NVIDIA Tegra 4 Specifications Sheet Leaks

Subject: General Tech, Mobile | April 7, 2012 - 07:11 PM |
Tagged: tegra 4, tegra, SoC, nvidia, mobile

The Chinese language VR-Zone website has allegedly managed to get their hands on a leaked specifications sheet for NVIDIA’s upcoming Tegra 4 System-on-a-chip (SoC) aimed at mobile tablets. Codenamed “Wayne,” the new SoC will come in several flavors and will arrive next year.

The upcoming chips will have 10x the performance of NVIDIA’s original Tegra and five times the performance of the current generation Kal-El Tegra 3 chip. NVIDIA has run into several hurdles in integrating an LTE cell radio into their SoCs, but if the leaked document is true, the company will finally release a Tegra chp with built-in LTE 100 and HSPA42 cell radio capabilities as early as the third quarter of 2013.

Badge_Tegra_3D_large.jpg

Further, the Tegra 4 SoCs will come in four flavors: T40, T43, AP40, and SP3X. T40 will represent the first Tegra 4 chp that manufacturers and consumers will be able to get their hands on -- as early as Q1 2013. It is a quad core part with one companion core and will run at 1.8 GHz. T43 is an evolution of the T40 and will bump up the clockspeed to 2.0 GHz. The AP40 chip will be the first budget Tegra 4 processor and will run anywhere between 1.2 GHz and 1.8 GHz. The T43 and AP40 SoCs are reportedly coming out in Q3 2013. All three chips -- The T40, T43, and AP40 -- are based on the ARM Cortex A15 architecture.

  T40 T43 AP40 SP3X
Release Date Q1 2013 Q3 2013 Q3 2013 Q3 2013
Markets Aimed At Flagship Flagship Mainstream Mainstream
Tablet Device Screen Size 10" 10" 10" 7"
Processor Clockspeed 1.8 GHz 2.0 GHz 1.2-1.8 GHz 1.2-2.0 GHz
Core Count 4+1 4+1 4+1 4+1
Chip Architecture A15 A15 A15 A9
Cell Radio       LTE100/HSPA42

 

The final Tegra 4 chip is called SP3X, and it will arrive in Q3 2013. Aimed at mainstream tablets with 7” or smaller screens, the upcoming SoC will feature LTE support and will have a clockspeed of 1.2 GHz to 2.0 GHz. It is a quad core (plus one companion core) part but is reportedly based on the ARM Cortex A9 architecture.  The leaked release dates do seem to be in line with earlier reports, though they should still be taken with your daily dose of salt.

 

Right now Tegra delivers on performance and many high end mobile devices have incorporated the NVIDIA chip. Even so, they still have very little market share, and the two mainstream Tegra 4 chips -- especially the SP3X with LTE radio -- should help them make inroads against Qualcomm and Samsung who hold a great deal of market share.

Overclocked Gigabyte WindForce GTX 680 GPU Pictured

Subject: Graphics Cards | April 5, 2012 - 11:19 PM |
Tagged: windforce, overclock, nvidia, gtx 680, gpu, gigabyte, custom gtx 680

Popular motherboard manufacturer Gigabyte is the latest company to debut a custom version of the NVIDIA GTX 680 reference graphics card. Gigabyte’s unique take on the GTX 680 starts off with a custom dark blue PCB and ripping out the puny two six pin PCI-E power connectors. They are then replaced with one eight pin and one six pin PCI-E power connector. Then, they top it off with a custom three fan cooler. The heatsink uses three copper heatpipes with direct contact with the GPU, and two arrays of aluminum fins.

P1050203.jpg

The cooler and blue PCB via VR-Zone

The extra power provided by the eight pin PCI-E connector allows for potentially higher overclocks (depending on the particular chips), and the custom cooler keeps the overclocked card nice and cool. In fact, Gigabyte is shipping the card with a factory overclock. Although they did not overclock the 2 GB of GDDR5 memory from stock, they have set the base clock frequency and boost frequency at 1071 MHz and 1124 MHz boost respectively. Compared to the reference specs of 1006 MHz base and 1058 MHz boost, that amounts to a respectable 65 MHz base overclock and 66 MHz boost overclock out of the box. Further, depending on the chip, they may be capable of overclocking much higher.

