Maxwell and Kepler and...Fermi?
Covering the landscape of mobile GPUs can be a harrowing experience. Brands, specifications, performance, features and architectures can all vary from product to product, even inside the same family. Rebranding is rampant from both AMD and NVIDIA and, in general, we are met with one of the most confusing segments of the PC hardware market.
Today, with the release of the GeForce GTX 800M series from NVIDIA, we are getting all of the above in one form or another. We will also see performance improvements and the introduction of the new Maxwell architecture (in a few parts at least). Along with the GeForce GTX 800M parts, you will also find the GeForce 840M, 830M and 820M offerings at lower performance, wattage and price levels.
With some new hardware comes a collection of new software for mobile users, including the innovative Battery Boost that can increase unplugged gaming time by using frame rate limiting and other "magic" bits that NVIDIA isn't talking about yet. ShadowPlay and GameStream also find their way to mobile GeForce users as well.
Let's take a quick look at the new hardware specifications.
|GTX 880M||GTX 780M||GTX 870M||GTX 770M|
|GPU Code name||Kepler||Kepler||Kepler||Kepler|
|Rated Clock||954 MHz||823 MHz||941 MHz||811 MHz|
|Memory||Up to 4GB||Up to 4GB||Up to 3GB||Up to 3GB|
|Memory Clock||5000 MHz||5000 MHz||5000 MHz||4000 MHz|
Both the GTX 880M and the GTX 870M are based on Kepler, keeping the same basic feature set and hardware specifications of their brethren in the GTX 700M line. However, while the GTX 880M has the same CUDA core count as the 780M, the same cannot be said of the GTX 870M. Moving from the GTX 770M to the 870M sees a significant 40% increase in core count as well as a jump in clock speed from 811 MHz (plus Boost) to 941 MHz.
Subject: Graphics Cards | March 3, 2014 - 10:33 PM | Tim Verry
Tagged: nvidia, maxwell, gtx 750, evga
EVGA recently launched two new GTX 750 graphics cards with 2GB of GDDR5 memory. The new cards include a reference clocked GTX 750 2GB and a factory overclocked GTX 750 2GB SC (Super Clocked).
The new graphics cards are based around NVIDIA’s GTX 750 GPU with 512 Maxwell architecture CUDA cores. The GTX 750 is the little brother to the GTX 750 Ti we recently reviewed which has 640 cores. EVGA has clocked the GTX 750 2GB card’s GPU at reference clockspeeds of 1020 MHz base and 1085 MHz boost and memory at a reference speed of 1253 MHz. The “Super Clocked” GTX 750 2GB SC card keeps the memory at reference speeds but overclocks the GPU quite a bit to 1215 MHz base and 1294 MHz boost.
|EVGA GTX 750 2GB||EVGA GTX 750 2GB Super Clocked|
|GPU||512 CUDA Cores (Maxwell)||512 CUDA Cores (Maxwell)|
|- GPU Base||1020 MHz||1215 MHz|
|- GPU Boost||1085 MHz||1294 MHz|
|Memory||2 GB GDDR5 @ 1253 MHz on 128-bit bus|
1 x DVI, 1 x HDMI, 1 x DP
Both cards have a 55W TDP sans any PCI-E power connector and utilize a single shrouded fan heatsink. The cards are short but occupy two PCI slots. The rear panel hosts one DVI, one HDMI, and one DisplayPort video output along with ventilation slots for the HSF. Further, the cards both support NVIDIA’s G-Sync technology.
The reference clocked GTX 750 2GB is $129.99 while the factory overclocked model is $139.99. Both cards are similar to their respective predecessors except for the additional 1GB of GDDR5 memory which comes at a $10 premium and should will help a bit at high resolutions.
Subject: General Tech, Graphics Cards | March 2, 2014 - 05:20 PM | Scott Michaud
Tagged: passive cooling, maxwell, gtx 750 ti
The NVIDIA GeForce GTX 750 Ti is fast but also power efficient, enough-so that Ryan found it a worthwhile upgrade for cheap desktops with cheap power supplies that were never intended for discrete graphics. Of course, this recommendation is about making the best of what you got; better options probably exist if you are building a PC (or getting one built by a friend or a computer store).
