.

RX 580 GTX vs 1060 and GTX 970 cards are somehow equal in performances. GTX 970 is the oldest card in the lineup and it was released in 2014. GTX 1060 was released in 2016 and RX 580 in 2017. Even though all these cards have nearly equal benchmarks the prices, some features, VRam  and base clocks have different variations. So, which is good for gaming, which is good for mining and which card is best for video rendering?

Introduction

RX 580 is a high end directX 12 supporting graphic card using GDDR5 memory technology. 580 is manufactured by AMD (advance micro devices). This card has 8 GB of vram and the effective maximum memory speed is 8 Gbps. RX 580 is considered as a high end gaming card which is capable of gaming at HD, FHD, UHD and 4K. The card can output game videos at 50-60 frames per second under HD resolution. GTX 1060 is also a high end gaming graphic card using GDDR 5 technology and available in 6GB and 3GB models.  GTX 970 is bit older than 580 and 1060 but still it is capable and competitive. It is also using GDDR 5 memory technology and has 4 GB vRam.

GPU processor comparison RX 580 vs GTX 1060 and 970

Radeon RX 580 has a base frequency of 1257 MHz ( Boost clock 1340 MHz) while GTX 1060 has a frequency of 1506 MHz (Boost – 1708 MHz) . The old grandpa GTX 970 has a base clock of 1050 MHz which is boostable up to 1753 MHz.

Comparing base frequencies GTX 970 < RX 580 < GTX 1060

Number of GPU cores

RX 580 has 2304 cores or processing units while GTX 1060 has 1280 cores in 6 GB model and 1152 cores in the 3 GB model. GTX 960 has 1024 cores. Likewise a CPU has four to 8 cores in general GPUs have small cores in thousands. A core is a processing unit which can do a single task at a time. With this small cores GPU can do many small tasks at a time. But a CPU can only do a few tasks at a time but the capacity of the individual task is way huge than a GPU.

Comparing number of cores GTX 970 < GTX 1060 < RX 580

Processor architecture nm Technology (Transistor size)

We are currently in the beginning of 5th computer generation which is the AI era. The 4th computer generation was based on VLSI or very large scale integrated circuits. Inside of these integrated circuits or IC’s the functioning unit is transistors.

nm technology means what is the size of a transistor inside a processor. So, smaller the number more transistors a processor can have in a specific area.  Smaller the nm technology more the speed a processor can get. Also, it improves power efficiency. For an example 7 nm AMD processor technology is 35% faster and 65% power saving comparing to the 10 nm technology.

Rx 580 processor is built based on 14 nm architecture while the GTX 1060 has 16 nm technology. The old guy gtx 970 has much slower and inefficient 28 nm technology.

Comparing architecture – GTX 970 <<< GTX 1060 < RX 580

What is base clock?

Clock speed is the frequency which clock generator of a processor can generate pulses.  Well if you don’t get that. Think it is somethings like the amount of work a processor can do within a second. Base clock is where the processor of GPU runs at recommended power levels.

What is Boost clock?

Some processors (Unlocked or over clockable) can run tasks at higher frequencies than the manufacturer recommended frequencies. Under this mode the processor of the GPU consume more electricity and do more work than usual.

What GPU can overclock?

Over clocking generate more heat than usual. So, if you GPU is not well cooled the generating heat can damage the electronic components on the board. So, you have to be very careful if you are about to over clock your GPU or even the CPU.  You may have seen some GPUs have “OC” part in their name. This means this GPU is overclocked in the factory and it has capable cooling system to run at boost clocks.

Power consumption RX 580 vs GTX 1060 Vs GTX 970

The graphic card power of GTX 1060 is 120 W and for GTX 970 it is 145 W. The RX series cards are known as power eaters. The users are usually complaining about the high power consumption of cards like rx 570 and rx 580. However according to AMD the card power of RX 580 is 185 W which is way too high comparing to the other cards.

For RX series graphic cards AMD is recommending a minimum of 500 W power supply. For the GTX 1060 Nvadia is recommending a 400 W PSU and for GTX 970 it is a minimum of 500 W.

Lower the power consumption better it is to your pocket.

So comparing the power consumption, RX 580 < GTX 970 < GTX 1060 ( GTX 1060 has the best power consumption of all three cards)

GPU memory

RX 580 has a physical memory size of 8 GB. The effective memory speed is about 8 Gbps and the maximum memory bandwidth is 256 GB/s. RX 580 is using 256-bit memory interface and the technology is GDDR5.

Geforce GTX 1060 is available in two memory sizes 6 GB and 3 GB.  Both models have an effective memory speed of 8 Gbps and the maximum bandwidth is 192 GB/s. 1060 is using 192 bit memory interface and the memory technology is GDDR5 which is same as RX 580.
Geforce GTX 970 has 4 GB physical memory and the maximum memory bandwidth is 223.4 GB/s. GTX 970 is also using 256-bit memory interface and the memory technology is GDDR5.