Gigabyte WindForce 3x2GB GTX 680.jpg

The assembled card showing the video outputs via Guru3D

So long as you can find one in stock, the NVIDIA GTX 680 GPU is shaping up to be an interesting card, especially the custom versions! More photos of the previewed Gigabyte GTX 680 WindForce edition is available here and here.

Source: Guru3D

NVIDIA Introduces Two New Rebranded 600 Series Cards

Subject: Graphics Cards | April 5, 2012 - 08:37 PM |
Tagged: rebranded gpu, nvidia, gtx 620, gtx 605, gpu, fermi

NVIDIA is continuing the infuriating practice (though they aren't alone in doing so) of re-branding older graphics cards into the following generation to present “new” cards (or to confuse consumers and tech journalists to no end, though I suspect that’s just a side benefit). Specifically, they are taking two lower tier desktop OEM cards and rebranding them as 600 series "Kepler" cards. The NVIDIA GT 520 and GT 510 will be renamed the NVIDIA GT 620 and GT 605 respectively. Even more confusing is that the “new” cards will have less hardware, and the only addition is the support for the OpenGL 4.2 standard (versus 4.1 on the 520 and 510).  Other than that, they are two Fermi based cards in Kepler clothing.

GeForce-GT-520-oem-3qtr-1000x580_gallery_preview.png

The NVIDIA GT 620 replaces the GT 520 and features half of the graphics memory as the 500 series card, meaning that users will get 512 MB or 1 GB on the 620 instead of the 1 GB / 2 GB options of the GT 520. The card still features VGA, DVI, and HDMI video outputs. The remaining specifications can be seen in the chart below. Despite halving the memory, the new card has a very slightly higher TDP at 30 watts versus the rated 29 watts of the GT 520.

GeForce-GT-520-oem-F-1000x580_gallery_preview.png

On the other hand, the NVIDIA GT 605 is the new version of the GT 510. The 600 series part also halves the amount of memory of the GT 510 counterpart with 512 MB and 1 GB versions compared to 1 GB and 2 GB versions of the GT 510. The GT 605 also has VGA, DVI, and HDMI ports. It is rated at a TDP of 25 watts like the GT 510.

  GT 510 GT 605 GT 520 GT 620
Graphics Clock 523 MHz 523 MHz 810 MHz 810 MHz
Processor Clock 1046 MHz 1046 MHz 1620 MHz 1620 MHz
Memory Clock (up to) 898 MHz 898 MHz 898 MHz 898 MHz
Memory (DDR3) 1 or 2 GB 512 or 1024 MB 1 or 2 GB 512 or 1024 MB
OpenGL 4.1 4.2 4.1 4.2
Maximum Resolution 2560x1600 2560x1600 2560x1600 2560x1600

Reference GT 500 and GT 600 Series Specifications (changes in bold).

According to Tom’s, the “new” cards are still Fermi based despite the new implied Kepler generation naming scheme. Granted, these are OEM cards but it still is a bit dishonest to rebrand them, especially in the case of the GT 620 where it is the same rank but with the Kepler generation digit at the beginning. There have been some comments around the Internet that the two new rebranded cards were brought into play to allow OEMs to sell PCs with new 600 series discrete graphics. At this level, it really doesn’t matter per se as they will still do HTPC and desktop graphics well enough and are not going to be purchased by customers directly, but it’s still annoying (heh). What do you guys think about the graphics card rebranding in general, whether it’s on the desktop or mobile market?

Source: NVIDIA

NVIDIA urges you to program better now, not CPU -- later.

Subject: Editorial, General Tech, Graphics Cards, Processors | April 4, 2012 - 04:13 AM |
Tagged: nvidia, Intel, Knight's Corner, gpgpu

NVIDIA steals Intel’s lunch… analogy. In the process they claim that optimizing your application for Intel’s upcoming many-core hardware is not free of effort, and that effort is similar to what is required to develop on what NVIDIA already has available.

