As we mentioned in the Tegra K1 portion of our coverage, the super mobile processor wasn’t just the only thing NVIDIA brought out as they also gave us a glimpse at the very first graphics cards that run on their latest Maxwell architecture, the GTX 750 and GTX 750 Ti.

nvidia-press-tour-geforce-gtx-750-ti-6

This is a comparison between Kepler and Maxwell's architecture

This is a comparison between Kepler and Maxwell’s architecture

Both these cards comes with a very reasonable price tags and this is because NVIDIA wanted to give gamers an affordable graphics card that still comes with all the great features available in higher-end cards such as NVIDIA ShadowPlay, G-Sync and Game Stream. Other than that, both cards are also designed to be super power efficient as the GTX 750 Ti only uses a TPD of 60W  while the GTX 750 only uses 55W. Both cards don’t even have power connectors. However, NVIDIA have said that certain partner cards that are designed for power users will feature power connectors. With Maxwell, NVIDIA might have just unlocked the secret to producing even more impressive graphics cards as Justin Walker, Senior Product Manager, GeForce, NVIDIA said “A very critical part when it comes to developing graphics processing units and graphics cards is that to advance performance, you have to advance power efficiency.”

nvidia-press-tour-geforce-gtx-750-ti-2

Yes, this is the GTX 750 Ti (the GTX 750 looks identical) and it is able to easily fit in the palm of a hand. Small yet very powerful

Both cards will share the standard ports that we see in most graphics cards today

Both cards will share the standard ports that we see in most graphics cards today

What makes this all possible is the new Streaming Multiprocessor (SM) design that improves performance per watt and performance per area. NVIDIA have made significant improvements to control logic partitioning, workload balancing, clock-gating granularity, compiler-based scheduling, number of instructions issued per clock cycle and many more enhancements that allow the Maxwell SM to exceed the Kepler SMX design. The new Maxwell SM architecture boasts 135% Performance/Core and 2x Performance/Watt compared to Kepler. Furthermore, Maxwell also boasts a much larger L2 cache design; 2048KB in GM107 compared to 256KB in GK107. With such a large cache, this means there will be fewer requests to the graphics card DRAM therefore, reducing overall board power and improving performance.

This little empty space (where the two holes are) is where the power connectors are supposed to be

This little empty space (where the two holes are) is where the power connectors are supposed to be but since this card consumes very little power, it doesn’t need one

nvidia-press-tour-geforce-gtx-750-ti-11

It uses less power but the new Maxwell cards are able to run even the latest games with close to high-end graphics settings with no problems whatsoever

Like most of NVIDIA’s Kepler-based graphics cards, the GTX 750 and GTX 750 Ti also support GPU Boost 2.0, which means that they have a minimum core clock speed of 1,020MHz. Compared t other entry-level cards, this is very high. On top of all that, the Maxwell features an improved NVENC block that provides faster encode (6-8x real-time for H.264 x 4x real-time for Kepler) and 8-10x faster decode. Also, there’s a new local decoder cache and when combined with the higher memory efficiency per stream for video decoding, power consumption is much lower when decoding videos.

On a final note regarding, the GTX 750 and GTX 750 Ti cards, both cards don’t have support for SLI but this is understandable as both cards are designed for mainstream users.

After the product presentations, we got a much closer look at G-Sync and while this photo doesn't do it justice, it's safe to say we were all very impressed

After the product presentations, we got a much closer look at G-Sync and while this photo doesn’t do it justice, it’s safe to say we were all very impressed

As mentioned above, the GTX 750 and GTX 750 Ti comes with many of NVIDIA’s staple features and this includes G-Sync. The G-Sync technology isn’t exactly new as NVIDIA announced it many months ago but what we saw in Ho Chi Minh is definitely something that all gamers should be excited for and read up as it will take gaming to a new level. G-Sync is essentially a monitor technology that is designed to address stutter, latency and tearing. Instead of waiting for a monitor to refresh, G-Sync turns things around as the GPU tells the monitor when to refresh.

This graph here shows what happens when users turn V-Sync on in a game

This graph here shows what happens when users turn V-Sync on in a game

This image shows how G-Sync eliminates things such as tearing, stutter and artifacts

This image shows how G-Sync eliminates things such as tearing, stutter and artifacts

So far, there are already four monitor manufacturers, namely ASUS, BenQ, Philips and Viewsonic have announced that they will be releasing monitors with G-Sync. For G-Sync to work, users will also have to have a GTX 600 and newer cards with a DisplayPort to utilise G-Sync. When asked about laptop displays with G-Sync, NVIDIA mentioned that it can work but so far they have no plans or announcements just yet.

 

NVIDIA's GeForce GTX 750 and GTX 750 Ti will enable gamers to build a very powerful mini-ITX rig

NVIDIA’s GeForce GTX 750 and GTX 750 Ti will enable gamers to build a very powerful mini-ITX rig

Without a doubt, the GeForce GTX 750 and 750 Ti are definitely a testament to NVIDIA’s promise of innovation. Both cards are small enough to easily fit into a mini-ITX chassis, has great power efficiency, performance and a slew of features. To top it all off, the GTX 750 has a retail price that starts at US$119 (~RM393). As for the GTX 750 Ti, the 1GB variant starts at a price of US$139 (RM460) while the 2GB variant starts at a price of US$149 (~RM492).

www.nvidia.com