The never ending story of TSMC's 28nm process

Subject: General Tech | April 17, 2012 - 04:12 PM |
Tagged: TSMC, 28nm

Once again DigiTimes is focusing on TSMC's 28nm process, a node which we have been hearing a lot about for a long time.  Today they are proposing that perhaps the industry has been a bit harsh on TSMC and the availability of working 28nm silicon from that manufacturer.  Their analyst suggests that the 28nm node is still new to the industry and that the production lines still do not have all of the bugs worked out, let alone optimizations to decrease the time it takes to produce a wafer.  This is at least partially true though AMD has been using TSMC for its 28nm Southern Islands GPUs for a while now, other companies have not been so successful in using TSMC.  That seems to have scared other companies, not only is NVIDIA looking elsewhere for chips, Qualcomm is as well.  On the other hand, ARM is trying to get their customers to do the opposite, and are optimising their processors for TSMC's 28nm node, as well as the older 40nm.

DigiTimes may be spot on when they describe TSMC's 28nm process production speeds as increasing faster than previous nodes have and that the problems are only for specific chips and not across the board like the 40nm issues were.  Since TSMC is predicting that they will be running at 95% capacity by the end of the year they had better hope that they can speed production and find a way to do so without having to shut down entire production lines in order to implement any optimizations they discover along the way as any drop in supply is going to be poorly received by customers.

No process transition goes smoothly.  TSMC may be in the news more frequently than other Fabs but those competitors are not without their difficulties as we saw with the limited amount of GLOBALFOUNDRIES produced Llano chips at the end of last year.  Hopefully the current yields do improve, not just for the sake of the GTX 680 but for all of the other customers planning on moving to this node.  In the meantime, it offers a tech-centric soap opera for enthusiast to watch and speculate on.

waffle.jpg

"Taiwan Semiconductor Manufacturing Company (TSMC) could ramp up its 28nm production capacity at a much faster pace than older 40/45nm and 65nm process nodes, according to Digitimes Research analyst Nobunaga Chai. To make such speculation about yield problems with TSMC's 28nm processes is unfair, said Chai, adding that the foundry is actually improving the process yield rate within its expectations."

Here is some more Tech News from around the web:

Tech Talk

 

Source: DigiTimes

ARM aims to make TSMC the Fab of choice for their customers

Subject: General Tech | April 16, 2012 - 01:47 PM |
Tagged: arm, TSMC, fab, cortex a15, cortex-a9, 28nm, 40nm

ARM has developed some optimizations for their chips, provided that the customer purchasing them uses TSMC to fabricate them.  ARM has licensed a large variety of fabrication companies to produce their chips but with their familiarity with TSMC's 28nm and 40nm processes they have been able to introduce performance enhancing optimizations specific to TSMC.  It could taste a bit like favouritism but is much more likely to stem from the volume of TSMC's production as well as the maturity of the 40nm process node.  The 28nm node could be a bit of a problem for ARM as we have seen that TSMC is not having an easy time producing enough good dies for their customers; this is why you cannot buy a GTX 680.  As The Inquirer points out, if ARM wants to make sure their customers can get their hands on reasonable volumes of chips, they will want to create optimizations specific to other manufacturers sooner rather than later.

arm-cortex-a15.jpg

"CHIP DESIGNER ARM has released a slew of optimisation packs for Taiwan Semiconductor Manufacturing Company's (TSMC) 28nm and 40nm process nodes.

ARM, which licenses designs to many chip designers, including Qualcomm, Texas Instruments, Nvidia and Samsung, has given TSMC a boost by offering processor optimisation packs for the firm's 28nm and 40nm process nodes. ARM claims the optimisation packs for its Cortex-A5, Cortex-A7, Cortex-A9 and Cortex-A15 processor cores help designers make use of TSMC's process node nuances to get the most out of their designs."

Here is some more Tech News from around the web:

Tech Talk

 

Source: The Inquirer

Want to stop the GTX 680 leaks? Release the card and they will go away!