A few months ago, Intel published an article on their software blog to urge developers to look to the future without relying on the future when they design their applications. The crux of Intel’s argument states that regardless of how efficient Intel makes their processors, there is still responsibility on your part to create efficient code.

nvidiainteltf2.jpg

There’s always that one, in the back of the class…

NVIDIA, never a company to be afraid to make a statement, used Intel’s analogy to alert developers to optimize for many-core architectures.

The hope that unmodified HPC applications will work well on MIC with just a recompile is not really credible, nor is talking about ease of programming without consideration of performance.

There is no free lunch. Programmers will need to put in some effort to structure their applications for hybrid architectures. But that work will pay off handsomely for today’s, and especially tomorrow’s, HPC systems.

It remains to be seen how Intel MIC will perform when it eventually arrives. But why wait? Better to get ahead of the game by starting down the hybrid multicore path now.

NVIDIA thinks that Intel was correct: there would be no free lunch for developers, why not purchase a plate at NVIDIA’s table? Who knows, after the appetizer you might want to stay around.

You cannot simply allow your program to execute on Many Integrated Core (MIC) hardware and expect it to do so well. The goal is not to simply implement on new hardware -- it is to perform efficiently while utilizing the advantages of everything that is available. It will always be up to the developer to set up their application in the appropriate way.

Your advantage will be to understand the pros and cons of massive parallelism. NVIDIA, AMD, and now Intel have labored to create a variety of architectures to suit this aspiration; software developers must labor in a similar way on their end.

Source: NVIDIA Blogs

The only thing better than an in stock GTX 680 would be two of them

Subject: Graphics Cards | March 29, 2012 - 06:18 PM |
Tagged: sli, nvidia, gtx 680

The name of the game in [H]ard|OCP's latest review is scaling at 5760 x 1200, specifically the scaling of two GTX 680s in SLI as well as a pair of HD 7970s in Crossfire.  Some games like Mass Effect 3 will not benefit much as the difference between 150fps and 170fps will be hard to do but others such as Battlefield 3 and Arkham City stress these cards somewhat at this resolution, but even 50fps is rather impressive when pushing about 7 million pixels.  Read on and be prepared to feel a little jealous, maybe jealous enough to snatch up a Galaxy model which is back in stock. (didn't last 2 minutes)

HgotSLI.jpg

"We've got two GeForce GTX 680 video cards to test SLI performance against Radeon HD 7970 CrossFireX. Will these less expensive GTX 680 video cards offer a better gameplay experience or choke at high resolutions due to a smaller VRAM footprint? We will prove to you which solution offers better efficiency and performance."

Here are some more Graphics Card articles from around the web:

Graphics Cards

 

Source: [H]ard|OCP

Podcast #195 - GTX 680 Review, the MSI HD 7970 Lightning, and a 4GB GTX 680!

Subject: General Tech | March 29, 2012 - 03:29 PM |
Tagged: ssd, podcast, nvidia, Intel, gtx680, amd, 7970, 680

PC Perspective Podcast #195 - 03/29/2012

Join us this week as we talk about our GTX 680 Review, the MSI HD 7970 Lightning, and a 4GB GTX 680!

You can subscribe to us through iTunes and you can still access it directly through the RSS page HERE.

The URL for the podcast is: http://pcper.com/podcast - Share with your friends!

  • iTunes - Subscribe to the podcast directly through the iTunes 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 Malvantano

This Podcast is brought to you by MSI Computer, and their all new Sandy Bridge Motherboards!

