Opinionated - China Chipping Away to Semiconductor Dominance

[Keel]

Forget it man. You're not going to catch up to me with negative rating.

are you serious? look at mine
all 11 +1 (read further) from one pathetic guy - one of the foremost Chinese haters on the forum in which I called him loser, laughed at their ignorance in economics on their complete confusion of "currency peg and currency swap" when they were assuming their "authority" on the subjects, challenged lightly on @Gabriel about his ehinicity, argued that Chinese miltary strength is better than that of France, criticised some right-wing japanese parading in imperial soldiers uniform as losers ...WTF ....He is insulting Chinese all the time while he cant tolerate the tiniest bit of riposte and what's more, he thinks his horrendous English standard is good enough to give lecture on Chinese posters from time to time. What a joke!

And a double ( one from the pathetic guy abovesaid ) and one from the japanese member about the same comment in which I said the indian judge who had given innocent verdicts to the japanese war criminals in the War Crimes Tribunal was a disgrace to human race and the judge was revered in japan

I didnt even bother to lodge complaints to the mods. It is a failing system wherein a lot of unqualiified and poor quality "professional" " think tanks" like those above are abusing it like gangsters running amok in the streets chopping up people whom they dislike

 

[Bussard Ramjet]

SMIC Expects First 28nm Revenue by Year End
51.1% of SMIC's revenue now comes from China

Alan Patterson

8/12/2015 00:00 AM EDT
2 comments


SMIC HQ in Shanghai


The company will start commercial production of its most advanced technology node in the third quarter and expects its first revenue contribution from 28nm in the fourth quarter of 2015, SMIC CEO Tzu-Yin Chiu said on a conference call to announce the company’s second-quarter results.

SMIC has been planning to make Qualcomm’s Snapdragon processors with its 28nm process. While Qualcomm wrestles with a business downturn, SMIC said it is engaged with four customers in TVs, set-top boxes and other consumer applications that plan to make chips on SMIC’s latest technology node.

The company said it had to turn some customers away during the second quarter as its utilization rate exceeded 100%. SMIC, whose top rival is Taiwan Semiconductor Manufacturing Co (TSMC), said that it has benefitted from strong demand in China, which for the first time accounted for more than half of SMIC’s revenue during the second quarter this year.

“SMIC has achieved two quarters of consecutive growth in 2015, and we are guiding an additional quarter of growth for the third quarter,” SMIC CEO Chiu said. “With some customers undergoing inventory adjustments, SMIC has successfully ramped up new products, keeping our fabs well utilized.”

While overall smartphone growth has been slowing, SMIC said some Chinese handset customers have been increasing their market share.

SMIC’s sales to customers in China rose consecutively in the second quarter to 51.1% of revenue, while sales to North America dropped by about 9 percentage points to 32.0% during the same period.

The company said it expects sales to North America to improve as it ramps up the 28nm process.

While SMIC forecast that its third quarter sales will rise by as much as 3% from the $546.6 million the company recorded in the second quarter, it declined to offer an outlook for financial results in the final quarter of this year.

The company said it will ramp up two fabs in China during the second half of 2015 to meet customer demand. As a result, SMIC said its utilization rate is likely to drop to the high 90% range starting in the third quarter.

The company reiterated that its capital expenditure for 2015 foundry operations will be $1.5 billion, part of which will come from the Chinese government. SMIC’s planned capex is about a tenth that of TSMC, which in April cut its 2015 capex budget by $1 billion to fall within a range of $10.5 billion and $11 billion.

Playing Catch Up
SMIC needs to catch up with leading foundry rivals such as Samsung and TSMC, which are ramping 14nm and 16nm FinFET technology this year. TSMC has dominated the 28nm node for nearly five years.

SMIC’s most advanced technology node in the second quarter, 40/45nm, accounted for 15.3% of its total revenue, down from 16.0% in the first quarter this year. Its second-quarter revenue from 55/65nm products also dropped to 25.2% from 26.1% in the first quarter, while the company saw sales gains in less advanced 90nm and 0.13um products.

SMIC is planning to make its first 14nm FinFET products by 2020, helped by a Chinese government initiative to boost the domestic semiconductor industry. The company said it may be possible to achieve the target sometime ahead of 2020.

The world doesn’t stand still. In July, TSMC said it is on track with 10 nm development for a volume ramp in the fourth quarter of 2016. The company reiterated expectations for its 7 nm risk production to start in the first quarter of 2017.