Subject: Graphics Cards | March 22, 2012 - 02:15 PM |
Tagged: nvidia, kepler, gtx 680, 28nm

After a fun filled week of speculation, accidental launches and more leaks than a cruise ship, we can finally talk about the GTX 680 and how it performs.  Start out at [H]ard|OCP who put four monitors on a GTX 680 and started benchmarking.  The card is made with TSMC's 28nm process, sports 3.54 billion transistors and both its GPU and 1536 CUDA cores run at the same 1.006GHz and the 2GB of memory is running at 6GHz on a 256-bit bus.  That is enough power to surpass the performance of AMD's HD 7970 and surprisingly it often draws less power than the Radeon making the card more efficient than AMD's offering and ruining NVIDIA's reputation for power hungry, hot running cards.

You can catch not only the print version of Ryan's GTX 680 review but if your timing is good you can catch a recording of the live stream he did earlier today!

The Inquirer is also worth checking out as they offer a dissenting opinon which places the performance of the GTX 680 as lower than an HD 7870, let alone the 7970!

H_680.jpg

"The silicon we have all been waiting for is here. NVIDIA is launching its next generation GeForce GTX 680 video card, poised with the new flagship GPU from NVIDIA. Can it compete with AMD's Radeon HD 7970? We were somewhat amazed at how this NVIDIA GPU does when it comes to gaming, pricing, efficiency, and features."

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

Graphics Cards

 

Source: [H]ard|OCP
Author:
Manufacturer: NVIDIA

The Kepler Architecture

Join us today at 12pm EST / 9am CST as PC Perspective hosts a Live Review on the new GeForce GTX 680 graphics card.  We will discuss the new GPU technology, important features like GPU Boost, talk about performance compared to AMD's lineup and we will also have NVIDIA's own Tom Petersen on hand to run some demos and answer questions from viewers.  You can find it all at http://pcper.com/live!!

NVIDIA fans have been eagerly waiting for the new Kepler architecture ever since CEO Jen-Hsun Huang first mentioned it in September 2010. In the interim, we have seen the birth of a complete lineup of AMD graphics cards based on its Southern Islands architecture including the Radeon HD 7970, HD 7950, HD 7800s and HD 7700s.  To the gamer looking for an upgrade it would appear that NVIDIA had fallen behind; but the company is hoping that today's release of the GeForce GTX 680 will put them back in the driver's seat.

This new $499 graphics card will directly compete against the Radeon HD 7970, and it brings quite a few "firsts" to NVIDIA's lineup.  This NVIDIA card is the first desktop 28nm GPU, the first to offer a clock speed over 1 GHz, the first to support triple-panel gaming on a single card, and the first to offer "boost" clocks that vary from game to game.  Interested yet?  Let's get to the good stuff.

The Kepler Architecture

In many ways, the new 28nm Kepler architecture is just an update to the Fermi design that was first introduced in the GF100 chip.  NVIDIA's Jonah Alben summed things up pretty nicely for us in a discussion stating that "there are lots of tiny things changing (in Kepler) rather than a few large things which makes it difficult to tell a story." 

arch01.png

GTX 680 Block Diagram

Continue reading our review of the new NVIDIA GeForce GTX 680 2GB Graphics Card!!

... and here's Newegg with a GTX 680 leak

Subject: General Tech, Graphics Cards | March 21, 2012 - 09:51 PM |
Tagged: pelly wants revenge, nvidia, leak, gtx 680, fermi, 28nm

Not to be outdone by NCIX, NewEgg also managed to jump the gun on the GTX 680 earlier today. The screengrab that was sent tp Bright Side of News shows their pricing of the soon to be released GTX 680 with models ranging from $500 to $535.  The specs are there for all to see, a GPU running at 1.006GHz, Shader clock of 2.012GHz, effective memory of 6.008GHz and 1536 Stream Processors.  Contrast that with the last GTX 580 that Josh reviewed which had a 782 MHz core, 1.564GHz shader, memory at 4.008GHz and 512 SPs and you can see it is a big step up!

Oh_Newegg.jpg

If you visit NewEgg now you will be greeted with a different result, a page describing the GTX 680's various features and a Buy Now button which unfortunately doesn't work at this moment.  That is a situtation which obviously cannot last as NewEgg would not have put it up.  Of course the realization that you can pick up a pair of GTX 570's for the same price might just mean some recalculations will be in order once we see the performance of the actual card.

2012-03-21_2137.png

Psst ... Hey buddy. Ya wanna see the GTX 680 under my jacket?

