To test out the support for DirectX 11 gaming, we still roped in Tomb Raider, to ascertain how the IGPs handle it under the game's normal quality settings.
We have updated the benchmark titles and the stalwart 3DMark (2013) is the sole artificial testing software title.
Windows 10 Home (64-bit) with Intel INF 9.
Kingston HyperX series 240GB SATA 6Gbps solid state drive (one single NTFS partition OS + benchmarks + games)
Intel Iris Pro Graphics 6200 (Intel Graphics Driver 10.) MSI Z97 Gaming 9 AC (Intel Z97 Express chipset), BIOS version 1.CĢ x 4GB Crucial Ballistix Elite Series DDR3-1600 (CAS 9-9-9-27) Intel Core i7-5775C test system configuration Seagate 600 series 240GB SATA 6Gbps solid state drive (one single NTFS partition OS + benchmarks + games) The details of each test rig are as follows:-ĪSUS A88X-Pro (AMD A88X Bolton D4 chipset), BIOS version 2103Ģ x 4GB G.Skill RipjawsX DDR3-2133 (Auto timings: CAS 9-10-11-28) The video memory size for the Intel Iris Pro 6200 iGPU was set to the maximum of 512MB. As we had mentioned earlier, the video memory for the A10-7870K Radeon R7 graphics cores was set to the maximum of 2GB.įor the Intel Core i7-5775C, the MSI Z97 Gaming AC motherboard was fielded, and it was paired with Crucial Ballistix Elite Series DDR3-2133 RAM, which were downclocked to operate at 1600MHz (the official spec of the Intel processor). We used the automatic timings for the memory modules after we set their operating frequency to the official supported clock speed of 2,133MHz. The AMD test rig's motherboard was the ASUS A88X-Pro and we paired with G.Skill RipjawX DDR3-2133 memory modules. In terms of their dynamic boost clock speeds, the Godavari APU CPU cores peaked at 4.1GHz, which is just 100MHz higher than its Kaveri compatriot. It still uses Steamroller CPU cores, just like the older Kaveri APUs however, the A10-7870K’s CPU cores operate at a higher base clock speed of 3.9GHz, in comparison with the Kaveri A10-7850K, which has a base clock speed of 3.7GHz. Parked under the Godavari series, the APU made its appearance in June 2015. The A10-7870K APU was introduced as a breath of fresh air to the Kaveri desktop APUs. It supports DDR3 memory modules that are rated up to 1,600MHz. On the CPU side of things, this quad-core Broadwell processor has a base clock of 3.3GHz, and a Turbo Boost clock speed of 3.7GHz. In this case, it was the MSI Z97 Gaming 9 AC. However, we were only able to "reserve" up to a maximum of 512MB of video memory from the UEFI BIOS utility of our test Intel Z97 motherboard. The graphics core is stated to support up to 1.7GB of video memory, and this is shared with the system RAM. So in comparison to the Intel camp, AMD gets points for its product clarity i.e., if you need its top-end integrated graphics solution, just go for its top-end desktop A-series processor. The top-dog (as of December 2015) is the A10-7870K and its Radeon R7 integrated graphics engine. In the AMD camp, the company refreshed its A-series APUs in mid 2015 with the new Godavari APUs.
As such, the Intel Iris Pro 6200 graphics engine is the best that Intel has to offer and this is featured in the 5th generation Intel Broadwell Core i7-5775C as its host CPU. In fact, the current best graphics class of GT3e is featured in Iris Pro Graphics family. For the curious case of Intel, its most current 6th-gen Skylake desktop CPUs feature the Intel HD Graphics 530 GPU, which is far from bring the best in class.
This test (GT2) focuses more on particle and GPU simulation with higher required pixel processing throughput, whereas GPU Test 1 focuses more on lighting and geometry.For rig builders who are looking to build a desktop PC without the need for a discrete graphics card, they primarily have two current desktop processor series to choose from the AMD A-series desktop APUs or the Intel 5th or 6th generation processors. In the GPU Test 2, AMD's Radeon R7 core in the A8 is much closer to Intel's Iris Pro 5200 Graphics. In 3DMark Fire Strike, the Gigabyte BRIX Pro with the Intel Core i7-4770R once again edges out the competition but by only a small margin of 5 - 8% at most. For the following tests, we ran the benchmark in Normal mode since we were testing integrated graphics solutions, versus high-end discrete GPUs. The more taxing Extreme mode targets 1.5GB of frame buffer memory and increases detail levels across the board. GPU target frame buffer utilization for normal mode is 1GB and the benchmark uses tessellation, ambient occlusion, volume illumination, and a medium-quality depth of field filter. 3DMark Fire Strike has two benchmark modes: Normal mode runs at 1920x1080, while Extreme mode targets 2560x1440.