Image Credit: Tom's Hardware
Tom's Hardware went another route: make it fanless.
After wrecking a passively-cooled Radeon HD 7750, which is probably a crime in Texas, they clamped it on to the Maxwell-based GTX 750 Ti. While the cooler was designed for good airflow, they decided to leave it in a completely-enclosed case without fans. Under load, the card reached 80 C within about twenty minutes. The driver backed off performance slightly, 1-3% depending on your frame of reference, but was able to maintain that target temperature.
Now, if only it accepted SLi, this person might be happy.
Subject: General Tech, Graphics Cards | February 20, 2014 - 05:45 PM | Ken Addison
Tagged: nvidia, mining, maxwell, litecoin, gtx 750 ti, geforce, dogecoin, coin, bitcoin, altcoin
As we have talked about on several different occasions, Altcoin mining (anything that is NOT Bitcoin specifically) is a force on the current GPU market whether we like it or not. Traditionally, Miners have only bought AMD-based GPUs, due to the performance advantage when compared to their NVIDIA competition. However, with continued development of the cudaMiner application over the past few months, NVIDIA cards have been gaining performance in Scrypt mining.
The biggest performance change we've seen yet has come with a new version of cudaMiner released yesterday. This new version (2014-02-18) brings initial support for the Maxwell architecture, which was just released yesterday in the GTX 750 and 750 Ti. With support for Maxwell, mining starts to become a more compelling option with this new NVIDIA GPU.
With the new version of cudaMiner on the reference version of the GTX 750 Ti, we were able to achieve a hashrate of 263 KH/s, impressive when you compare it to the performance of the previous generation, Kepler-based GTX 650 Ti, which tops out at about 150KH/s or so.
As you may know from our full GTX 750 Ti Review, the GM107 overclocks very well. We were able to push our sample to the highest offset configurable of +135 MHz, with an additional 500 MHz added to the memory frequency, and 31 mV bump to the voltage offset. All of this combined to a ~1200 MHz clockspeed while mining, and an additional 40 KH/s or so of performance, bringing us to just under 300KH/s with the 750 Ti.
As we compare the performance of the 750 Ti to AMD GPUs and previous generation NVIDIA GPUs, we start to see how impressive the performance of this card stacks up considering the $150 MSRP. For less than half the price of the GTX 770, and roughly the same price as a R7 260X, you can achieve the same performance.
When we look at power consumption based on the TDP of each card, this comparison only becomes more impressive. At 60W, there is no card that comes close to the performance of the 750 Ti when mining. This means you will spend less to run a 750 Ti than a R7 260X or GTX 770 for roughly the same hash rate.
Taking a look at the performance per dollar ratings of these graphics cards, we see the two top performers are the AMD R7 260X and our overclocked GTX 750 Ti.
However, when looking at the performance per watt differences of the field, the GTX 750 Ti looks more impressive. While most miners may think they don't care about power draw, it can help your bottom line. By being able to buy a smaller, less efficient power supply the payoff date for the hardware is moved up. This also bodes well for future Maxwell based graphics cards that we will likely see released later in 2014.
Subject: General Tech | February 20, 2014 - 02:17 PM | Ken Addison
Tagged: podcast, video, toshiba, raptr, R9 290X, r9 290, pcper, OEM, maxwell, gtx 750 ti, desktop pc, 750 ti, 5TB
PC Perspective Podcast #288 - 02/20/2014
Join us this week as we discuss the release of the NVIDIA GTX 750 Ti, Upgrading Crappy Desktops, 5TB Hard Drives and more!
The URL for the podcast is: http://pcper.com/podcast - Share with your friends!