Program length: 1:00:26

Program Schedule: 

  1. Introduction
  2. 1-888-38-PCPER or podcast@pcper.com
  3. http://pcper.com/podcast
  4. http://twitter.com/ryanshrout and http://twitter.com/pcper
  5. Gigabyte GA-X79-UD5 LGA 2011 EATX Motherboard Review
  6. NVIDIA GeForce GTX 680 2GB Graphics Card Review - Kepler in Motion
  7. Acer Aspire Timeline Ultra M3 Review: Kepler's First Laptop
    1. PC Perspective Live Review Recap: NVIDIA GeForce GTX 680
  8. MSI R7970 Lightning Review: AMD's HD 7970 Gets the Treatment
  9. This Podcast is brought to you by MSI Computer, and their all new Sandy Bridge Motherboards!
  10. Galaxy Readying 4GB and Hall Of Fame Edition GTX 680 GPUs
  11. About that pricing AMD; you sure you want to stick with it?
  12. Super Talent Releases New RAIDDrive upStream PCI-E SSD
  13. Hardware / Software Pick of the Week
    1. Ryan: is lame and didnt have one
    2. Jeremy: I got a better deal but this is close
    3. Josh: Two SSDs for the price of one 
    4. Allyn: Khan Academy
  14. 1-888-38-PCPER or podcast@pcper.com
  15. http://pcper.com/podcast   
  16. http://twitter.com/ryanshrout and http://twitter.com/pcper
  17. Closing

Source:

Digital Storm Shows Off Custom Water Cooled Aventum PC

Subject: Systems | March 26, 2012 - 04:06 PM |
Tagged: water cooling, nvidia, Ivy Bridge, gtx 680, Digital Storm

Digital Storm, a custom PC Manufacturer founded in 2002 today revealed their latest system lineup. The new Aventum computers employ the company’s Cryo-TEC sub-zero cooling solution and the latest in PC hardware in a custom full tower chassis. The custom Aventum systems come in several tiers, including three systems with Intel Sandy Bridge-E processors, NVIDIA GTX 680 graphics cards, solid state drives, and at least 16 GB of RAM. Digital Storm further does not skimp on the power supplies. The Aventum computers are powered by either Corsair or Silverstone PSUs.

01-Featured-Image.jpg

The hardware inside the chassis is impressive from a performance standpoint, and Digital Storm is including high end hardware as part of several tiers. The lowest tier is an Intel Sandy Bridge Core i7 2700K and a single EVGA NVIDIA GTX 680 graphics card on an Asus P8Z68-V Pro motherboard. On the other hand, the top tier system moves up to a dual socket EVGA SR-X motherboard, two Intel Xeon E5-2630 processors and three EVGA NVIDIA GTX 680 GPUs in a triple SLI configuration. The other hardware differences are less pronounced - like the upgrade to faster or more RAM and a bit more SSD capacity and PSU wattage. At launch, there will be four system configuration levels which you can see in the chart below.

  Level 1 Level 2 Level 3 Level 4
Processor Intel Core i7 2700K Intel Core i7 3930K Intel Core i7 3960X 2x Dual Intel Xeon E5-2630 Six-Core
Memory 16 GB DDR3 1600 MHz 16 GB DDR3 1600 MHz 16 GB DDR3 2133 MHz Corsair GT 32 GB DDR3 ECC REG 1333 MHz
Graphics Card(s) 1x EVGA GTX 680 2x  Dual SLI GTX 680 3x Triple SLI GTX 680 3x Triple SLI GTX 680
Storage 120 GB SSD & 1 TB HDD 120 GB SSD & 1 TB HDD 120 GB SSD & 1 TB HDD 180 GB SSD & 1 TB HDD
Motherboard ASUS P8Z68-V
PRO/GEN3
ASUS Sabertooth
X79
ASUS Rampage
IV Extreme X79
EVGA Classified
SR-X
Power Supply Corsair 1050W Pro Silver Corsair 1200W Pro Gold Silverstone 1500W SST-ST1500 Silverstone 1500W SST-ST1500
Optical Drive Slot Loading DVD Writer Slot Loading DVD Writer Slot Loading DVD Writer Slot Loading DVD Writer
OS Windows 7 HP x64 Windows 7 HP x64 Windows 7 HP x64 Windows 7 Pro x64
Price $3,859 $4,985 $6,687 $7,850

 

The hardware is nice, but it is not the only interesting aspect of the new Aventum PCs. Rather, it is the custom chassis that holds the Digital Storm hardware. The metal full tower ATX case is divided up into sections and supports three 420mm (3x140mm) radiators, and 13 case fans to keep the Cryo-TEC thermo-electric cooler from overheating. The cooler is placed directly on the CPU and then is itself cooled by a water cooling loop. There are two 420mm radiators in the bottom of the chassis along with the computer’s power supply.

cryo-tec-04.jpg

The Digital Storm Cryo-TEC cooler installed in a system.

