News Posts matching "TSMC"

Return to Keyword Browsing

TSMC to Build World's First 3 nm Fab in Taiwan

TSMC has announced the location for their first 3 nm fab: it will be built in the Tainan Science Park, southern Taiwan. Rumors pegged the new 3 nm factory as possibly being built in the US, due to political reasons; however, TSMC opted to keep their production capabilities clustered in the Tainan Science Park, where they can better leverage their assets and supply chain for the production and support of the world's first 3 nm semiconductor factory. It certainly also helped the Taiwanese government's decision to pledge land, water, electricity and environmental protection support to facilitate TSMC's latest manufacturing plan. It's expected that at least part of the manufacturing machines will be provided by ASML, a Netherlands-based company which has enjoyed 25% revenue growth already just this year.

As part of the announcement, TSMC hasn't given any revised timelines for their 3 nm production, which likely means the company still expects to start 3 nm production by 2022. TSMC said its 7 nm yield is ahead of schedule, and that it expects a fast ramp in 2018 - which is interesting, considering the company has announced plans to insert several extreme ultraviolet (EUV) layers at 7 nm. TSMC has also said its 5 nm roadmap is on track for a launch in the first quarter of 2019.

Sources: EETimes, Tweakers.net, Thanks @ P4-630!

AMD To Change Suppliers for Vega 20 GPUs on 7nm, HBM2 Packaging for Vega 11

AMD's RX Vega supply has seen exceedingly limited quantities available since launch. This has been due to a number of reasons, though the two foremost that have been reported are: increased demand from cryptocurrency miners, who are looking towards maximizing their single node hashrate density through Vega's promising mining capabilities; and yield issues with AMD's Vega 10 HBM2 packaging partner, Advanced Semiconductor Engineering (ASE). It's expected that chip yield for Vega 10 is also lower per se, due to it having a 484 mm² die, which is more prone to defects than a smaller one, thus reducing the amount of fully-enabled GPUs.

AMD's production partner, GlobalFoundries, has historically been at the center of considerations on AMD's yield problems. That GlobalFoundries is seemingly doing a good job with Ryzen may not be much to say: those chips have incredibly small die sizes (192 mm²) for their number of cores. It seems that Global Foundries only hits problems with increased die sizes and complexity (which is, unfortunately for AMD, where it matters most).

Demand for EUV Fabrication Systems Increasing; ASML Sees 25% Revenue Growth

Dutch company ASML may not be very known to us mortal users, but it has one of the greatest aces up its sleeve: it specializes in what are some of the most complex machines currently made by mankind. Extreme Ultraviolet Lithography Systems (EUV) are the kind of machines that make you look in wonder and amazement at man's ingenuity - ASML, which specializes in this type of systems, has a production capability for 2017 that numbers just 12 of these. That means on average, they take a whole month putting one of these together. That really goes to show the complexity inherent to these systems. And it shows: EUV machines are about the size of a city bus, and typically cost more than 100 million euros ($115.3 million) each.

The revenue growth forecast is spurred by an additional 8 EUV systems being ordered by ASML's clients, which include Intel, Samsung, and TSMC - some of the biggest players in the semiconductor business. The new orders brought the company's order backlog to 27 machines - more than double their current annual output. ASML is taking steps to to ensure an increase in production capability to keep up with the multi million-dollar demand: the company is set to expand its system production capability to 24 in 2018, before reaching an expected capacity of around 40 systems in 2019. Third-quarter revenue will be about 2.2 billion euros ($2.5 billion), the Veldhoven, Netherlands-based maker of chip-making machines predicts. The company's stock valuation has increased some 30% over the past year - the company's valuation currently stands at around €53 billion ($61 billion.)

Sources: Bloomberg, Tweakers.net, Thanks @ P4-630!