 

[cirr]

Tsinghua arm readies to take on Qualcomm

2015-08-12 10:38

China Daily Editor: Si Huan

Tsinghua Holdings Co Ltd, a technology conglomerate backed by Tsinghua University, plans to invest at least 30 billion yuan ($4.76 billion) in developing mobile chip technology, highlighting the company's ambition to challenge Qualcomm Inc's dominance in the country's chip market.

"To catch up with Qualcomm as soon as possible, we will pour 30 billion yuan, and probably even more, into the research and development of mobile chips in the next few years," Xu Jinghong, chairman of Tsinghua Holdings, told China Daily in an interview on Tuesday.

Tsinghua Holding said a certain proportion of the money will come from government funding and its partners. In February, its unit Tsinghua Unigroup Ltd said it has received 10 billion yuan from governments to invest in chip companies.

"Frankly, compared with global competitors, we are still three-to-five years behind in technology, especially in cutting-edge 4G and 5G products," Xu said, "but if we don't close the technological gap, we will never win."

The comment came after Tsinghua Unigroup Ltd filed a plan to buy US chipmaker Micron Technology Inc for $23 billion. Xu said earlier it was still in discussions for a potential deal.

"We will continue to expand our presence in the integrated circuit industry through both acquisition and self-research," he said, adding chips will be one of the focuses of the Beijing-based company.

Roger Sheng, senior analyst at research firm Gartner Inc, said the government's emphasis on the semiconductor sector is the biggest advantage for Tsinghua Unigroup.

"Few tech companies in China have such a large amount of capital at their disposal as Tsinghua Holdings does," Sheng, said adding "despite its current technological weakness, it has ambitions and is acting very quickly".

"Qualcomm's pioneering efforts in the sector also established a successful business patten which Tsinghua can follow," Sheng said.

Tsinghua Holdings' intensified efforts to boost its chip-related resources come as the Chinese government seeks to reduce the country's reliance on foreign technology, on worries that it may hurt national security.

The company evolved into the largest chip firm in China after it acquired Spreadtrum Communications Inc, the world's third-largest mobile phone chip maker, and RDA Microelectronics Inc, the fourth-largest, in 2013.

"If we can manage to catch up with Qualcomm technologically, our innovation capability, the huge smartphone market in China as well as the labor cost, which starts rising but is still lower than that of the US, can offer us considerable commercial opportunities," Xu said.

Earlier this month, Tsinghua Holdings announced it has made a breakthrough in chemical mechanical polishing, an important process in manufacturing chips. One of its unit successfully developed the first 12-inch polishing machine in China which could planarize semiconductor wafers to an extent that every square of a nanometer (billionths of a meter) is flat.

"The machine shows that we are the first Chinese company to have mastered the technology, and enables us to produce extremely tiny chips for smart wearable gadgets," said Li Zhongxiang, vice-president of Tsinghua Holdings.

Tsinghua arm readies to take on Qualcomm

 

[Keel]

SMIC Expects First 28nm Revenue by Year End
51.1% of SMIC's revenue now comes from China

Alan Patterson

8/12/2015 00:00 AM EDT
2 comments


SMIC HQ in Shanghai


The company will start commercial production of its most advanced technology node in the third quarter and expects its first revenue contribution from 28nm in the fourth quarter of 2015, SMIC CEO Tzu-Yin Chiu said on a conference call to announce the company’s second-quarter results.

SMIC has been planning to make Qualcomm’s Snapdragon processors with its 28nm process. While Qualcomm wrestles with a business downturn, SMIC said it is engaged with four customers in TVs, set-top boxes and other consumer applications that plan to make chips on SMIC’s latest technology node.

The company said it had to turn some customers away during the second quarter as its utilization rate exceeded 100%. SMIC, whose top rival is Taiwan Semiconductor Manufacturing Co (TSMC), said that it has benefitted from strong demand in China, which for the first time accounted for more than half of SMIC’s revenue during the second quarter this year.

“SMIC has achieved two quarters of consecutive growth in 2015, and we are guiding an additional quarter of growth for the third quarter,” SMIC CEO Chiu said. “With some customers undergoing inventory adjustments, SMIC has successfully ramped up new products, keeping our fabs well utilized.”

While overall smartphone growth has been slowing, SMIC said some Chinese handset customers have been increasing their market share.

SMIC’s sales to customers in China rose consecutively in the second quarter to 51.1% of revenue, while sales to North America dropped by about 9 percentage points to 32.0% during the same period.

The company said it expects sales to North America to improve as it ramps up the 28nm process.