Digital Storm has designed it such that three 140mm fans draw cool air in from outside of the case, through the radiator, and then channels the heated air out of the back of the case via vent under the power supply. The 13 case fans provide cooling for five cooling “zones” and are monitored and controlled by temperature probes using Aventum software in Windows. System and temperature information is also displayed on a built in LCD on the right side of the case.

04-Exhaust-Chamber.jpg

Another interesting aspect of the Aventum chassis is that the hardware is installed “backwards” in the case such that it can be viewed through a window on the right side of the case (instead of the left in the majority of cases). It also features a removable drive cage with four 3.5” drive bays. There is also support for two internal 2.5” drives and a slot loading DVD writer optical drive accessed on the top of the case. Power and reset buttons are located just under the DVD drive while four USB ports and two audio jacks (1 mic, 1 headphone) are located on the right side of the case near the DVD drive.

Aventum_01.jpg

The case also features plenty of mesh patterned ventilation holes and cut out Digital Storm logos. Also, there is a Digital Storm logo on the front of the case that is back-lit by a customizable LED color. Digital Storm’s Director of Product Development Rajeev Kuruppu noted that their research department has worked for months with thermal imaging cameras to ensure that the high end components are cooled as efficiently as possible. ”Every integral component and every zone is constantly being monitored so our customers can ensure their dream machine is always delivering optimal performance.”

02-Left-Panel.jpg

The Aventum systems are available now and range in price from $3,859 to $7,856 depending on the particular configuration. More information will be posted on the Digital Storm website later today.

About that pricing AMD; you sure you want to stick with it?

Subject: General Tech | March 26, 2012 - 01:45 PM |
Tagged: gpu, pricing, nvidia, amd, radeon

AMD has spent a lot of money developing GCN and it shows with products that provide better performance than the previous generation and do so with less power consumption, a hard trick to pull off.  There are also numerous other architectural changes in the three current families of Southern Island cards which benefit users, but most will be focused on faster graphics without the need to upgrade their PSU.  Until last week, since AMD had the fastest GPU going period, as well as much better price/performance numbers than NVIDIA's choice, there was no reason for AMD to consider changing their pricing structure as they need to recuperate the amount of dollars spent on R&D as well as manufacturing.

Last week the GTX 680 changed that, as not only did NVIDIA steal the performance crown back from AMD but they also successfully reduced the power consumption which was the Achilles Heel of Fermi.  Even worse news for AMD was the pricing that NVIDIA attached to their flagship Kepler product, at $500 they are priced below AMD's HD 7970 by between $50 to $100.  AMD's only hope is that the process problems at TSMC will keep the availability of the GTX 680 down, which it seems to have as NewEgg has run out of that card.  Hoping that your competitor cannot keep their stock up is not exactly a good model to run your business.

Unfortunately any price change AMD makes will have repercussions on many models.  The 7950 averages about $460 which is far too close to the GTX 680's price since the performance is not that close, however dropping the HD 7950 towards $400 makes the HD 7870 at $360 a little uncomfortable.  That is going to have an effect on AMD's profitability, since they likely set out their accounting based on the current pricing of the Radeon series and will have to recalculate a lot of numbers to lower price and still remain profitable.  However painful a process that might be they need to think of it sooner, rather than later; NVIDIA has more Kelper cards in store and they are not going to cost more than the GTX 680.