Xbox One X Hardware Specs Give Gaming Desktops a Run for their Money

Microsoft Sunday dropped its mic with the most powerful game console on paper, the Xbox One X, formerly codenamed "Project Scorpio." The bottom-line of this console is that it enables 4K Ultra HD gaming at 60 Hz. Something like this requires you to spend at least $1,200 on a gaming desktop right now. Unlike a Windows 10 PC that's been put together by various pieces of hardware, the Xbox One X is built on a closed ecosystem that's tightly controlled by Microsoft, with heavily optimized software, and a lot of secret sauce the company won't talk about. The console still puts up some mighty impressive hardware specs on paper.

To begin with, at the heart of the Xbox One X is a semi-custom SoC Microsoft co-developed with AMD, built on TSMC's 16 nm FinFET node (the same one NVIDIA builds its "Pascal" GPUs on). This chip features a GPU with almost quadruple the single-precision floating point compute power as the one which drives the Xbox One. It features 40 Graphics CoreNext (GCN) compute units (2,560 stream processors) based on one of the later versions of GCN (likely "Polaris"). The GPU is clocked at 1172 MHz. The other big component of the SoC is an eight-core CPU based on an unnamed micro-architecture evolved from "Jaguar" rather than "Bulldozer" or even "Zen." The eight cores are arranged in two quad-core units of four cores, each; with 4 MB of L2 cache. The CPU is clocked at 2.30 GHz.

AMD Doesn't Regret Spinning off GlobalFoundries

AMD co-founder Jerry Sanders, in 2009 was famously quoted as stating that "real men have fabs," a jibe probably targeted at the budding fab-less CPU designers of the time. Years later, AMD spun-off its silicon fabrication business, which with a substantial investment of the Abu Dhabi government through its state-owned Advanced Technology Investment Company (ATIC), became GlobalFoundries (or GloFo in some vernacular). This company built strategic partnerships with the right players in the industry, acquisitions such as IBM's fabs, and is now at the forefront of sub-10 nm fab development. It remained one of AMD's biggest foundry partners besides TSMC and Samsung, and is manufacturing its AMD processors at a brand new facility in Upstate New York, USA.

AMD, on the other hand, doesn't regret spinning off GloFo. Speaking at Merrill Lynch Global Technology and Investment Conference, CTO Mark Papermaster said, that going fab-less has helped AMD focus on chip-design without worrying about manufacturing. Production is no longer a bottleneck for AMD, as it can now put out manufacturing contracts to a wider variety of foundry partners. Its chip-designers aren't limited by the constraints of an in-house fab, and can instead ask external fabs to optimize their nodes for their chip-designs, Papermaster said. 14 nm FinFET has added a level of standardization to the foundry industry.

Source: Expreview

AMD to Continue Working With TSMC, GLOBALFOUNDRIES on 7 nm Ryzen

In the Q&A section of their 2017 Financial Analyst Day, AMD CEO Lisa Su answered an enquiry from a Deutsche-bank questioner regarding the company's aggressive 7 nm plan for their roadmap, on which AMD seems to be balancing its process shrinkage outlook for the foreseeable future. AMD will be developing their next Zen architecture revisions on 7 nm, alongside a push for 7 nm on their next-generation (or is that next-next generation?) Navi architecture. This means al of AMD's products, consumer, enterprise, and graphics, will be eventually built on this node. This is particularly interesting considering AMD's position with GLOBALFOUNDRIES, with which AMD has already had many amendments to their Wafer Supply Agreement, a remain of AMD's silicon production division spin-off, the latest of which runs from 2016 to 2020.

As it is, AMD has to pay GLOBALFOUNDRIES for its wafer orders that go to other silicon producers (in this case, TSMC), in a quarterly basis since the beginning of 2017, based on the volume of certain wafers purchased from another wafer foundry. In addition, AMD has annual wafer purchase targets from 2016 through the end of 2020, fixed wafer prices for 2016, and a framework for yearly wafer pricing in this amendment, so the company is still bleeding money to GLOBALFOUNDRIES. However, AMD is making the correct decision in this instance, I'd wager, considering GLOBALFOUNDRIES' known difficulties in delivering their process nodes absent of quirks.

