Industry Information Station

Earlier this week, some people may have seen the AMD TressFX physics physics and rendering techniques “flyers”. The chip manufacturer today announced details of the TressFX–which is a set of new tools for creating realistic effects. The system is expected to render the hair at the same time, but also for the players to show the appearance of the real and lifelike action. TressFX is AMD and Crystal Dynamics of fruitful co-operation, the Game Studio building is the Tomb Raider series of iconic barrier control, about ten years ago.

“Next-generation” Lara Croft: Tomb Raider (basically duplicate), with its iconic “tied in a bunch of” different female hero Lara also has fine rendering and long flowing hair. Throughout the history the game, “hair” is really only some “hair” of a small number of polygon texturing.

TressFX has created a system where developers can have tens of thousands of separate and flexible hair and may be subject to gravity, wind, movement, and even the effects of humidity. This allows the TressFX to render special real effects.

AMD advanced “salons” tool, based on its current OIT “no rule of transparency” (Order Independent Transparency), which uses the same PPLL “per-pixel list” (Per-Pixel Linked-List) of data, the complexity of managing memory usage, and rendering. FressFX can also take full advantage provided by AMD GCN DirectCompute for hardware-based massively parallel processing capability.

Although AMD is not stated, but also support DirectCompute and Nvidia OIT (PDF). This may indicate that TressFX or will work on a wider variety of GPU, and not just limited to AMD’s Radeon series. However, if it is not “exclusive”, I’ll bet TressFX Radeon GPU performance, will be the best.

[Compiled from: TechSpot]

AMD code Jaguar (Jaguar) processor is the Bobcat (Bobcat) successors, and will take on 28nm process technology, is designed specifically for mobile device chips prepared by solution, used Jaguar architecture chips ranging from less than 5W to 25W power consumption products. Jaguar in the schema for each CU unit consists of 4 core, each of which is the 512KB cache, cache 512KB Bobcat, but cannot be shared between core, Jaguar was in the cache is shared between four core.

In the presence of AMD at the International Solid State Circuits Conference (ISSCC), more than 4 core processors AMD did not directly address the future plan, but hinted at future publishing chip may use more than one CU unit, 6-core or 8-core processor also has the advent of May.

Jaguar schema instruction execution efficiency higher than the Bobcat, and clock frequency, power consumption, there are different degrees of increase. Access units compared to the previous generation schema to achieve performance gains were also redesigned, also supports 128-bit floating point unit. AMD also said that Jaguar will play all, frequency you want to over 1.85Ghz, but eventually is limited by power consumption of SoC issues, each core power-0.5W.

In November 2012, when AMD introduced the code for “Abu Dhabi” Opteron server CPU Opteron 6,300 series, but at that time we did not sample that can be audited. Data provided by AMD some confusing, its best score is difficult to assess the SPECjbb2005 benchmark; SPEC CPU2006 benchmark tests are also very general. According to mentioned in AMD, Opteron with 2.4GHz compared to 6,378, Opteron 6380 (2.5GHz)-leading 24%, up to 40% performance-per-watt.

Contrary to the, SPECint_Rate2006 only enhanced the cent, SPECCfp_Rate2006 upgrade-only 8%. However, SEPC CPU2006 rating scale with the clock rate is not so much. For example, 6380 vs 6,376 8% clock speed differences between affected only 3.5% SPECint_Rate2006 and SPECfp_Rate2006 score high and low; but the SPEC CPU2006 benchmark tests show the best side of Interlagos Opteron. (Click here to read the AnandTech analysis report)

A far cry from the two tests with the actual server workloads, we had this two conclusions – first of all, improved performance per GHz and power; second, we can only confirm its optimized very well, even (potentially) fully recompiled the code. SPEC CPU2006 compilation set, and Specjbb JVM settings: all code that does not exist on the server that is running application.

Therefore, when you run the code, the new “Abu Dhabi” Opteron (Opteron) is only a little faster, also clearly faster? Increase in energy consumption are the best or that can be quantified? Well, we (Anandtech) in most of the benchmarks that are in the real world, if you are interested in, you can read the full version of Opteron 6,300-benchmark evaluation articles .

