Intel Haswell Refresh
Hardware

We have been telling you about this in our previous leaks for quite some time now, but the official confirmation (by VR Zone) comes no less the sweeter. All our previous Haswell Refresh and Intel 9 Series Chipset leaks have been more or less confirmed, and not only that, but the dates for the Intel Haswell, 9 Series Chipset and Devil’s Canyon has been confirmed.Intel Haswell Refresh Devil's Canyon

Intel Haswell Refresh and 9 Series Chipset launching on May 11th – Devils Canyon and Pentium K Following 2nd June 2014

I guess I don’t really have a lot of information for this particular post apart from the fact that I can now completely confirm the launch dates of all the upcoming Intel SKUs. Basically the Intel 9 Series Chipset Z97 and H97 will be launching along with the Haswell Refresh lineup. The Devils Canyon ‘K’ Unlocked Haswell Refresh will be following along with their Pentium ‘K’ counterparts on the second day of June 2014. Several Z97 motherboard lineups have so far been leaked which include MSI’s Z97 Gaming, ASUS Z97 ROG, ASUS Z97 TUF, ASUS Z97 Legacy and Gigabyte’s Gaming lineup. The Intel Haswell Refresh processors launching on 11th May are listed below:

AMD logo
Hardware

Disclaimer: The conclusion and guidance given at the end has been given in a completely non-professional capacity. Since this is not intended to be professional investment advice, we cannot be held liable in any way.

AMD (NYSE: AMD) has just reported its Q1 Results for 2014. About 3 months back, I advised against investing (short) because of my forecast of a significant decrease in revenue next quarter. And as far as shorts are concerned, the chipmaker has infact reported a loss. Interestingly however, the reported GAAP diluted LPS (Loss Per Share) of 0.03$ and consequently the reported non-GAAP diluted EPS of 0.02$ is higher than analysts expectation of a  break even on non-GAAP.

Advanced Micro Devices, Inc. (NYSE: AMD) nets Loss Per Share of 3 Cents on GAAP basis – SCRYPT Mining Sales Drying Off, CPU Portion in Decline while GPU portion sees Significant Growth

AMD StockNumbers aren’t everything, because even though the numbers might make it appear as if this Quarterly Report depicts a bad result, the reality is quite different. What we are seeing are signs of an actual, possible company turnabout. I.e. this report could be the first sign that the Profit of Q4 FY13 was not a mere fluke.

AMD logo
Hardware

During their quarterly earnings call for Q1 2014, AMD revealed that they would stick with 28nm across all their products as opposed to 20nm which is the process reserved for their future products. Many have their eyes on AMD’s next generation of graphics cards which will feature new and improved 20nm process technology but we have to wait a bit more to see those GPUs in action.AMD 20nm

AMD To Stick With 28nm Across All Products This Year

Lisa Su (Senior Vice President and General Manager of Global Business Units at AMD) revealed that their graphics and semi-custom business products are both on 28nm. We have seen several 28nm products across AMD graphics and processors designs which include the Radeon R9 and Radeon R7 series for graphics while Steamroller, Jaguar powering AMD’s processors and accelerated processing units. The transcript of the session can be seen below:

Lisa Su – SVP and General Manager of Global Business Units

Sure, Chris. So let me take that and give you a little bit of our thinking.

So in terms of product and technology selection, certainly we need to be at the leading-edge of the technology roadmap. So what we’ve said in the past is certainly this year all of our products are in 28-nanometer across both, you know, graphics client and our semi-custom business. We are, you know, actively in the design phase for 20-nanometer and that will come to production. And then clearly we’ll go to FinFET. So that would be the progression of it.