While SMIC forecast that its third quarter sales will rise by as much as 3% from the $546.6 million the company recorded in the second quarter, it declined to offer an outlook for financial results in the final quarter of this year.

The company said it will ramp up two fabs in China during the second half of 2015 to meet customer demand. As a result, SMIC said its utilization rate is likely to drop to the high 90% range starting in the third quarter.

The company reiterated that its capital expenditure for 2015 foundry operations will be $1.5 billion, part of which will come from the Chinese government. SMIC’s planned capex is about a tenth that of TSMC, which in April cut its 2015 capex budget by $1 billion to fall within a range of $10.5 billion and $11 billion.

Playing Catch Up
SMIC needs to catch up with leading foundry rivals such as Samsung and TSMC, which are ramping 14nm and 16nm FinFET technology this year. TSMC has dominated the 28nm node for nearly five years.

SMIC’s most advanced technology node in the second quarter, 40/45nm, accounted for 15.3% of its total revenue, down from 16.0% in the first quarter this year. Its second-quarter revenue from 55/65nm products also dropped to 25.2% from 26.1% in the first quarter, while the company saw sales gains in less advanced 90nm and 0.13um products.

SMIC is planning to make its first 14nm FinFET products by 2020, helped by a Chinese government initiative to boost the domestic semiconductor industry. The company said it may be possible to achieve the target sometime ahead of 2020.

The world doesn’t stand still. In July, TSMC said it is on track with 10 nm development for a volume ramp in the fourth quarter of 2016. The company reiterated expectations for its 7 nm risk production to start in the first quarter of 2017.

Congrats to our Taiwanese brothers

 

[Bussard Ramjet]

Congrats to our Taiwanese brothers

Well they are not your brothers until you take them over, which is arguably not in near sight.

So what I posted this article was to tell that TSMC is planning to reach 7nm in 2017, while SMIC is planning to reach 16 nm by 2020.

The gap seems to be increasing.

 

[Keel]

Well they are not your brothers until you take them over, which is arguably not in near sight.

So what I posted this article was to tell that TSMC is planning to reach 7nm in 2017, while SMIC is planning to reach 16 nm by 2020.

The gap seems to be increasing.

Jealous comments ignored!

 

[FairAndUnbiased]

Well they are not your brothers until you take them over, which is arguably not in near sight.

So what I posted this article was to tell that TSMC is planning to reach 7nm in 2017, while SMIC is planning to reach 16 nm by 2020.

The gap seems to be increasing.

The gap between the haves and the have nots is indeed increasing:

1980: neither China nor India could make integrated circuits.
1990: India still could not make integrated circuits, China could make 1970's US level.
2000: India still could not make integrated circuits, China could make 1990's US level.
2010: India still could not make integrated circuits, China could make 2005 US level.
2015: India still could not make integrated circuits, China could make 2010 US level.
2020: Think India will start making integrated circuits?

 

[Shotgunner51]

SMIC needs to catch up with leading foundry rivals such as Samsung and TSMC, which are ramping 14nm and 16nm FinFET technology this year. TSMC has dominated the 28nm node for nearly five years.

SMIC is planning to make its first 14nm FinFET products by 2020, helped by a Chinese government initiative to boost the domestic semiconductor industry. The company said it may be possible to achieve the target sometime ahead of 2020.
The world doesn’t stand still. In July, TSMC said it is on track with 10 nm development for a volume ramp in the fourth quarter of 2016. The company reiterated expectations for its 7 nm risk production to start in the first quarter of 2017.

Sportsmanship spirit in tech advancement, in business.
Congrats to Taiwanese bros of TSMC! @TaiShang
Wish SMIC catch up fast with the league of TSMC, Samsung and Hynix!

 

[hans]

Actually 28 nm is OK for most chip application...
The better process is mainly for high performance CPU with lower power consummation.
Like CPU in computer and smartphones.
Other application like SOC is enough with current process.
Also most company listed here are design house for SOC with special application.
So the gap will not be so big.

Well they are not your brothers until you take them over, which is arguably not in near sight.
So what I posted this article was to tell that TSMC is planning to reach 7nm in 2017, while SMIC is planning to reach 16 nm by 2020.
The gap seems to be increasing.

 

[cirr]

SMIC is doing exceedingly well revenue and profit wise，in a market which has seen many semiconductor majors‘ earning ability collapse。With the help of the government，SMIC’s immediate focus should be on developing advanced technology、building more fabs and ramping up production through a doubling or even tripling of annual capex in the next couple of years。

SMIC’s Q2 net profit up 35% at US$76.7m

By Zhu Shenshen | August 13, 2015, Thursday

HONG Kong-listed Semiconductor Manufacturing International Corp, said yesterday that its second-quarter net profit grew 35 percent year on year to US$76.7 million, as surging domestic demand drove its revenue to a record high.