Subject: General Tech | March 20, 2012 - 11:26 PM |
Tagged: pelly, nvidia, leak, gtx 680, fermi, 28nm

The gang over at Tweaktown managed to get pictures of a retail Gigabyte GTX 680, which is not only better than candid snaps from Las Vegas making it to the web, it also solidifies a few facts.  For instance, as you can see below there are two 6-pin PCIe power connectors which pegs the maximum supplimental power that this card can draw at 150W.  That is a big difference from the two 8-pin PCIe connectors that could deliver up to 275 to a GTX 580; NVIDIA has obviously made a huge step forward in power savings with the move to 28nm regardless of any design or manufacturing problems they may have had to overcome to deliver this card to retailers.  

TT_23087_02_pictures_of_a_retail_gtx_680_have_been_leaked.png

Tweaktown didn't stop there either GPU fans; it seems that the online\brick and mortar computer chain NCIX made a little mistake and let the GTX 680 appear on their wishlist app.  Both an EVGA and an MSI model of the GTX 680 could be added to your wishlist ... for the price of $578.20 USD plus delivery.  That same retailer currently sells HD 7970's for between $564.99 USD to $619.99.  If only there had been some leaked benchmarks which might indicate which way AMD might have to adjust their pricing.

TT_23087_01_pictures_of_a_retail_gtx_680_have_been_leaked.png

Lucas is so going to sue you!

Source: Tweaktown

TSMC's 28nm process is going to be around for the long haul

Subject: General Tech | March 20, 2012 - 01:12 PM |
Tagged: TSMC, nvidia, amd, southern islands, kepler, 28nm, maxwell, llano

TSMC's 28nm process has been in the news for a long time, sometimes this was a good thing but more often it was not.  Back in May of 2009 the first announcements of TSMC's brand new 28nm process hit the news with major production slated to start in early 2010.  That didn't happen on time, much to several companies dismay as Josh unhappily discussed towards the end of 2010.  This set a trend for TSMC's 28nm process for a while, for instance AMD did not quite meet their promise of readily available 28nm GPUs in 2011, though a late December launch for the HD7970 did meet the spirit of the agreement.  The delays and issues on TSMC's 28nm lines had a variety of causes, perhaps one of the worst being TSMC's overly optimistic attitude about their production capabilities especially when AMD had a surprise for them.  Add to that the long line of woes during the development and production of NVIDIA's 28nm Kepler GPU as well as the recent shutdown of the production line, and you can see why TSMC's 28nm process has spent a lot of time being maligned in the news.  It almost makes you forget about the 40nm process woes, but that is ancient news.

All that effort is not going to waste as DigiTimes reports that TSMC is planning on expanding their 28nm capacity this year and expects that process to account for 10% of their 2012 revenue.  The next question on most peoples minds is the progress on TSMC's 22nm process which in 2010 they announced would be ready by Q3 2012, something which NVIDIA's Maxwell team is probably anticipating with great anxiety.

TSMC.jpg

"With current capacity for 28nm processes filled up, Taiwan Semiconductor Manufacturing Company (TSMC) is likely to expand the leading-edge process capacity later in 2012, according to industry sources.

TSMC reportedly is running at full capacity at its 12-inch fabs due to strong orders for 28nm as well as 40nm and 65nm designs. In order to avoid orders to rivals such as United Microelectronics (UMC) and Samsung Electronics, TSMC will have to speed up the pace of its leading-edge capacity expansion in particular its 28nm capacity, the sources said."

Here is some more Tech News from around the web:

Tech Talk

 

Source: DigiTimes

Alleged NVIDIA GK104 Kepler GTX 670 Ti Photo Leaked

Subject: Graphics Cards | March 8, 2012 - 03:41 PM |
Tagged: nvidia, kepler, graphics card, gpu, GK104, gaming, 28nm

GDC 2012 is upon us, and in addition to the Samaritan demo and some gaming goodness, we spotted a leaked image over at Legit Reviews that is allegedly a photo of a production NVIDIA GTX 670 Ti graphics card.

GTX670Ti.jpg

The cooler looks to cover the whole PCB and be of the blower design, funneling hot air of the of the front of the card and out of the case.  The connectors include two DVI, one HDMI, and one Display Port.  Rumors suggest that the latest NVIDIA cards will be capable of multi-display (>2) from a single card much like AMD cards have been doing for some time.