Lisa Su – SVP and General Manager of Global Business Units

David, I think what I said earlier sort of what we’re doing in terms of technology strategy, we are 28 this year, we have 20-nanometer in design, and then FinFET thereafter. So that’s the overall product portfolio. via Seeking Alpha

Intel Haswell Refresh
Hardware

The retail prices of Intel’s upcoming Haswell Refresh processors have been revealed by Computerbase which show variants ranging from the top $303 US Core i7-4790 to the entry tier Celeron G1840T which will cost $42 US. Intel’s Haswell Refresh processors will launch on 10th of May and would be available for both Mobile and Desktop consumer platforms.Intel Haswell Refresh

Intel Haswell Refresh Desktop and Mobile Processors Prices Confirmed

Intel initially launched their fourth generation Haswell processors in June 2013 at Computex. The processors featured the latest 22nm core 3D Tri-gate transistors technology which improved the IPC and efficiency by a decent amount of Ivy Bridge. The upcoming Haswell Refresh processors are no different then these featuring bumped up clock speeds and more SKU design schemes such as low TDP and low power variants available at similar price points.

A total of 26 desktop processors have started shipping to retailers across the globe and  would be available for sale in the upcoming days prior to their official launch planned for next month. The Core i7-4790 will be the flagship processor of the Haswell Refresh lineup and the first 22nm consumer CPU to obtain clock frequency of 4 GHz. It will feature 4 cores and 8 threads. The Core i7-4790 will also be available in ‘S’ and ‘T’ variants which will retail at the same price of $303 US but will feature lower clock speeds since they are aimed towards low TDP and low power platforms. Other models include Core i7-4785T, Core i5-4690, Core i5-4590, Core i5-4460, Core i3-4360, Core i3-4350, Core i3-4150 and several Pentium and Celeron variants all of which would be available in “T” and “S” variants.

Intel-Logo
Hardware

Back when Intel first entered the scene at Costa Rica, it jump started the country’s Tech Industry. It also gave Costa Rica a much needed boost out of the shadows of its neighboring countries, and into the lime light. Infact Intel was seen proclaiming on more than one occasion how its presence had increased the Tech exports of said country by such a large amount that the traditional primary export of coffee. Now with Intel closing shop and laying off 1500 workers, this comes as nothing short of a miniature financial crisis for the Costa Rican Republic.

Intel facility in Costa Rica

Intel lays of 1500 workers and closes operation in Costa Rica – Could result in a miniature financial crisis for the Republic

Intel’s Microprocessors account for roughly 25% of the exports of the Republic of Costa Rica and roughly 5% of its GDP. You don’t need to be an Economist to see how big a chunk of its wealth will go missing after Intel departs Costa Rica. The Bank of America, probably following Intel’s cue and sensing the changing tide, immediately layed of 1400 workers after this announcement. I am tempted to quote a well known movie phrase but I will restrain myself, however the fact remains that such a major decision will have tidal consequences. Interestingly this means that Intel is actively scaling down. This is probably one of the most alarming signs that actually put some weight on the claims that the PC Market is shrinking.

Intel recently halted work on its Fab 42 and now it seems it has closed shop in Costa Rica too. I remember that the Costa Rican batch of processors was being favored a few years back and this shows how much intrinsic value this fab had to Intel. Closing it down cannot be good in any way. Unless, the rumors are correct and Intel is headed for China. China has cheap labor. And if Intel is truly chasing cheap rates per hour then I must express my distaste. Not that the Chinese market shouldn’t be entered but the way in which it is being entered is borderline unethical.

chips-graphic_fire_0_0_0
Hardware

Chip maker Qualcomm, whose products are used in some of the world’s top smartphones, including the Samsung Galaxy S5, has announced two 64-bit additions to its line of SoCs manufactured. The Snapdragon 810 and 808 both feature ARM’s A57 and A53 cores. The 810 has a 4 big + 4 Little configuration while the 808 has a 2 big + 2 Little configuration. Qualcomm explained the shift to the new configuration as a reaction to the changing demands of the market in the demand of the 64-bit models supported by ARM cores; a change possibly started by Apple’s release in the iPhone 5S.

Snapdragon 810_slides for press_689_2

The Snapdragon 810 to feature a slew of new features including LPDR4 memory, rendering of 4K video at 30 fps and 1080p at 120 fps.

In addition to the adding of the A57 for high power tasks and the A53 for lower level ones, the 810 also has the modem integrated directly within the SoC die. In terms of connectivity, the Cat 6/7 on both the chips, has the capability for 20 Mhz exaggeration capability to achieve 300 Mhz down link speeds. In terms of GPU, in the 810 the GPU has been upgraded to the Adreno 430, which should bring significantly improved performance and lower power consumption when compared to the 805. It also shows Qualcomm’s focus to the 4K display and an increase in 3D graphics capability. The 810 supports two 4K displays at 30 fps, while the 808 has support capability for 2560/4Mp primaries.

