The Author
Diehard PC Gamer and Hardware Enthusiast.

Today, NVIDIA officially announces their GeForce GTX 750 Ti and GeForce GTX 750 graphics cards which are based on the first generation Maxwell core architecture. With almost twice the performance per watt of its predecessors, the NVIDIA Maxwell core architecture is going to become one of the most innovative GPU core architecture after Kepler in terms of efficiency.NVIDIA GeForce GTX 750 Ti Front

NVIDIA Launches GeForce GTX 750 Ti and GeForce GTX 750

The first generation Maxwell core would be integrated inside two graphics cards known as the GeForce GTX 750 Ti and the GeForce GTX 750. These are the first mid range cards NVIDIA has launched in their GeForce 700 series lineup since the rest of the graphics cards were based off the GK110 or the GK104 Kepler cores. The Maxwell GPU for both cards is codenamed GM107 so we will talk about it first before moving in the specifications of the cards.

The NVIDIA Maxwell GM107 core is the first of two Maxwell based chips which will be based on the 28nm process . The GM107 which is already available on the GeForce GTX 750 Ti and the GeForce GTX 750 while the other chip is codenamed GM108 and is the base Maxwell model which would be available as a part of NVIDIA’s GeForce 800M mobility lineup. The GM108 would be fused in the entry level mobility chips which include the GeForce GT 840M and possibly a few more models.

Maxwell brings new levels of power efficiency, something that NVIDIA has strive to deliver with their Kepler generation of graphics processing units. Maxwell will have approximately three Tiers basically, Tier One will consist of 200W+ Discrete GPUs. The lowest tier will have 2W and 3W SoC Tegra Parts and ultra low voltage gpus. The middle tier will cover everything in between. The first GPU will of course be the GM107 and will come in both Desktop and Mobile parts. With such a large focus put towards power efficiency, some may doubt that the performance of Maxwell generation of graphics card will not be great but its clear now that the GPU still packs a high performance ration compared to its Kepler based siblings plus the tendency to support higher clock speeds.

The NVIDIA Maxwell GM107 based graphics cards will have a TDP of just 60W but we have seen manufacturer’s adding an PCI-E power connector to add stability to their cards when in fact they won’t need one at all. Maxwell will deliver up to twice the performance per watt as compared to Kepler which seems true as witnessed on the official NVIDIA GPU roadmap presented at GTC 2013.

NVIDIA Maxwell GM107 Architecture

The NVIDIA Maxwell GM107 architecture is most clearly built from scratch yet looks like an hybrid of both Kepler and Fermi. The SMM or Streaming Multiprocessor of Maxwell will replace the SMX of Kepler and each of the smm are assembled into four blocks yet are defined as part of a single SMM which means the core architecture has got quite dense with Maxwell. Each of these blocks hold 32 CUDA cores so a single SMM with four of these operation units results in 128 CUDA cores. The SMM has 128 CUDA cores compared to 192 CUDA cores on the SMX.

The GM107 GPC ‘Graphics Processing Cluster’ consists of five of these streaming multiprocessors which are connected to a Raster Engine. Each SMM consists of Polymorph Engine 2.0 which includes the Vertex Fetch, Tessellator, Viewport transform, attribute setup and stream output. Each SMM has 8 texture mapping units which equates to 40 on the whole chip and 16 ROPs while connected to two 64-bit memory controllers.

There’s also a handful of cache on Maxwell which is 2 MB in total compared to 256 KB on Kepler GK107 which reduces GPU queuing. IPCC has been increased on the GPU core and balancing improvement to workload has been done. NVIDIA has improved their H.264 encoding and decoding with NVENC on Maxwell so that would result in better performance in ShadowPlay technology and improved sleep states have been implemented to reduce power input while the GPU is running idle. All of this is packed in a die which measures 148mm2  which means transistor density has been upped by 15% on the same 28nm process design.

Advertisements

NVIDIA First Generation Maxwell Core – Block Diagram:NVIDIA-Maxwell-GM107-block-diagram-635x557

NVIDIA GeForce GTX 750 Ti

The NVIDIA GeForce GTX 750 Ti will be the top most GM107 graphics card featuring 640 CUDA Cores, 40 TMUs, 16 ROPs and a 2/1 GB GDDR5 memory operating along a 128-bit interface. The clock speeds would be maintained at 1020 MHz for core, 1085 MHz boost and 1400 MHz memory clock along a 128-bit interface with an effective memory clock of 5.4 GHz. The card would ship in both 1 GB and 2 GB GDDR5 variants with display outputs that include 2 DL-DVI, mini HDMI and a Display port 1.2. Total TDP of the card is rated at 60W so no power pins are required for the card to function yet some manufacture’s would add on their non-reference cards just to add stability. The GeForce GTX 750 Ti would launch on 18th February 2014 for a price around $149 US.

NVIDIA GeForce GTX 750

The NVIDIA GeForce GTX 750 (Non-Ti) is also based on the GM107 core architecture featuring 512 CUDA Cores, 32 TMUs, 16 ROPs and a 1 GB GDDR5 memory interface. The clock speeds for this graphics card are configured at the same 1020 MHz core and 1085 MHz boost but the memory runs slower at 5.0 GHz effective clock speed (1250 MHz QDR). The GeForce GTX 750 has an even lower TDP of 55W which makes it an highly efficient card for its price range. The GeForce GTX 750 is also planned for launch on 18th February and would stick to the $119 US pricing.GeForce GTX 750 Series Front

NVIDIA GeForce GTX 750 and GeForce GTX 750 Ti  Reviews:

Note - Reviews will be live in less than 6 hours so stay tuned for this panel to be updated with links from various tech sites. The first review is up folks, Tweak Town wins the race.

Official Page – NVIDIA GeForce GTX 750 Ti / NVIDIA GeForce GTX 750

NVIDIA GeForce GTX 750 and GeForce GTX 750 Ti  Specifications:

NVIDIA GeForce GTX 750 Ti NVIDIA GeForce GTX 650 Ti NVIDIA GeForce GTX 750 NVIDIA GeForce GTX 650
GPU Codename Maxwell GM107 Kepler GK106 Maxwell GM107 Kepler GK107
GPU Process 28nm 28nm 28nm 28nm
Die Size 148mm2 221mm2 148mm2 118mm2
Transistors 1.87 Billion 2.54 Billion 1.87 Billion 1.3 Billion
Cores 640 CUDA 768 CUDA 512 CUDA 384 CUDA
Core Clock 1020 MHz 928 MHz 1020 MHz 1058 MHz
Boost Clock 1085 MHz N/A 1085 MHz N/A
Memory 2 GB GDDR5 1 GB GDDR5 1 GB GDDR5 1 GB GDDR5
Memory Clock 5.4 GHz 5.4 GHz 5.0 GHz 5.0 GHz
Memory Bandwidth 86.4 GB/s 86.4 GB/s 80 GB/s 80 GB/s
Power Connectors None 6-Pin None 6-Pin
TDP 60W 110W 55W 64W
Launch 18th February 2014 2013 18th February 2014 2013
Price $149 US $149 US $119 US $119 US

NVIDIA GeForce GTX 750 and GeForce GTX 750 Ti Performance Chart:

GeForce GTX 750 Series Performance

 

NVIDIA Maxwell Architectural and Presentation Slides:NVIDIA Maxwell Power efficency NVIDIA Maxwell Power Consumption NVIDIA Maxwell Performance Per Watt NVIDIA Maxwell SMM