---

· 1 Introduction (introduction)
• 2 Positioning: Skus and Price (Location and price)
· 3 Benchmarking Configuration (test configuration)
· 4 Virtualization Performance:Linux VMs on ESXi (virtualization)
· 5 Measuring Real-World Power Consumption (actual power consumption)
· 6 SAP Sales&Distribution (SAP accounting software)
· 7 Java Server Performance (Java Server)
· 8 Rendering Performance:Cinebench (rendering performance)
· 9 3DS MAX 2013 performance test
· 10 LS-DYNA Neon-Refined Revised and a three-car collision test
• 11 LS-DYNA Power Consumption (LS-DYNA-power test)
· 12 Encryption and Decryption (encryption and decryption)
· 13 Compression and Decompression (compression decompression)
· 14 Conclusions (conclusion)

[Compiled from: AnandTech]

AMD has revealed its Turbo Dock technology, which is for the Tablet PC and ultra portable notebook market latest attempt, based on a Jekyll&Hyde APU–when docked to the base (dock) when the priority property; while in tablet mode, you give priority to battery life. Their debut was on the AMD Temash SoC this year, Turbo Dock was actually a dynamic throttling system. Suppose you like more system intensive content creation, when he is a Hybrid PC access, could bring to CPU and GPU performance gains of up to 40%.

AMD Turbo Dock technology

When the screen section be removed, Turbo Dock automatically throttled to extend battery life. In fact, the adjustment of the system depends on the user which application is running. This means that if you just watch the video, using the APU is enough.

At the same time, Temash latest low power AMD processors, based on Windows 8 tablet and mix. Thanks to 28nm processor, its commitment to make 5 times times the graphics performance of the Intel Atom Z2760-for example, Acer Iconia W510 mix this-and dual-core and quad-core version.

AMD Turbo Dock Technology

At next week’s Mobile World Congress (MWC), AMD will display A6-1450 and other completely Temash SoC. Although there is no specific brand, first Temash and supports Turbo function of Dock plates and mix, will be launched at the end of 2013.

[Compiled from: SlashGear]

AMD Fitch Ratings lowered b before the company’s long-term debt rating to CCC. In accordance with the definition of Fitch Ratings, b-grade two lower than the investment-grade rating, CCC is three less than the investment-grade level. Fitch explained that the CCC level means that the company’s debt repayment holds large changes, and high dependence on the external economic environment. Fitch AMD cash flow is expected to reach negative values, target below the amount could drop AMD, and close to their operating minimum requirement level.

Fitch said that AMD’s sales growth may be slowed, mainly because of the slowdown in consumer demand for PC products, more people have begun to buy products such as iPad. At the end of 2012, AMD has about US $ 1.18 billion in cash, enough AMD to maintain normal operation. But with cash flow negative, and needs it in 2013 and 2014, respectively, to production partners cover the cost of 215 million dollars and 200 million dollars, AMD will face greater financial pressures. AMD cash liquidity to set itself a target of 1.1 billion dollars, a minimum of US $ 700 million. At the end of 2012, AMD’s total debt stood at US $ 2.1 billion.

After AMD announced the restructuring strategy, outside it is expected that these measures will help AMD to cut $ 450 million cost. But the company’s cash expenditures will continue to be outside the focus of the future.

Fitch said in its report, while AMD had previously set a target high-growth my product sales ratio increased to 40% per cent of total sales, but due to the traditional PC market domination in the AMD sales status, the company this transition may take several years to complete.

AMD will update the series a “Trinity” APU, leads to 20%~40% is said to be performance-enhancing. The company code-named the “Richland” new APU, had wanted to OEM manufacturers shipments, and show more detail. APU process and structure of the same named Richland, CPU is Piledriver hammer, and remains a 32nm SOI process, graphics are upgraded to HD series of 8,000, now more about Richland model and specification were also exposed.

In CES exhibition Eve AMD announced has this year of products route figure, in addition to focus care of low power platform zhiwai, desktop level APU products will has two times update, code Richland of APU technology and schema does not variable, CPU also is Piledriver piling machine, system way still is 32nm SOI technology, video will upgrade to HD 8,000 series, second half is is code Kaveri of APU, technology upgrade to 28nm.