- iTunes - Subscribe to the podcast directly through the Store
- RSS - Subscribe through your regular RSS reader
- MP3 - Direct download link to the MP3 file
Hosts: Ryan Shrout, Jeremy Hellstrom, Josh Walrath and Allyn Malventano
Week in Review:
News items of interest:
Hardware/Software Picks of the Week:
Subject: Graphics Cards | February 19, 2014 - 04:43 PM | Jeremy Hellstrom
Tagged: geforce, gm107, gpu, graphics, gtx 750 ti, maxwell, nvidia, video
We finally saw Maxwell yesterday, with a new design for the SMs called SMM each of which consist of four blocks of 32 dedicated, non-shared CUDA cores. In theory that should allow NVIDIA to pack more SMMs onto the card than they could with the previous SMK units. This new design was released on a $150 card which means we don't really get to see what this new design is capable of yet. At that price it competes with AMD's R7 260X and R7 265, at least if you can find them at their MSRP and not at inflated cryptocurrency levels. Legit Reviews contrasted the performance of two overclocked GTX 750 Ti to those two cards as well as to the previous generation GTX 650Ti Boost on a wide selection of games to see how it stacks up performance-wise which you can read here.
That is of course after you read Ryan's full review.
"NVIDIA today announced the new GeForce GTX 750 Ti and GTX 750 video cards, which are very interesting to use as they are the first cards based on NVIDIA's new Maxwell graphics architecture. NVIDIA has been developing Maxwell for a number of years and have decided to launch entry-level discrete graphics cards with the new technology first in the $119 to $149 price range. NVIDIA heavily focused on performance per watt with Maxwell and it clearly shows as the GeForce GTX 750 Ti 2GB video card measures just 5.7-inches in length with a tiny heatsink and doesn't require any internal power connectors!"
Here are some more Graphics Card articles from around the web:
- MSI GTX 750 Ti Gaming Video Card Review @HiTech Legion
- NVIDIA GeForce GTX 750 Ti @ Benchmark Reviews
- ASUS GTX 750 OC 1 GB @ techPowerUp
- MSI GTX 750 Ti Gaming 2 GB @ techPowerUp
- NVIDIA GeForce GTX 750Ti the Arrival of Maxwell @HiTech Legion
- Palit GTX 750 Ti StormX Dual 2 GB @ techPowerUp
- The GTX 750 Ti Review; Maxwell Arrives @ Hardware Canucks
- Nvidia GeForce GTX 750 Ti vs. AMD Radeon R7 265 @ Legion Hardware
- MSI GTX750Ti OC Twin Frozr @ Kitguru
- NVIDIA GeForce GTX 750 Ti 2 GB @ techPowerUp
- NVIDIA GeForce GTX 750 Ti "Maxwell" On Linux @ Phoronix
- A quick look at Mantle on AMD's Kaveri APU @ The Tech Report
- Sapphire Radeon R9 Tri-X OC video card @ Hardwareoverclock
- AMD Radeon R9 290: Still Not Good For Linux Users @ Phoronix
- AMD Radeon R7 265 2GB Video Card Review @ Legit Reviews
- Sapphire Radeon R7 260X OC 2GB Graphics Card Review @ Techgage
- XFX Double Dissipation R9 280X @ [H]ard|OCP
An Upgrade Project
When NVIDIA started talking to us about the new GeForce GTX 750 Ti graphics card, one of the key points they emphasized was the potential use for this first-generation Maxwell GPU to be used in the upgrade process of smaller form factor or OEM PCs. Without the need for an external power connector, the GTX 750 Ti provided a clear performance delta from integrated graphics with minimal cost and minimal power consumption, so the story went.
Eager to put this theory to the test, we decided to put together a project looking at the upgrade potential of off the shelf OEM computers purchased locally. A quick trip down the road to Best Buy revealed a PC sales section that was dominated by laptops and all-in-ones, but with quite a few "tower" style desktop computers available as well. We purchased three different machines, each at a different price point, and with different primary processor configurations.
The lucky winners included a Gateway DX4885, an ASUS M11BB, and a Lenovo H520.
What we know about Maxwell
I'm going to go out on a limb and guess that many of you reading this review would not have normally been as interested in the launch of the GeForce GTX 750 Ti if a specific word hadn't been mentioned in the title: Maxwell. It's true, the launch of GTX 750 Ti, a mainstream graphics card that will sit in the $149 price point, marks the first public release of the new NVIDIA GPU architecture code named Maxwell. It is a unique move for the company to start at this particular point with a new design, but as you'll see in the changes to the architecture as well as the limitations, it all makes a certain bit of sense.