NVIDIA Announces Its Volta-based Tesla V100

Today at its GTC keynote, NVIDIA CEO Jensen Huang took the wraps on some of the features on their upcoming V100 accelerator, the Volta-based accelerator for the professional market that will likely pave the way to the company's next-generation 2000 series GeForce graphics cards. If NVIDIA goes on with its product carvings and naming scheme for the next-generation Volta architecture, we can expect to see this processor on the company's next-generation GTX 2080 Ti. Running the nitty-gritty details (like the new Tensor processing approach) on this piece would be impossible, but there are some things we know already from this presentation.

This chip is a beast of a processor: it packs 21 billion transistors (up from 15,3 billion found on the P100); it's built on TSMC's 12 nm FF process (evolving from Pascal's 16 nm FF); and measures a staggering 815 mm² (from the P100's 610 mm².) This is such a considerable leap in die-area that we can only speculate on how yields will be for this monstrous chip, especially considering the novelty of the 12 nm process that it's going to leverage. But now, the most interesting details from a gaming perspective are the 5,120 CUDA cores powering the V100 out of a total possible 5,376 in the whole chip design, which NVIDIA will likely leave for their Titan Xv. These are divided in 84 Volta Streaming Multiprocessor Units with each carrying 64 CUDA cores (84 x 64 = 5,376, from which NVIDIA is cutting 4 Volta Streaming Multiprocessor Units for yields, most likely, which accounts for the announced 5,120.) Even in this cut-down configuration, we're looking at a staggering 42% higher pure CUDA core-count than the P100's. The new V100 will offer up to 15 FP 32 TFLOPS, and will still leverage a 16 GB HBM2 implementation delivering up to 900 GB/s bandwidth (up from the P100's 721 GB/s). No details on clock speed or TDP as of yet, but we already have enough details to enable a lengthy discussion... Wouldn't you agree?

TSMC Trade Secrets Stolen - Former Engineer Arrested In China

In the highly competitive, high-stakes scene of the business world - and particularly so in the silicon giants of the era - trade secrets, specifications, and protecting one's intellectual property that give the leg-up on competitors is key towards success. And while most companies work within the meanders of law (even if sometimes skirting it ever so lightly), some don't. And things like this happen: the steal (or purported steal, because no one has been convicted yet) of trade secrets by former employees is one of the most dreaded occurrences in the tech world - remember Zenimax and Carmack's "dovetailing"?

Chinese manufacturers are looking to enter the high-performance computing market with their own products, designs, and manufacturing capability. In this case, former TSMC engineer Hsu is being accused of stealing proprietary information and other materials related to the foundry's 28 nm process technology. The goal would be to pass them to China-based Shanghai Huali Microelectronics (HLMC), with which he accepted a job offer, according to the Hsinchu District Prosecutors Office. Digitimes reports that HLMC had been aggressively headhunting for talent to kick start its 28 nm manufacturing process, though if true, this sound like a little too aggressive of a headhunting.

Source: Digitimes, WCCFTech

NVIDIA to Build "Volta" Consumer GPUs on TSMC 12 nm Process

NVIDIA's next-generation "Volta" GPU architecture got its commercial debut in the most unlikely class of products, with the Xavier autonomous car processor. The actual money-spinners based on the architecture, consumer GPUs, will arrive some time in 2018. The company will be banking on its old faithful fab TSMC, to build those chips on a new 12 nanometer FinFET node that's currently under development. TSMC's current frontline process is the 16 nm FFC, which debuted in mid-2015, with mass-production following through in 2016. NVIDIA's "GP104" chip is built on this process.

