Intel Iris Plus Graphics 640 vs NVIDIA Quadro P2000 Max-Q vs NVIDIA Quadro P3200

Intel Iris Plus Graphics 640

The Intel Iris Plus Graphics 640 (GT3e) is a processor graphics card for Kaby Lake models announced in January 2017. As the successor to the Intel Iris Graphics 540 (Skylake), the Iris Plus Graphics 640 is used for 15-Watt ULV models and is equipped with 64 MB eDRAM cache. There should not be any big changes compared to the Iris Graphics 540.

The so called GT3e version of the Kaby Lake GPU still has 48 Execution Units (EUs), which can reach up to 1050 MHz depending on the model. Besides eDRAM cache, the Iris 640 can also access the system memory (2x 64 Bit DDR3L-1600/LPDDR3-1866/DDR4-2400) via processor interface.

Compared to the Iris Plus 650 from the 28-Watt models, the Iris 640 only has a slightly lower maximum clock as well as the lower TDP, so the utilization of the Turbo Boost potential should be worse.

Performance

The exact performance of the Iris Plus Graphics 640 depends on the CPU model, because maximum clock as well as the size of the L3 cache can differ a bit. The system memory (DDR3/DDR4) will influence the performance as well.

The fastest chips are high-clocked Core i7 models like the Core i7-7660U. Depending on the game, the Iris Plus 640 will probably be similar to the previous model somewhere between a dedicated GeForce 920M and GeForce 940M and is usually limited to low settings in modern games.

Features

The reworked video engine now fully supports hardware decoding of H.265/HEVC videos. Contrary to Skylake, however, Kaby Lake can now also decode H.265/HEVC Main 10 with a 10-bit color depth as well as Google's VP9 codec. The video output is possible via DP 1.2/eDP 1.3 (up to 3840 x 2160 @60 Hz), whereas HDMI is also supported in the older 1.4a standard. An HDMI 2.0 output can be added via converter from DisplayPort. The GPU can drive up to three displays simultaneously.

Power Consumption

The Iris Plus Graphics 640 is used for 15-Watt ULV processor and therefore thin notebooks and Ultrabooks.

NVIDIA Quadro P2000 Max-Q

The Nvidia Quadro P2000 is a mobile mid-range workstation graphics card for notebooks. Similar to the consumer GeForce GTX 1050 Ti (Laptop), it is based on the GP107 chip with 768 shaders. Compared to the normal P2000, the Max-Q variant offers a lower clock speed at a reduced power consumption. The graphics card is designed for the Kaby Lake / Coffee Lake generation.

The Quadro GPUs offer certified drivers, which are optimized for stability and performance in professional applications (CAD, DCC, medical, prospection, and visualizing applications). The performance in these areas is therefore much better compared to corresponding consumer GPUs.

Performance

The theoretical performance should be similar with the GeForce GTX 1050 Ti Max-Q.

NVIDIA Quadro P3200

The Nvidia Quadro P3200 is a mobile high-end workstation graphics card for notebooks. As the slightly faster Quadro P4000, it is based on the GP104 chip and uses 1792 CUDA cores. The memory bus however is slightly narrower at 192 versus 256 Bit. The clock rate should be similar, as Nvidia also specifies 5.3 TFLOPs of peak performance.

The Quadro GPUs offer certified drivers, which are optimized for stability and performance in professional applications (CAD, DCC, medical, prospection, and visualizing applications). The performance in these areas is therefore much better compared to corresponding consumer GPUs.

Performance

The performance should be slightly below the older Quadro P4000, due to the same peak performance of the shaders but a slower memory bus. Compared to the old Quadro M3000M, the P3200 should offer a 93% better overall performance in SpecViewPerf 12.

Power Consumption

The power consumption of the Quadro P3200 is 75 Watts TGP (including memory)  and therefore similar to the Quadro P3000.