So far we have not heard any substantiated rumours about price changes from AMD but you can speculate that they must be coming.  For now you should first decide how much your budget can manage and then start looking for specials at retailers that bring the cards down to the price you have decided you can afford.  If they aren't low enough today then wait a few days as the GPU market is going to be decidedly unstable for the next while. 

AMD-VS-Nvidia.jpg

Here is some more Tech News from around the web:

Tech Talk

 

Galaxy Readying 4GB and Hall Of Fame Edition GTX 680 GPUs

Subject: Graphics Cards | March 26, 2012 - 11:36 AM |
Tagged: nvidia, gtx680, gpu, galaxy, 4gb gtx 680

A new article over at Chinese site EXP Review suggests that graphics card manufacturer Galaxy is gearing up with three new NVIDIA GTX 680 graphics cards. Among the new GTX 680 GPUs pictured, Galaxy is planning a reference card, a heavily overclocked Hall Of Fame card, and models with 4 GB of GDDR5 memory.

galaxy_68s.jpg

Galaxy is upping the memory ante with a new NVIDIA GTX 680 that has 4 GB of memory - twice that of the reference cards. The card will use Hynix memory chips (8 on the front, 8 on the rear of the card), an improved 5+2 phase power supply with DirectFET replacements for the usual MOSFET design. In addition, the card features 6+8 pin PCI-E power connectors and an aftermarket Galaxy Gemini cooler. The Gemini HSF uses heatpipes and an aluminum fin array cooled by two 90mm fans to cool the card. The extra cooling enabled Galaxy to offer the new card with a 10% factory overclock, for a base clockspeed of 1.1 GHz.

GTX6804GB_03.jpg

The other interesting card is the upcoming Galaxy GTX 680 Hall of Fame edition. This card is based on a white PCB with two 8 pin PCI-E power connectors (like those of the Asus DirectCU II and MSI Lightning). Further, it is cooled with three 90mm fans and five heatpipes leading to an aluminum fin array. The card will come equipped with dual BIOS support and overclocking software. It is not directly stated, but the Hall of Fame edition should be more overclockable thanks to the expanded cooling solution. Also, it may come with 4 GB of memory like the card above.

GTX6804HOF_01.jpg

In our PCPer Live Review, it was stated that while NVIDIA could do reference cards with 4 GB of memory, they chose not to in order to hit certain price points and profit margins. Instead, they left it up to the Add In Board partners to offer cards with extra RAM and factory overclocks. Galaxy is prepping their 4 GB cards, but theya re not likely to be the only vendor offering cards with increased memory.  More photos are available over at EXPReview.

Source: EXPReview

Intel becomes a 22nm foundry -- no, not for NVIDIA.

Subject: General Tech, Mobile | March 25, 2012 - 08:42 PM |
Tagged: TSMC, nvidia, Intel

NVIDIA would like Intel to be their fab partner for ARM processors. Turns out NVIDIA-produced ARM products are not tempting to Intel.

Last month we reported that Intel would open up their fabrication plants to contracts from other companies. We stated that the world would likely end if Intel were to ever produce products from NVIDIA. It turns out that the world is safe.

intel-cmon-not-you.png

Turn out the lights, pretend we’re not home.

Intel is far and away the most advanced semiconductor fabricators in the world and many companies would love to have their components created in their factories. Intel is very aware of how sophisticated their technique is relative to their competitors and exercises that advantage.

NVIDIA currently fabricates their chips at TSMC. That partnership has proven to be slightly problematic to NVIDIA’s business goals. Their Kepler launch turned out to not be nearly as soft of a launch as was proposed by SemiAccurate -- but that is to be expected from a website by that name (especially with NVIDIA news).

Perhaps you were a little too greedy in requesting that Intel manufacture your ARM processors, NVIDIA? Maybe you should test the waters with a discrete GPU order or, you know, some other market that Intel does not compete in try as they might.

Even still, there was a rumor going around when Intel partnered with AMD for hardware-accelerated physics support. It does not seem like Intel really want to be friends. Plenty of fish in the sea, though.

Source: Forbes