China’s leading contract chipmaker said in a filing to the stock exchange that its revenue increased 6.9 percent in the period to US$546.6 million.

It has benefited significantly from a decision by the central government to add the semiconductor industry to its national development strategy.

China sales accounted for 51 percent of SMIC’s revenue in the period, up from 44 percent in the second quarter of 2014.

“This is a breakthrough for SMIC,” said Gu Wenjun, chief analyst at Shanghai-based consulting company iCwise.

“The national strategy (on semiconductor development) will continue to benefit it going forward,” he said.

The company’s domestic customers comprise mostly “fabless” semiconductor companies, which design chips, but outsource their fabrication to foundries like SMIC.

The company said in the filing that it expects its revenue to grow by between 1 and 3 percent in the third quarter.

SMIC’s Q2 net profit up 35% at US$76.7m | Shanghai Daily

 

[Bussard Ramjet]

SMIC is doing exceedingly well revenue and profit wise，in a market which has seen many semiconductor majors‘ earning ability collapse。With the help of the government，SMIC’s immediate focus should be on developing advanced technology、building more fabs and ramping up production through a doubling or even tripling of annual capex in the next couple of years。

SMIC’s Q2 net profit up 35% at US$76.7m

By Zhu Shenshen | August 13, 2015, Thursday

HONG Kong-listed Semiconductor Manufacturing International Corp, said yesterday that its second-quarter net profit grew 35 percent year on year to US$76.7 million, as surging domestic demand drove its revenue to a record high.

China’s leading contract chipmaker said in a filing to the stock exchange that its revenue increased 6.9 percent in the period to US$546.6 million.

It has benefited significantly from a decision by the central government to add the semiconductor industry to its national development strategy.

China sales accounted for 51 percent of SMIC’s revenue in the period, up from 44 percent in the second quarter of 2014.

“This is a breakthrough for SMIC,” said Gu Wenjun, chief analyst at Shanghai-based consulting company iCwise.

“The national strategy (on semiconductor development) will continue to benefit it going forward,” he said.

The company’s domestic customers comprise mostly “fabless” semiconductor companies, which design chips, but outsource their fabrication to foundries like SMIC.

The company said in the filing that it expects its revenue to grow by between 1 and 3 percent in the third quarter.

SMIC’s Q2 net profit up 35% at US$76.7m | Shanghai Daily

If you go a bit further up, I already posted this information.

 

[cirr]

If you go a bit further up, I already posted this information.

SMIC will leapfrog 22nm node and be only 1-1.5 generation behind the top players by 2018 or 2019.

SMIC’s 28nm Process Puts the Pressure on UMC

Peter Brown

12 August 2015

Semiconductor Manufacturing International Corp.’s (SMIC) has announced that its 28 nanometer (nm) process technology has been successfully adopted into mainstream smartphones. This marks a significant step in the commercial usage of 28nm core chips and a new era of advanced mobile phone chip manufacturing in China.

That puts SMIC in a good position to appeal to the Chinese government’s demand to increase domestic electronics content (Read: Why China Is Shopping for Silicon Valley Chip Companies). However, it also places a heavy burden on competing foundry United Microelectronics Corp. (UMC) to stave off SMIC as the third overall global foundry, according to IHS.

While SMIC is still several generations behind the likes of Taiwan Semiconductor Manufacturing Corp. (TSMC) and Globalfoundries that are both qualified for 20nm and 14nm production, the Chinese foundry is only slightly behind UMC, which has been in volume production of 28nm chips for six quarters.

Len Jelinek, senior director of semiconductor manufacturing at IHS.“The key question is can [SMIC] surpass UMC in terms of 28nm revenue?” ask Len Jelinek, senior director of semiconductor manufacturing at IHS. “I believe with the pressure being placed by the Chinese government that within a year or so, SMIC may be able to significantly increase their revenue on 28nm and possibly surpass UMC’s revenue,” he says.

Jelinek says the SMIC process technology is a big deal since the company had to develop the technology independently, as Taiwan is not allowed to transfer technology to mainland China. While SMIC is the only foundry currently at this manufacturing process node, it may not be for long as Shanghai Huali Microelectronics Corp. is also working to develop and qualify 28nm technology.