This could also mean that NVIDIA could slug it out against AMD with its current GeForce GTX 10-series "Pascal" GPUs throughout 2017-18, even as AMD threatens to disrupt NVIDIA's sub-$500 lineup with its Radeon Vega series, scheduled for Q2-2017. NVIDIA's "Volta" architecture could see stacked DRAM technologies such as HBM2 gain more mainstream exposure, although competing memory standards such as GDDR6 aren't too far behind.
Sources: Commercial Times (Taiwan), TechReport

TSMC to Build New $15.7 Billion Fab in Taiwan, for 3 nm and 5 nm Chips

TSMC (Taiwan Semiconductor Manufacturing Co.), one of the foremost semiconductor producers in the world - which controls a leading 55% share of the global market - said on Wednesday it plans to build a new, $15.7 billion facility in Taiwan that would churn out 5 nm and 3 nm chips. If TSMC were to achieve these production nodes in a timely fashion (with "timely" meaning "before their competitors"), that would prove a huge boon for the company, as everyone - and especially deep-pocketed smartphone chip designers such as Apple and Qualcomm - is looking towards evolution in process nodes, which allows for improvements in power consumption, performance, size and cost of chips per wafer.

"We're asking the government to help us find a plot that is large enough (123 to 197 acres) and has convenient access so we can build an advanced chip plant to manufacture 5 nm and 3 nm chips," TSMC spokesperson Elizabeth Sun said. The spokesperson declined to provide details about the timing of the construction and production, though it's seemingly still a few years away (yet close enough for it to merit an official request). TSMC co-CEO Mark Liu had already mentioned that the company was working on 5 nm chips, and had assigned the task of developing 3 nm technology and conducting research on 2 nm technology to upwards of 300 engineers. Delays on EUV (Extreme Ultraviolet) lithography have slowed expected advancements in further miniaturization of the process nodes. It remains to be seen which technology TSMC is counting on towards aiding them in their goals for 5 nm, 3 nm and the mentioned 2 nm chip production, especially since at those sizes, we start leaving the usual realm of plane old physics, crossing the threshold towards their exotic cousins, quantum physics.

Source: Nikkei Asian Review, Thanks P4-630!

SoC Powering Xbox One S Leverages 16 nm FinFET from TSMC

Microsoft's new slim Xbox One S console achieves its slimness - including its inbuilt power-supply, by significantly reducing thermal load of its key components. This begins at the heart of the console, its SoC. A semi-custom chip by Microsoft and AMD, the SoC powering the Xbox One S is built on the 16 nm FinFET process at TSMC. The chip powering the original Xbox One was built on the same foundry's 28 nm node.

The new SoC isn't merely an optical shrink of the original 28 nm chip down to 16 nm FinFET, Microsoft added a few components to the chip, including an HEVC hardware decoder, hardware CODECs for Blu-ray UHD with HDR; and a revamped display controller with HDMI 2.0 and HDCP 2.2. The chip also performs 1080p to 4K UHD upscaling, with a native upscaling algorithm. The eSRAM memory bandwidth is increased slightly from 204 GB/s from 219 GB/s.

Source: Eurogamer

NVIDIA Accelerates Volta to May 2017?

Following the surprise TITAN X Pascal launch slated for 2nd August, it looks like NVIDIA product development cycle is running on steroids, with reports emerging of the company accelerating its next-generation "Volta" architecture debut to May 2017, along the sidelines of next year's GTC. The architecture was originally scheduled to make its debut in 2018.

Much like "Pascal," the "Volta" architecture could first debut with HPC products, before moving on to the consumer graphics segment. NVIDIA could also retain the 16 nm FinFET+ process at TSMC for Volta. Stacked on-package memory such as HBM2 could be more readily available by 2017, and could hit sizable volumes towards the end of the year, making it ripe for implementation in high-volume consumer products.

Source: WCCFTech

Softbank Acquires ARM for $32 Billion

Japanese conglomerate Softbank acquired British CPU architecture designer ARM in a USD $32 billion deal on Monday. Softbank's bid of $32 billion is a 43 percent premium over ARM's current valuation of $22.3 billion, and the Cambridge-based firm will recommend its shareholders to approve of its acquisition. Shares of ARM surged 45% on the LSE, adding £7.56 billion to its market value. The company reported revenues of $1.49 billion in 2015. ARM founder Herman Hauser, however, isn't happy with the board's decision. "This is a sad day for me and a sad day for technology in Britain," he stated.