TSMC-logo
Hardware

Taiwan Semi-Conductor Manufacturing Company, TSMC, is expected to launch its 16nm FinFET manufacturing process plus at the end of 2014. This will add two more advanced processes to its portfolio. Its competitors, Intel and Samsung already produce the 14nm nodes, and this move will likely put TSMC in a better position to compete with them. 

tsmc2

16nm FinFet plus an addition to the 16nm FinFet introduced last year by TSMC.

In September last year, TSMC released the 16nm FinFet designs. These additions were expected to provide methodologies for boosting power and area in 16nm. This included low voltage operation analysis and high resistance layer routing optimization. According to the original road map released by TSMC, the 16nm FinFet was expected to enter production by the end of this year. According to CEO Andy Liu in January, the 16nm FinFet plus has 15% better performance than its predecessor.

The 16nm FinFet plus is expected to enter production in early 2015. According to digitimes, this may help TSMC win A9 processor orders from Apple. According to JP Morgan Securities, some mobile fabless computer may directly adopt the 16nm and migrate from the 20nm, since it gives an additional die shrink.

Graphene Transistors
Hardware

Graphene is slated as the major breakthrough of this century. Infact it could very well propel the semiconductor a couple of decades easily (compared to the performance trend via Moore’s law ). Graphene transistors are more than capable of being clocked at 500Ghz so you get the idea of what Samsung is claiming to have achieved: a replicateable production process of Graphene nodes.

Intel Graphene TransistorsGraphene.

Experimental gFET Graphene Production – Scientific breakthrough of this century to be used in CPUs* of wearable devices

OK, I admit, I was being slightly sarcastic when I wrote the headline. It seems sort of ironic that if Samsung’s claims turn out to be true then the first Graphene processors will be in future wearable devices instead of lets say, desktop computers. Of course, Samsung can be forgiven for saying this considering its a primarily mobile company after all, but it still stings to a PC Enthusiast like me. Graphene, aka the miracle material of the century, to be used in future smart watches; seems like an extract from Douglas Adam’s Hitch Hiker’s Guide to the Galaxy. Anyways, enough of my rant. On with the specifics.

This breakthrough was claimed at the Samsung Advanced Institute of Technology and Sungkyunkwan University in South Korea. Basically they used a common silicon wafer with Germanium substrate to manufacture a mono layer mono crystal Graphene impression on top of it and then remove the silicon wafer (and germanium) from below. The silicon wafer can then be reused which is pretty great. Also since the Graphene is removed from the germanium using a completely dry process the Graphene is completely wrinkle free which basically means that the construction of the crystal is completely clean and low in defects. Since both the germanium substrate and silicon wafer can now be reused ( in previously known production processes the germanium substrate had to be burned off) it will exponentially increase, mass production capabilities. Samsung has started production of Graphene Field Effect Transistors (gFET)

Intel-Logo
Hardware

At the Intel Developer forum taking place currently in China, the company announced the successor to the Bay Trail SoC. Bay Trail is currently found on Intel’s high performance tablets and entry level computing devices. Braswell will feature a 14 nm architecture. This will be similar to its cousin, the next generation Broadwell and an update from the 22 nm architecture currently used in Bay Trail.

intel-braswell-tablet-mobile-cpu-processor-chip-620x354

New chip to be available for Windows, Linux, Chrome and Android. Support for Chromebook announced as well.

At the developer forum in China, Intel PC chief Kirk Skaugen told Intel employees about Intel’s hopes to help Google grow its Chromebook and Chromebox business as well. Chromebook, as we know is a growing sector of the notebook market. The notebooks run Google’s Chrome OS and are generally cheaper than $ 400. Intel also claimed that the chip will now boost the notebook designs of Chromebook to 20 in 2014 from four previous year. The move might  also be aimed at attracting more manufactures to manufacture the Chromebook. We can also expect better power efficiency and a significant power boost from Braswell as well. This is due to the shrinkage in its architecture. It will therefore be able to power budget laptops and desktops much effectively. The chip will also make use of the Intel Iris. The SoC design, Intel claims will reduce the overall size of the final product as well as the number of supporting chips required.