SMIC’s 28nm process technology is being used initially to manufacture Qualcomm’s Snapdragon 410 processors for use in smartphones. While SMIC claims this is a pivotal step toward advancing mobile phone chip manufacturing in China, Qualcomm could elect to use the chips globally as well. However, Jelinek believes this might not be the case.

“Given the strong desire of the Chinese government to have domestic suppliers, I would think that Qualcomm would preferentially provide these chips to domestic suppliers,” he says.


Qualcomm and SMIC have developed the first 28nm process in China. Source: SMIC

Dr. Tzu-yin Chiu, CEO and executive director at SMIC, says in a statement that for the first time China’s mainland manufacturers will be able to introduce mainstream smartphones that are the result of the close collaboration between a Chinese foundry and smartphone chip maker built domestically.

Chiu says SMIC will work on developing more advanced processing technologies in the future in order to meet the demand from the domestic mainland. In fact, Qualcomm and SMIC have already formed an agreement—along with research institute IMEC and wireless technology giant Huawei—to develop a 14nm process (Read: SMIC, Qualcomm, IMEC in 14nm Process Development

SMIC’s 28nm Process Puts the Pressure on UMC - IHS Electronics360

 

[Jlaw]

The gap between the haves and the have nots is indeed increasing:

1980: neither China nor India could make integrated circuits.
1990: India still could not make integrated circuits, China could make 1970's US level.
2000: India still could not make integrated circuits, China could make 1990's US level.
2010: India still could not make integrated circuits, China could make 2005 US level.
2015: India still could not make integrated circuits, China could make 2010 US level.
2020: Think India will start making integrated circuits?

So what? India has democracy. Democracy is so great that many Indians if given the chance would leave their country.

 

[cirr]

As the old saying goes：它山之石可以攻玉

Spreadtrum Guns for Intel’s 14nm FinFET in 2016

Junko Yoshida

5/27/2015 00:28 AM EDT

SHANGHAI, China — Qualcomm and MediaTek, you better watch out. Here comes Spreadtrum, riding piggyback on Intel’s foundry business and gunning for 14nm finFET, with sights set on 10nm.

China’s Spreadtrum Communications will use Intel Corp.’s 14nm finFET process technology, for both the low- and high-end mobile chips the company plans to launch in 2016, Leo Li, chairman and CEO of Spreadtrum, told EE Times Tuesday (May 26).

For Spreadtrum, using Intel as its foundry has apparently superseded its potential adoption of Intel Architecture in future mobile chips.

Intel’s $1.5 billion investment in Tsinghua Unigroup last fall resulted in the U.S. chip giant owning 20 percent of China’s combined Spreadtrum Communications and RDA Microelectronics. Spreadtrum’s quid pro quo for Intel, under the agreement, is a matter of intense speculation among semiconductor industry observers.

Li, during the one-on-one interview, insisted, “I am under no obligation” to use Intel technologies “unless they prove to be competitive on the market.”

Nowhere in a series of agreements the two companies signed last fall is it stipulated that Spreadtrum must switch from ARM-based architecture to Intel Architecture in future chips.

“They can’t force us,” said Li. But that’s not to say that Li isn’t interested in a war chest full of Intel’s technologies. “Intel is a great company. It really has a lot to offer.”


Leo Li, Chairman and CEO of Spreadtrum

External force

Li sees Spreadtrum’s role as “an external force” to change Intel’s culture and mentality. He believes Spreadtrum can help commercialize a lot of technologies that haven’t gone beyond Intel’s R&D lab.

Li holds the view that the U.S. CPU behemoth is too accustomed to “being served” by others in the industry, rather than serving them. “BK (Intel CEO Brian Krzanich) understands this. But not everyone [inside Intel] gets it,” said Li. He indicated that Intel is “a professional company” but sometimes it moves “too slow.”

For now, Spreadtrum will use Intel’s mobile chip SoFIA to gauge “customer engagement,” said Li. Intel designed SoFIA for a super-cheap smartphone last year. Packed with a dual-core Atom Silvermont processor, the SoC also crams in a modem. Spreadtrum will be selling SoFIA with a 4G modem, in addition to Spreadtrum’s existing product lines. They include single-core and quad-core ARM-based 3G mobile chips; and quad-core and octa-core based 4G mobile processors.

“If our customers like SoFIA, I’d take it,” said Li. But Spreadtrum testing Intel’s chip for “customer engagement” doesn’t sound like a ringing endorsement.