ARM designs CPU architectures, which it then licenses to other processor and SoC manufacturers, many of which are fabless themselves, making it an intellectual property giant. None of ARM's products are "tangible" or physical. While Intel, for example, designs CPU architectures (eg: x86), implements it (eg: Core i7, Celeron), and manufactures it (at its Costa Rica and Malaysia fabs) ARM's product is not tangible. It has a CPU architecture, which clients such as Samsung, Huawei, and Qualcomm license, implement (eg: Exynos, Kylin, Snapdragon), and contract-manufacture, through fabs such as GlobalFoundries, TSMC, and ST Microelectronics.

Source: BBC

NVIDIA GP104 "Pascal" ASIC Pictured

Here are two of the first pictures of NVIDIA's upcoming "GP104" graphics processor. This chip will drive at least three new GeForce SKUs bound for a June 2016 launch; and succeeds the GM204 silicon, which drives the current-gen GTX 980 and GTX 970. Based on the "Pascal" architecture, the GPU will be built on TSMC's latest 16 nm FinFET+ node. The chip appears to feature a 256-bit wide GDDR5 memory interface, and is rumored to feature a memory clock of 8 Gbps, yielding a memory bandwidth of 256 GB/s.

Source: ChipHell, AnandTech Forums

TSMC to Begin 7 nm Trial Production in 2017

Taiwan's premier semiconductor foundry TSMC could begin 7 nanometer (nm) trial production in as early as the first half of 2017. Co-CEO Mark Liu, speaking at the company's investor-meet held earlier this month, stated that TSMC is currently engaging with over 20 companies on 7 nm development, with over 15 tape-outs within 2017, leading up to volume-production by early-2018. In the run-up to 7 nm, the company is also developing a 10 nm node for lower-powered devices (eg: mobile baseband). The company has already begun tape-outs of 10 nm chips in Q1-2016. TSMC is currently handling volume-production of 16 nm FinFET Plus chips.

Source: DeliddedTech

NVIDIA "GP104" Silicon to Feature GDDR5X Memory Interface

It looks like NVIDIA's next GPU architecture launch will play out much like its previous two generations - launching the second biggest chip first, as a well-priced "enthusiast" SKU that outperforms the previous-generation enthusiast product, and launching the biggest chip later, as the high-end enthusiast product. The second-biggest chip based on NVIDIA's upcoming "Pascal" architecture, the "GP104," which could let NVIDIA win crucial $550 and $350 price-points, will be a lean machine. NVIDIA will design the chip to keep manufacturing costs low enough to score big in price-performance, and a potential price-war with AMD.

As part of its efforts to keep GP104 as cost-effective as possible, NVIDIA could give exotic new tech such as HBM2 memory a skip, and go with GDDR5X. Implementing GDDR5X could be straightforward and cost-effective for NVIDIA, given that it's implemented the nearly-identical GDDR5 standard on three previous generations. The new standard will double densities, and one could expect NVIDIA to build its GP104-based products with 8 GB of standard memory amounts. GDDR5X breathed a new lease of life to GDDR5, which had seen its clock speeds plateau around 7 Gbps/pin. The new standard could come in speeds of up to 10 Gbps at first, and eventually 12 Gbps and 14 Gbps. NVIDIA could reserve HBM2 for its biggest "Pascal" chip, on which it could launch its next TITAN product.

TSMC Damaged by Earthquake, Could Impact AMD and NVIDIA GPU production

The recent 6.4 magnitude Taiwan earthquake, which hit the island nation on February 6th, affected TSMC worse than expected. Taiwan's premier semiconductor foundry, TSMC, had initially expected semiconductor wafer shipments to be down by less than 1%, but it is now emerging that the drop in shipments could be higher, because the damage to one of its facilities, Fab-14, is worse than originally assessed.