Currently more information about Richland APU has also surfaced. First name Trinity generation model A-5000 series, so naturally Richland series A-6000 series, earlier had been exposed A10-6800K, now there are more types of specifications.

A10-6800K: quad-core, TDP 100W integrated HD 8670D video card

A10-6700: quad-core, TDP 65W, integrated HD 8670D video card

A8-6600K: quad-core, TDP 100W integrated HD 8570D video card

A8-6500: quad-core, TDP 65W, integrated HD 8570D video card

A6-6400K: dual-core, TDP 65W, integrated HD 8470D video card

A4-6300: dual-core, TDP 65W, integrated HD 8370D video card

Like the current Trinity TDP, dual-core 65W, whether with k. K is 100W in quad-core, with no k is 65W.

Integration of video is HD 8,000 series of, but cannot determines is HD 7,000 series of vest also is real of HD 8,000 Island schema, after all Richland to until Q2 quarter to publishing, AMD of route figure in the said its is Radeon Core 2 schema, which HD 8370 HD is 128 a flow processor unit, HD 8470D is 192 a flow processor unit, HD 8,570 and HD 8,670 no specific data, Guess respectively 256, 384 stream processor units.

If the schema is based on HD 8,000, Richland APU in HD graphics performance may 7750-HD 7,770, is very exciting.

Platforms, Richland matching chipset is upgraded, replacing A85X,A78 A88X replacement A75,A68 replacement A55, slot type or a Socket FM2, however this FM2 with Trinity APU use FM2 is compatible or not say, somewhat similar to the AM3 and AM3+ relations.

Other specifications change, memory support for 1866MHz+, according to AMD official documents on the Richland was actually to 2133MHz and even support of 1.25V low-voltage DDR3 memory. 8 SATA 6Gbps,USB SATA interface 3 interface 4, USB 2, 14.

More information Kavri no, FX series will not have any new products this year, neither new processor or motherboard new products next year.

[via：TechSpot]

AMD fourth-quarter losses widen. In the context of weak demand for personal computers, the company is still difficult situation. The company expects income of around 1.05 billion dollars in the current fiscal first quarter would be higher than four-quarter decline of about 9%. A recent Thomson Reuters (Thomson Reuters) survey of analysts forecast first-quarter revenue of us $ 1.1 billion.

AMD announced that the fourth-quarter loss of $ 473 million, or a loss of 63 cents; a loss of 177 million dollars in the same period of last year, or a loss of 24 cents. Four-quarter results included $ 90 million restructuring charge, compared to the same period takes into account the impairment of Globalfoundries stake held by related expenditures of $ 209 million. Excluding restructuring charges and other special items, the company lost 14 cents per share for the first quarter, compared to earnings per share of 19 cents in the same period. Analysts expected before the company lost 20 cents a share excluding special items.

When the quarter down 32%, to 1.16 billion dollars, in line with the company in October last year to about 1.1 billion-US $ 1.21 billion in expected.

---

AMD was making beyond, in parallel with strengthening solid profits from the PC market, the company will be more and more attention has been given in the Smartphone and Tablet PC.

AMD heavyweight APU (accelerated processor) is market-led by industry is expected in the future, at present, AMD has been your focus to the lucrative field of mobile devices, this area is currently being led by the ARM based on RISC chips.

So, AMD has recently hired two engineers working in the Qualcomm and Apple would not be too surprising.

Citing Reuters, Charles-Marta Vieira da Silva, previously serving in high-throughput low power and embedded chip design has a wealth of experience in the field, and now he is an AMD system on chip (System-on-chip), Vice President of development.

At the same time, Wayne-mailaitesikai from Apple as Vice President in charge of software IP development, he had involvement at Apple iOS device processor development.

It is interesting, two engineers from the earlier AMD’s work experience. Before them, in August last year came to AMD chip design guru Jim Keller became Chief Designer.

In response, AMD spokesman drew-Plailly (Drew Prairie) stated that this appointment will help AMD to expand in new markets.

Noteworthy is, since Lorette (Rory Read) after 2011 as AMD’s CEO, head of the company has undertaken major adjustments. Last October, in order to streamline its operations, AMD announced 15% job cuts, which are the layoffs for the second time within a year’s time.