Whatever happened to the Intel Architecture-based SoCs that Spreadtrum and Intel were supposedly developing together? They were scheduled for rollout in the second half of 2015, but the project has apparently slipped.

Before the end of this year, the new SoC that will actually come out of Spreadtrum will use ARM-based octa-cores. That processor will be made by using Taiwan Semiconductor Manufacturing Co.’s 16nm process technology.

Spreadtrum Guns for Intel's 14nm FinFET in 2016 | EE Times

 

[cirr]

Phytium? A new kid on the block? We will soon hear more about this company and its CPU for HPCs.

Phytium to detail 64-core ARMv8-A processor for HPC in August

May 1st, 2015 at 4:52 pm - Author Anton Shilov

The U.S. government recently banned sales of Xeon Phi co-processors for high-performance computing applications to numerous supercomputer sites in China, slowing down development of local research. But perhaps, China will not need any HPC chips from the world’s renowned suppliers in the future as domestic companies are developing a number of promising microprocessors.

Phytium, a startup processor developer from Guangdong, China, is working on a special 64-core central processing unit specifically designed for high-performance computing applications. The company is going to present the supercomputer chip at the Hot Chips technology conference in August, which indicates that the CPU is either ready and is about to be released, or its development is largely completed.



Nothing particular is known about Phytium’s upcoming CPU. The company itself describes it as “a 64-core ARMv8 processor for HPC”, which essentially means that this is 64-bit with many cores. However, this product is not that simple. In fact, Phytium’s CPU appears to be a highly-custom supercomputer chip that contains a lot of truly innovative technologies developed in-house.

Modern ARM cores allow to build SMP [symmetric multiprocessing] quad-core clusters as well as create many-core chips featuring up to 12 coherent SMP processor clusters through CoreLink CCN, AMBA 5 CHI, AMBA 4 ACE as well as other technologies. The total number of cores within a modern ARMv8-A system-on-chip cannot exceed 48 units if the SoC designers uses off-the-shelf technologies designed by ARM. Developers, who plan to integrate more than 48 ARM Cortex-A57 or ARM Cortex-A72 chips, will have to design their own high-speed on-chip ring-bus communication and interconnection technologies that support higher amount of processing engines (which may involve redesign of SMP clusters and necessity to develop special commutators for every cluster, or a group of cores, depending on exact approach to design).


An example of ARM-designed 48-core ARMv8-A microprocessor

Many-core chips also require sophisticated memory controllers and extreme memory bandwidth (i.e., they need to feature multiple controllers that can support high-speed memory). Such memory sub-systems are hard to design. ARM offers quite advanced DMC-520 controllers with enterprise-class class RAS (reliability, availability and serviceability) features such as ECC for x72 DRAM, TrustZone security and end to end Quality of Service (QoS). Unfortunately, the DMC-520 in the current implementation supports up to quad-channel DDR4 3.20GHz memory providing up to 102.4GB/s of bandwidth, which may not be enough for a 64-core HPC processor. It should be possible to install more than four DMC-520 controllers inside a chip, but it requires a lot of additional work. Developers of ARMv8-A-based processors for server and HPC applications, such as AMD or Cavium, use dual-channel or quad-channel memory controllers for their Opteron 1100A “Seattle” and ThunderX system-on-chips.

Apparently, Phytium has managed to design an advanced ring-bus interconnect technology, a custom L3 cache sub-system with snoop filter, a sophisticated memory sub-system and a lot of other unique things necessary for many-core microprocessors.

Unfortunately, nothing particular is known about the cores that Phytium plans to use inside its 64-core HPC chip. It is possible that the company will use off-the-shelf ARM cores, such as Cortex-A57, but it is also possible that the company has created a custom implementation of ARMv8 architecture. If the company is designing its chip for supercomputers, it has to be competitive against things like Intel Xeon Phi or Nvidia Tesla, hence, we are talking about ~0.5~1.0 TeraFLOPS-class double precision floating point performance per processor. It is, of course, possible that the chip will be slower because of architecture limitations or design trade-offs (power vs. throughput).

Phytium is not the only Chinese developer working on microprocessors for supercomputers. However, it will be the only CPU developer from China to present at this year’s Hot Chips conference. Since the symposium is about leading-edge technologies, it is logical to assume that Phytium is to unveil something very advanced.

Phytium is a part of Guangzhou Higher Education Mega Center, an area featured by higher education institutions, located on Xiaoguwei Island in Panyu District, Guangzhou, China.

Phytium to detail 64-core ARMv8-A processor for HPC in August | KitGuru