TSMC, in an official communication to its clients, assured that 95% of the foundry machines could return to functionality within 2-3 days after the earthquake. To that effect, machines in Fab-6 and Fab-14B have been fully restored. Despite the disaster, the company appears confident of reaching revenue targets of US $5.9-6.0 billion for Q1-2016. TSMC is the primary foundry partner of major fabless semiconductor companies, such as Qualcomm, NVIDIA, and AMD. AMD recently moved its next-generation GPU manufacturing to Korean silicon giant Samsung, while NVIDIA is building its next "Pascal" GPU family on TSMC's process.

Source: DigiTimes

TSMC to Launch its 5 nm Fab by 2020

Taiwan's premier semiconductor foundry, TSMC, announced that it is on track to begin production of chips on its 7 nanometer silicon fab process by the first half of 2018. The company also announced that production on an even newer 5 nanometer process should commence two years later, in 2020. The company has currently cleared all decks for mass-production of chips on its 16 nm FFC (FinFET compact) node, with the company hoping to grab over 70% of the worldwide 14/16 nm production market-share by the end of 2016.

Source: DigiTimes

Samsung to Fab AMD "Zen" and "Arctic Islands" on its 14 nm FinFET Node

It has been confirmed that Samsung will be AMD's foundry partner for its next generation GPUs. It has been reported that AMD's upcoming "Arctic Islands" family of GPUs could be built on the 14 nanometer FinFET LPP (low-power Plus) process. AMD's rival NVIDIA, meanwhile, is building its next-gen "Pascal" GPU family on 16 nanometer FinFET node, likely at its traditional foundry partner TSMC.

It gets better - not only will Samsung manufacture AMD's next-gen GPUs, but also its upcoming "Zen" family of CPUs, at least a portion of it. AMD is looking to distribute manufacturing loads between two foundries, Samsung and GlobalFoundries, perhaps to ensure that foundry-level teething trouble doesn't throw its product launch cycle off the rails. One of the most talked about "Arctic Islands" GPUs is codenamed "Greenland," likely a successor to "Fiji." Sales of some of the first chips - GPUs or CPUs - made at Samsung, will begin some time in Q3 2016. Some of the other clients for Samsung's 14 nm FinFET node are Apple and Qualcomm. The company plans to speed up development of its more advanced 10 nm node to some time in 2017.

Source: ETNews

NVIDIA GP100 Silicon Moves to Testing Phase

NVIDIA's next-generation flagship graphics processor, codenamed "GP100," has reportedly graduated to testing phase. That is when a limited batch of completed chips are sent from the foundry partner to NVIDIA for testing and evaluation. The chips tripped speed-traps on changeover airports, on their way to NVIDIA. 3DCenter.org predicts that the GP100, based on the company's "Pascal" GPU architecture, will feature no less than 17 billion transistors, and will be built on the 16 nm FinFET+ node at TSMC. The GP100 will feature an HBM2 memory interface. HBM2 allows you to cram up to 32 GB of memory. The flagship product based on GP100 could feature about 16 GB of memory. NVIDIA's design goal could be to squeeze out anywhere between 60-90% higher performance than the current-generation flagship GTX TITAN-X.

Source: 3DCenter.org

NVIDIA "Pascal" GPUs to be Built on 16 nm TSMC FinFET Node

NVIDIA's next-generation GPUs, based on the company's "Pascal" architecture, will be reportedly built on the 16 nanometer FinFET node at TSMC, and not the previously reported 14 nm FinFET node at Samsung. Talks of foundry partnership between NVIDIA and Samsung didn't succeed, and the GPU maker decided to revert to TSMC. The "Pascal" family of GPUs will see NVIDIA adopt HBM2 (high-bandwidth memory 2), with stacked DRAM chips sitting alongside the GPU die, on a multi-chip module, similar to AMD's pioneering "Fiji" GPU. Rival AMD, on the other hand, could build its next-generation GCNxt GPUs on 14 nm FinFET process being refined by GlobalFoundries.