AMD expressed the hope that the increase in communications, and digital revenues in areas such as micro-server. In the next three or four years, AMD expects non-PC business revenue will account for about 50% of the company’s total revenues.

As successor of the Trinity processor, Richland APU has already shipped with the PC manufacturer. Updated products will have a lower power consumption and higher clock speeds, and AMD branded applications bundled and “optimized” after the game. Although Richland seems to be mainly for mobile applications, however the desktop version is on the way (using existing FM2 socket on the motherboard). But for integrated Radeon 8,000 series GPU, you pretty much have no, no happy too early, its frequency slightly higher than its predecessor, but not to 100MHz.

Richland APUs

Higher frequencies than the Trinity 300MHz the basis of Richland, and small Turbo Super 200-400MHz,TDP and 65W. Although it is not very compelling desktop product, it can be considered a step forward.

New APU was a transition, until the latest video of the arrival of the Kaveri are integrated (based on current GCN schema). Kaveri on 28nm process technology will be used, and Trinity and Richland was 32nm.

Related news:
[Map] AMD Richland APU: integrated HD 8,000 graphics, chipset update

[Compiled from: TechSpot]

On January 26, over the years, AMD APU platform has become increasingly mature. The second quarter of this year, AMD will transition from the current Trinity series to series of Richland . New series include the dual-core and quad-core product, TDP (thermal design power) 65W and 100W respectively.

Insiders disclosed that Richland APU 32nm manufacturing process and an enhanced version of the x86 pile (Piledriver) schema, FM2, A85X Bolton D3 chipset. New Richland APU lineup is as follows:

A10-6800K Radeon HD 8670D/quad-core four

A8-6600K Radeon HD 8570D/quad-core four

A10-6700 Radeon HD 8670D/quad-core four

A8-6500 Radeon HD 8570D/quad-core four

A6-6400K Radeon HD 8470D/dual-core 65W

A4-6300 Radeon HD 8370D/dual-core 65W

Flagship model A10-6800K Radeon HD 8670D iGP (integrated graphics processor) and fully unlock the overclocking of the design. A8-6600K uses lower-performance Radeon 8570D iGP. The two quad-core model TDP is 100W, but frequency of operation vary.

In addition to TDP 100W quad-core models, Richland APU lineup also includes models of 65W. A10-6700 and A8-6500 are 65W TDP, but in order to reduce power consumption, their relatively low operating frequency. A10-6700 and A8-6500 iGP HD and 8670D HD, respectively.

Richland APU in the lineup only two TDP 65W dual-core model, using low-end iGP. A6-6400K 192 GCN stream processors of the schema, A4-6300 has 128 stream processors. Two models with 1 MB L2 cache. AMD is already shipping Richland APU to OEM manufacturers, but high-volume supply until the second quarter of 2013.

Source: Guru3D

Thanks anonymous posting
In the 12 years of cooperation, launched 4 new cars, following the 6 champions, AMD and Ferrari have ended relationships. AMD no longer appear in the Ferrari on the list of technology partners in the ecosystem. AMD has now decided to use resources more wisely, instead of spending money on advanced motion. Long-term relationship between Ferrari and AMD, AMD also the Abu Dhabi Investment Group investments, the Group 5% of Ferrari shares owned at that time.

AMD and Ferrari will be cooperation between Taiwan manufacturers Acer involved, Acer introduced the Ferrari brand of laptops, which use AMD hardware. Among them, the carbon fiber shell 4,000 series won numerous international media awards.

According to informed sources, the parties terminate cooperation main reason is that AMD did not renew their contracts to continue to cooperate, Armstrong after scandal (AMD is one of the main sponsor of Armstrong’s cycling team), AMD decided to stay away from sponsoring sports, at the same time come up with resources to create additional value for consumers.

Coincidence is that left Ferrari in AMD ecosystem at the same time, Intel announced that starting from Acer Liquid C1 Smartphone with its Atom SoC on-chip system drive Acer Smartphone, before given Acer Ferrari Edition Liquid e smartphone launch, speculation Intel will replace AMD to cooperate with the Ferrari brand and sponsorship.