AMD A8 Logo
Hardware

A short update on the almost forgotten sibling of the A10-7850k: the Kaveri A8-7600. Basically, the retail SKU which is the AMD A8-7600 has been delayed to the second quarter of 2014. This is the same time frame the rest of Kaveri arrives. However the OEM variant is still on schedule for Q1 2014 release. AMD Kaveri

AMD Kaveri APU A8-7600 OEM on Schedule for Q1 – Retail delayed to Second Quarter

When the Kaveri APU launched a few months ago the engineering samples that were shipped to review sites included the A8-7600 along with the A10-7700K and A10-7800K. However though both A10 APUs were available on launch the A8-7600 wasn’t (mostly). The much more energy efficient and cheaper counterpart targeted at a very select niche of low-end products was supposed to debut in Q1 2014. However sweclockers.com now confirms that this will not be the case. Homebuilders and people who require a PIB (Processor in a box packaging) will have to wait till the second quarter of 2014, which is almost 6 months after launch and reviews. Here is the official statement of AMD that was given to our Swedish friends:

“Production of the AMD A8-7600 is on track and will be available first to AMD’s OEM customers in Q1 2014 to ENSURE the introduction of AMD based systems in Q2 2014. Additional parts of the AMD A8-7600 for PIB production will be available for sale in the component channel in 2H 2014.”

AMD FX
Hardware

When AMD announced the FX-9590 last year it was seen as an attempt to connect back with its enthusiast community. Boasting a clock speed of 5 Ghz, the chip was faster than anything else on AMD’s lineup. Even then, it would have a hard time competing with Haswell and Ivy-Bridge E. Now, according to our sources, the FX-9590 is reaching the end of its life cycle only after a mere 9 months.

FX-9590 Exiting before its time, is Heat Problem the real reason?

FX-9590Heat was naturally going to be a problem the 9590 would have to face after being able to clock at 5 Ghz. The TDP of the board is very high 220 Watts. This brings strong problems in areas of motherboard compatibility with a variety of hardware. Another problem is the cooler choice for this heat produced. However, perhaps the most valid reason for the exit of the FX 9590 is the release of Mantle.

Mantle, as we know, works by cutting down CPU bottlenecks and therefore allows applications to utilize the CPU and GPU capabilities more effectively. This allows greater control over the hardware installed. Therefore, due this better utilization, the clock speed of CPU’s in gaming and other applications could be less important, with greater focus being paid towards better utilization of available hardware resources. This leaves the 9590 rather in the middle, with it’s selling point left obsolete.

OLYMPUS DIGITAL CAMERA
Hardware Industry

Intel has announced improvements to its SD card size computer called the Edison. The Edison is targeted at wearable devices. The computer was announced in January this year and has the Intel Quark x86 CPU, with a clock speed of 400 MHz.

intel-edison-wearables

More sensors added to the chip. Upgrades are likely to expand chip usability.

The original Edison computer had limited sensing capabilities and could only be used on small, wearable devices such as watches and health monitoring. Intel has now added a sensor hub to the new edition. This will enable users to add sensors for location, environment and light, said Mike Bell vice president and general manager of New Devices group at Intel. Intel hopes the redesigned Edison will be used in the design and development of wearable products. Bell said he envisioned the Edison to be used in wearable devices such as a US football hat where the chip would be capable of doing real time analysis related to the impact of a hit.

This is the reason why Intel did not focus on size when adding improvements to the Edison. The improvements will make the rectangular circuit board 1 millimeter larger on each side. This is due to the addition of a circuit block with connectors to which additional sensors will be attached. “Smaller may not be better,” Bell sais, emphasizing upon the need to increase size in order to improve the capabilities of the computer. “You can connect [any sensor] to this”, he said.

AMD_APU_Kaveri_min
Hardware

We have covered some Kaveri benchmarks before and I keep thinking that they are over, but it turns out there is one aspect we did not cover. A Chinese site recently ran the Flagship Kaveri APU in a GPGPU showdown against Maxwell and other dGPUs. These benchmarks reveal surprising insights into the power and efficiency of the Kaveri’s architecture and in a few cases its weaknesses. Without any further ado (boy this line is getting old) lets begin.

