NVIDIA Preparing GeForce GTX 1050 Ti Max-Q Design To Tackle RX Vega M GL, Up To 2.18 TFLOPs Of Performance
Something pretty interesting popped up at Videocardz recently. NVIDIA has apparently let slip the existence of the GTX 1050 and 1050 Ti Max-Q design in their latest Linux changelog. This would mean the company is preparing to roll out the lineup soon and will be pitting it against the Kaby Lake G lineup’s RX Vega M GL. Since the name of the game is power efficiency and Max-Q is all about maximizing the thermal and power envelop, we expect it to be fairly competitive.
NVIDIA launching GTX 1050 and GTX 1050 Ti Max-Q variants soon
The change was spotted in the Linux display driver released earlier this month and lists not only the MX 130 and MX 110 (lower powered variants of the 150) but the 1050 Ti with Max-Q design as well. To those who remember Max-W is NVIDIA’s design philosophy involving constrained TDP settings. Max-Q technology has previously been used in an ultraportable gaming notebook to reduce the GPU power consumption by half.
It finds the most efficient tradeoff of performance and power for the GPU. The software balances work done on the CPU and GPU, optimizing the game settings and using advanced system design techniques for thermal management and power regulation. It also introduces WhisperMode. The new, ultra-efficient mode makes your plugged-in laptop run much quieter while gaming. It works by intelligently pacing the game’s frame rate while simultaneously configuring the graphics settings for optimal power efficiency.
The clock speed of the Max-Q is probably going to be somewhere around 1417 MHz to 1450 MHz, which translates to a theoretical graphics performance of 2.18 TFLOPs. This would put it within spitting distance of the newly launched Kaby Lake G series of graphics which house the Vega M. Keep in mind however, that while the Vega GL has a higher theoretical power, AMD and NVIDIA architectures are not directly comparable and as has been the case this generation, NVIDIA usually fares better even with lower theoretical FP32 performance.
The AMD Radeon RX Vega M GL graphics chip will be featured on a range of 8th Generation Core i7 and Core i5 processors. These feature 20 CUs which equals to 1280 stream processors, 80 texture units and 32 ROPs. The Vega 20 die is clocked at a base frequency of 931 MHz and boost frequency of 1011 MHz. These chips deliver a rated single precision output of 2.6 TFLOPs which is slightly up from a Radeon RX 560 reference design that has 2.4 TFLOPs of FP32 performance. The Radeon RX Vega 20 GPU is accompanied by 4 GB of HBM2 memory and this operates at 1.4 Gbps alongside a 1024-bit bus interface, pumping out 179.2 GB/s of bandwidth. For a single HBM package, this is lots of available bandwidth dedicated for the GPU alone.
The Max-Q design has previously been seen in the Zephyr notebooks which feature the distended keyboard and frills (which aren’t for everyone) and it remains to be seen whether this GPU will require the same style of aesthetic and cooling. If that is the case, then it could limit the total available market of the product since a lowered keyboard and the odd cooling style isn’t everyone’s cup of tea.