Source: BusinessKorea

TSMC to Commence 10 nm Volume Production by Q4-2016

Semiconductor foundry TSMC assured its clients that the company will be ready with a 10 nanometer manufacturing node for volume production, by the 4th quarter of 2016. Company president and joint-CEO Mark Liu made this announcement during the company's recent Q2-2015 earnings call. "The recent progress of our 10 nanometer technology development is very encouraging and on track with our plan," he said. With volume production of chips commencing in Q4, some of the first products based on them should begin appearing in early-2017. "We ramp up 10 nm in the Q4 2016 next year, but the real product shipment will be in Q1 2017," said C.C. Wei, co-CEO.

Source: Kitguru

NVIDIA Tapes Out "Pascal" Based GP100 Silicon

Sources tell 3DCenter.org that NVIDIA has successfully taped out its next big silicon based on its upcoming "Pascal" GPU architecture, codenamed GP100. A successor to GM200, this chip will be the precursor to several others based on this architecture. A tape-out means that the company has successfully made a tiny quantity of working prototypes for internal testing and further development. It's usually seen as a major milestone in a product development cycle.

With "Pascal," NVIDIA will pole-vault HBM1, which is making its debut with AMD's "Fiji" silicon; and jump straight to HBM2, which will allow SKU designers to cram up to 32 GB of video memory. 3DCenter.org speculates that GP100 could feature anywhere between 4,500 to 6,000 CUDA cores. The chip will be built on TSMC's upcoming 16 nanometer silicon fab process, which will finally hit the road by 2016. The GP100, and its companion performance-segment silicon, the GP104 (successor to GM204), are expected to launch between Q2 and Q3, 2016.

Source: 3DCenter.org

AMD to Skip 20 nm, Jump Straight to 14 nm with "Arctic Islands" GPU Family

AMD's next-generation GPU family, which it plans to launch some time in 2016, codenamed "Arctic Islands," will see the company skip the 20 nanometer silicon fab process from 28 nm, and jump straight to 14 nm FinFET. Whether the company will stick with TSMC, which is seeing crippling hurdles to implement its 20 nm node for GPU vendors; or hire a new fab, remains to be seen. Intel and Samsung are currently the only fabs with 14 nm nodes that have attained production capacity. Intel is manufacturing its Core "Broadwell" CPUs, while Samsung is manufacturing its Exynos 7 (refresh) SoCs. Intel's joint-venture with Micron Technology, IMFlash, is manufacturing NAND flash chips on 14 nm.

Named after islands in the Arctic circle, and a possible hint at the low TDP of the chips, benefiting from 14 nm, "Arctic Islands" will be led by "Greenland," a large GPU that will implement the company's most advanced stream processor design, and implement HBM2 memory, which offers 57% higher memory bandwidth at just 48% the power consumption of GDDR5. Korean memory manufacturer SK Hynix is ready with its HBM2 chip designs.Source: Expreview

Next AMD Flagship Single-GPU Card to Feature HBM

AMD's next flagship single-GPU graphics card, codenamed "Fiji," could feature High-Bandwidth Memory (HBM). The technology allows for increased memory bandwidth using stacked DRAM, while reducing the pin-count of the GPU, needed to achieve that bandwidth, possibly reducing die-size and TDP. Despite this, "Fiji" could feature TDP hovering the 300W mark, because AMD will cram in all the pixel-crunching muscle it can, at the expense of efficiency from other components, such as memory. AMD is expected to launch new GPUs in 2015, despite slow progress from foundry partner TSMC to introduce newer silicon fabs; as the company's lineup is fast losing competitiveness to NVIDIA's GeForce "Maxwell" family.

Source: The TechReport
Return to Keyword Browsing