For those of you that don't really care about the underlying magic that makes the GTX 750 Ti possible, you can skip this page and jump right to the details of the new card itself. There I will detail the product specifications, performance comparison and expectations, etc.
If you are interested in learning what makes Maxwell tick, keep reading below.
The NVIDIA Maxwell Architecture
When NVIDIA first approached us about the GTX 750 Ti they were very light on details about the GPU that was powering it. Even though the fact it was built on Maxwell was confirmed the company hadn't yet determined if it was going to do a full architecture deep dive with the press. In the end they went somewhere in between the full detail we are used to getting with a new GPU design and the original, passive stance. It looks like we'll have to wait for the enthusiast GPU class release to really get the full story but I think the details we have now paint the story quite clearly.
During the course of design the Kepler architecture, and then implementing it with the Tegra line in the form of the Tegra K1, NVIDIA's engineering team developed a better sense of how to improve the performance and efficiency of the basic compute design. Kepler was a huge leap forward compared to the likes of Fermi and Maxwell is promising to be equally as revolutionary. NVIDIA wanted to address both GPU power consumption as well as finding ways to extract more performance from the architecture at the same power levels.
The logic of the GPU design remains similar to Kepler. There is a Graphics Processing Cluster (GPC) that houses Simultaneous Multiprocessors (SM) built from a large number of CUDA cores (stream processors).
GM107 Block Diagram
Readers familiar with the look of Kepler GPUs will instantly see changes in the organization of the various blocks of Maxwell. There are more divisions, more groupings and fewer CUDA cores "per block" than before. As it turns out, this reorganization was part of the ability for NVIDIA to improve performance and power efficiency with the new GPU.
It wouldn’t be February if we didn’t hear the Q4 FY14 earnings from NVIDIA! NVIDIA does have a slightly odd way of expressing their quarters, but in the end it is all semantics. They are not in fact living in the future, but I bet their product managers wish they could peer into the actual Q4 2014. No, the whole FY14 thing relates back to when they made their IPO and how they started reporting. To us mere mortals, Q4 FY14 actually represents Q4 2013. Clear as mud? Lord love the Securities and Exchange Commission and their rules.
The past quarter was a pretty good one for NVIDIA. They came away with $1.144 billion in gross revenue and had a GAAP net income of $147 million. This beat the Street’s estimate by a pretty large margin. As a response, trading of NVIDIA’s stock has gone up in after hours. This has certainly been a trying year for NVIDIA and the PC market in general, but they seem to have come out on top.
NVIDIA beat estimates primarily on the strength of the PC graphics division. Many were focusing on the apparent decline of the PC market and assumed that NVIDIA would be dragged down by lower shipments. On the contrary, it seems as though the gaming market and add-in sales on the PC helped to solidify NVIDIA’s quarter. We can look at a number of factors that likely contributed to this uptick for NVIDIA.
Subject: General Tech, Graphics Cards | January 20, 2014 - 04:19 AM | Scott Michaud
Tagged: maxwell, nvidia
Well this is somewhat unexpected (and possibly wrong). Maxwell, NVIDIA's new architecture to replace Kepler, is said to appear in Feburary with the form of a GeForce GTX 750 Ti. The rumors, which sound iffy to me, claims that this core will be produced at TSMC on a 28nm fabrication technology and later transition to their 20nm lines.
As if the 700-series family tree was not diverse enough.
2013 may have been much closer than expected.
Swedish site, Sweclockers, have been contacted by "sources" which claim that NVIDIA has already alerted partners to prepare a graphics card launch. Very little information is given beyond that. They do not even have access to a suggested GM1## architecture code. They just claim that partners should expect a new videocard on the 18th of February (what type of launch that is is also unclear).
This also raises questions about why the mid-range card will come before the high-end. If the 28nm rumor is true, it could just be that NVIDIA did not want to wait around until TSMC could fabricate their high-end part if they already had an architecture version that could be produced now. It could be as simple as that.
The GeForce GTX 750 Ti is rumored to arrive in February to replace the GTX 650 Ti Boost.