Not a whole lot is known about the upcoming GK104 "Kepler" GPUs with a good deal of certainty, but we have reported on a few leaks including that the cards will have 2 GB of GDDR5 RAM on a 256 bit memory bus, and that the cards may just be coming out in May. Due to Epic using a working Kepler GPU in their Samaritan demo, that launch date does not sound too far fetched either. On the performance front, there are conflicting rumors; some rumors state that the cards will blow AMD out of the water and other people swear the cards will not be as powerful as the rumors suggest. I suppose we'll find out soon though!

Are you still waiting for NVIDIA's Kepler GPUs or have you jumped on the latest Radeon series?

Maybe that is why TSMC stopped production on 28nm?

Subject: General Tech | March 8, 2012 - 03:09 PM |
Tagged: TSMC, 28nm

The news broke yesterday that TSMC had halted production on 28nm production in mid February for an undisclosed reason.  SemiAccurate thought this strange as the only company that is admitting to problems with TSMC's 28nm process is NVIDIA, but at least production was scheduled to restart by the end of March.  Today from DigiTimes we hear a more positive story about not only the 28nm but also the 20nm process line as TSMC predicts that the demand for chips from their customers could go as high as 95% of their capacity.  Perhaps because of the lack of pressure because of a currently well stocked channel they shut down the line to ensure no preventable problems would prevent them from meeting this high demand?

tsmc_wafer.jpg

"Taiwan Semiconductor Manufacturing Company (TSMC) chairman and CEO Morris Chang has commented that development of 20nm and 28nm processes is progressing well. The market believes TSMC's capacity utilization rate in second-quarter 2012 should be close to 95%.

Chang indicated that 28nm processes will likely contribute 10% of 2012 revenues.

Industry sources pointed out that TSMC's capacity utilization of its 28nm and 45nm processes at its 12-inch wafer plant has not decreased and delivery usually takes 8-10 weeks.

Equipment makers also indicated that TSMC has been ordering equipment and installations will be completed in first-half 2012.

Industry sources added that other Taiwan-based IC design houses have been experiencing increasing orders. The market was pessimistic before the Lunar New Year holidays causing inventory levels to be low. This has induced the recent wave of increased orders."

Here is some more Tech News from around the web:

Tech Talk

 

Source: DigiTimes

NVIDIA is up for a rough year

Subject: General Tech | February 16, 2012 - 12:45 PM |
Tagged: nvidia, 28nm, TSMC, kepler, tegra, tegra 3

If you caught the podcast last night you would have heard Josh discussing NVIDIA's year end financial conference call, otherwise you will have to wait until the 'cast is posted later this week.  Until then you can read SemiAccurate's take on the call here.  There is a lot of news about NVIDIA and none of it is good, from wafer yields to chip demand nothing seems to have gone right for them.  Attempting to move off of their cursed 40nm line and switching to 28nm, NVIDIA has run into big yield problems as in entire wafers having issues as opposed to just some dies being bad.  

Tegra is not doing so well either, with sales of Tegra 2 dropping as we approach the release of Tegra 3, which is getting a lot of bad press.  SemiAccurate refers to the chip as bloated in size as well as being downright expensive to make.  Combine that with the fact that NVIDIA is lagging on A15 adoption and Samsung and Apple turning their backs on Tegra and it doesn't look good for NVIDIA's mobile plans. The one ray of sunshine is that even combined Samsung and Apple do not account for even half of smartphones on the market, so there is still room for NVIDIA and Tegra to grow.

Tegra3_Chip.jpg

"Nvidia seems to be so far ahead of the curve that they are experiencing problems that are unique in the industry. In their recent year end financial conference call, there was enough said to draw some very grim conclusions.

Today’s conference call was a near complete validation of all the things SemiAccurate has been saying about Nvidia. Remember when we asked if Nvidia could supply Apple? Anyone notice the part about dumping early 28nm capacity, and the disappearance of 28nm Fermi shrinks? Remember how 28nm was not an issue for Nvidia, even if their product roadmap slips said otherwise. How well does this mesh with the quotes from Jen-Hsun himself on the topic?"

Here is some more Tech News from around the web:

Tech Talk

 

Source: SemiAccurate