Tag Archives: CUDA

Accelerated computing / GPUs

By figaro

There are plenty of sources that explain the performance per watt of a computer. However, I wanted to investigate how accelerated computer components (notably GPUs) have become more efficient at a lower price over the years. I have thus defined a metric: performance per watt per price-unit, and plotted these by launch date and launch price.
The results are as follows:


  • GFLOPS are single precision
  • Prices are in euro as they were approximately at launch date; if no launch price is known, it has not been proxied with a current price and no plot point is shown
  • Data taken mostly from Comparison of Nvidia graphics processing units – Wikipedia, the free encyclopedia and collection started from the advent of multi-core GPU architectures. The choice of Nvidia was made, because of our professional interest in deploying CUDA and does not constitute an endorsement.
  • Data is on retail components as opposed to OEM components.
  • Launch prices are often artificially high, because it is the feature set that appeals to the enthusiasts who are also the first movers. The price degradation over time (loosely 10% per year) has not been taken into account.
  • The last two plot points are the GeForce GTX Titan and GeForce GTX 650 Ti Boost.

So while performance per watt has increased more than 5-fold over the observed period, the performance per watt per price-unit has not kept up accordingly: almost 4-fold.

In fact, there is even an inverse relationship between the number of cores and the performance metric:

Perhaps the high end cards do not drop in price as much and maintain their price level at launch to finance the development of the lower end cards.

From: http://www.unix.com/high-performance-computing/221051-accelerated-computing-gpus.html

NVIDIA Has Big Plans for the Next Few Years

By Steve Symington, The Motley Fool

Filed under:

It seems an understatement to say NVIDIA has been busy so far this year.

First, at the Consumer Electronics Show in January, the graphics specialist simultaneously unveiled both its beefy Tegra 4 processor, and a curious Tegra 4-powered handheld gaming device called Project Shield.

Then, in February, NVIDIA excited the smartphone market and sent shivers up the spines of Qualcomm investors when it announced the Tegra 4i. Why? The Tegra 4i stands tall as NVIDIA‘s first fully integrated 4G LTE mobile processor, featuring an integrated version of its i500 4G LTE modem, 60 NVIDIA GPU cores, and a quad-core CPU which was jointly designed by NVIDIA and processor specialist ARM Holdings.

At the same time, as I noted last month, NVIDIA provided new details on its Chimera digital photography architecture. In addition to being integrated into its Tegra 4 and 4i platforms, Chimera should also help even further solidify NVIDIA‘s grasp on the industry by allowing photographers to capture wide-angle HDR images while the camera is moving. In addition, Chimera also lets users choose a static focal point in their scenes, helping them to maintain a properly-focused shot regardless of whether the camera or its subjects are moved. 

Let’s talk about the future
Of course, there’s plenty more activity to note, from the use of NVIDIA‘s technology in the world’s leading supercomputers to its new GRID-based virtual graphics products. That doesn’t mean, however, that NVIDIA will be content to rest on its laurels anytime soon.

To the contrary, at its 2013 GPU Technology Conference, NVIDIA just described its future plans for both “Project Logan” and “Project Parker” — or, as many are tentatively calling them already, Tegra 5 and Tegra 6.


Source: NVIDIA 2013 GPU Technology Conference.

This announcement, mind you, comes before we’ve even seen the very first Tegra 4 devices hit the market. Even so, NVIDIA management was quick to point out during its latest earnings conference call that, “at this point, [they] already have more design wins with Tegra 4 than [they] had in total with Tegra 3″ — and that’s a great thing considering analysts at Needham downgraded the stock last month while complaining of the existing Tegra line’s “limited traction in smartphones.”

Harnessing the power of Wolverine
Aside from the fact the Tegra 4i should provide plenty of incentive for smartphone makers to change that, there’s even more to love about Project Logan.

Logan will not only be integrated with NVIDIA‘s next-gen Kepler GPU and OpenGL 4.3, but also will be the first mobile processor to incorporate the company’s high-performance CUDA parallel computing platform — and yep, that’s the same CUDA utilized by NVIDIA‘s Tesla processors for supercomputing applications. This, as NVIDIA CEO Jen-Hsun Huang stated, will enable mobile devices to do “everything that a modern computer ought to do.” What’s more, NVIDIA should have its first Logan processors ready later this year, with the line “easily” entering mass production by early 2014.

Hanging with …read more
Source: FULL ARTICLE at DailyFinance