AMD Kaveri APU A10-7850K GPGPU and Compute Showdown with Maxwell dGPUs

Know Maxwell made headlines for the impressive increase of Performance Per Watt and the no-external-power connector feature. However Nvidia hardware has always been very restricted in terms of Double Precision computing and GPGPU in general. The reason is of course deliberate DP restriction per GPU Cycle to usually 1/8 the SP amount. So just how well will the GTX 750 and TX 750 Ti fare against an iGPU, more specifically the AMD Kaveri A10-7850K, in a battle of Compute. What may normally be an unfair match (iGPU vs. dGPU) becomes surprisingly competitive as you will see.  Oh and I almost forgot, a big shout out to inpai.com.cn for providing us with the benches. A DDR3-2400 Mhz Dual Ranked RAM was used for the APU to make sure it performed at its best.

Intel-Logo
Hardware

Altera is one of the largest (if not the largest) FPGA and SoC maker out there right now. Their partnership with Intel has a good history and their build quality is usually remarkable. Interestingly, Altera’s FPGA are always about one generation ahead of desktop processors as Intel Fab’s give priority to silicon request from Altera. Intel has just deepened their relationship further with Altera’s 10th Generation FPGA based on the 14nm Tri Gate node with HMCs.

Altera Stratix 10 HMC SoC

Altera 10th Generation Stratix FPGA and Socs with 14nm FinFET  and HMC

Basically Altera’s SoCs and FPGAs are divided into two main lineups. The Arria 10 lineup is the low to mid-range of Altera’s products and is outsourced to TSIntel Altera 14nm FinFET Board HMC

Now the new relationship will usher in a new range of products will be a “System in a Package” form factor. Which basically means an SoC with a board. But interestingly these boards feature an HMC or a Hybrid Memory Cube which is one of the next-big-things in the Memory Industry.  Altera’s work with Intel will enable the development of multi-die devices that efficiently integrates monolithic 14 nm Stratix 10 FPGAs and SoCs with other advanced components, which may include DRAM, SRAM, ASICs, processors and analog components, in a single package. The integration will be enabled through the use of high-performance heterogeneous multi-die interconnect technology. . The devices will address the performance, memory bandwidth and thermal challenges impacting high-end applications in the communications, high-performance computing, broadcast and military segments.

57283_Tegra4ichipshot
Hardware

With the announce of the release of the Wiko Wax and the LG G2 Mini in March, Nvidia had started to make moves towards entering the smartphone processor market. Both the phones are powered by the Tegra 4i, Nvidia’s smartphone variant of the Tegra 4 range. So it was seen as a change of course for the company as  Jen-Hsun Huang announced that Nvidia plans to aim only towards the higher end devices where graphics and performance plays an important role.

nvidia_tegra4i_2

Nvidia Tegra ‘i’ LTE SoC’s Future in Jeopardy – Probably Discontinued.

It should be kept in mind that the mainstream SoC market is already very overcrowded, with major competitors including Qualcomm and Apple possessing 70% of the market share. A highly demanding market, with stringent requirements, the chips have to be designed to specific specification in order to gain any sort of penetration. If the chip is over designed, with a high price tag, there are no sales. Same is true for under designed chips, lacking performance capabilities.

With the Tegra 4i, featuring an integrated LTE baseband and a quadcore, Cortex A9 2.3 GHz CPU, Nvidia had hoped to significantly make a dent in the value smartphone market and become a potential, strong competitor to Qualcomm. “We are going to get design wins in tablets in very important companies, super-phones in important companies, and we’re going to continue to. I believe that the mainstream smartphone market is saturated and it’s not worthwhile to pursue.” Jen-Hsun Huang, Nvidia’s CEO said.



Creative Commons License This work by WCCF (Pvt) Ltd. is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
Permissions beyond the scope of this license may be available on the Terms and Conditions page. WCCF 2011 is designed by UzEE, Inc.
Ran 22 queries in 1.554  seconds.
Find us on Google+