Search the Community

According to Cleveland.com Intel is planning on building a semiconductor manufacturing plant in New Albany Ohio,.I can't wait for it to get up and running and begin churning out chips

Sad days, I had to RMA my Dead AMD Vega64 which still had a 3 year warranty with Gigabyte. The RMA was done via the online store I bought it from here in South Africa. So they sent me feedback ... In todays computer market, this kinda seems like im just being thrown out to fend for myself in the impossible GPU market. 1. They have no stock at all of any GPU in their store, only a GT710 lol 2. Even if I added extra money to the 5000 South African Rand they are offering as credit note. GTX1660 Super would cost R8000 (out of stock) and RTX3060 would cost R12500 (out of stock). Why does the consumer have to suffer when the manufacturer cannot fulfill their warranty, even with the Semiconductor shortage? All I really want is card similar to my Vega64 Thanks for coming to my ted talk I guess. ____________________________________________________________ ***EDIT*** They phoned me to offer a GTC 1660 Super if I pay a the difference , so they are atleast giving me another option now. Does anyone know if the 1660 Super is good enough for Twitch stream + VR Gaming like Beat Saber?

What advantages does Samsung have over TSMC on their 8nm manufacturing process other then the fact that TSMC is at maximum capacity? How factual that Nvidia's future RTX 30 series lineup refresh would be manufactured by TSMC instead of Samsung?

Summary New semiconductor fabrication plants in the USA (Intel, TSMC, Samsung) will have a hard time finding skilled employees to staff their plants when they are operational. The lack of a 'pipeline' and the competition and attractiveness of SWE doesn't make the job any easier for the industry. Quotes My thoughts I found this story, attributing the skill shortage to a lack of 'exposure' et al. interesting, after Linus mentioned that ALL the major semiconductor makers declined his request for a tour of their 'cool' fabrication plants and the potential for free PR & marketing. The other thing ('pipeline') is that it needed a crisis on several fronts, and an act of congress (loads of government money), to patch over some gaping holes in America's comparative competitiveness, and the semiconductors talent 'pipeline' (Vocational Education and Training). Compare this to mainland Europe and how close Universities and their leading industry companies, big and small (Mittelstand), collaborate to maintain their basic science lead and skills availability. And the entrenchment and history of VET in the European economy. Sources Protocol: New chip fabricating plants will need tens of thousands of skilled workers who don’t currently exist. Training them means persuading students to look away from jobs at big tech companies.

Samsung's semiconductor division has beat Intel in Q2 2017 earnings: With Ryzen, Windows on ARM with Qualcomm, and now this, it looks like Intel is going to have a rough year... Via: Engadget Source: Samsung Earnings, Intel Earnings (PDF)

Now this news isn't necessarily as sexy as vega , threadripper or volta news , but I still felt that it was important to share , as it has implications for the industry as a whole . For those who don't know , a quick recap: EUV is the next generation of lithography . Modern lithography uses a 193nm light source for etching , but the wave length of the light itself is starting to cause issues , forcing manufacturers to result to multi-patterning ( etching some parts of the chip multiple times ) , considerably driving up cost . This is one of the main reasons semiconductor costs have been rising since 28nm . EUV has a wavelength of 13.5nm , and should alleviate these issues , however it has been met with a lot of delays . One of the big issues is that wafer throughput has been very low on early light sources because ASML ( and it's competitors ) couldn't manage to increase the power of the light sources in their machines . This low throughput made EUV cost-prohibitive for high volume manufacturing . For a long time , it has been said that a light source of 250W or more would be needed to make EUV a viable alternative to standard 193nm ArF lithography . source: https://seekingalpha.com/article/4087896-critical-euv-milestone-implications-semiconductors ( the article goes on several pages , for those interested in the subject ) Here are a few videos explaining ( briefly ) how semiconductors are made and processed :

http://www.globalfoundries.com/newsroom/press-releases/2016/09/07/globalfoundries-extends-fdx-roadmap-with-12nm-fd-soi-technology Sorry, on phone and can't use quotes. Will edit when I can. It seems the goal of this node is to be a jack-of trades node that isn't as costly as FinFET design (though the cost of FDSOI substrate wafers is far higher than standard bulk silicon, so there's probably some creative booking to be able to say this). I thought IBM's Power 9 would be based on a 14nm FDSOI design, but with the chips due in another year, I'm now wondering if IBM will use GloFo's 14nm LPP or HP process instead or just stick with 22nm FDSOI. Interesting times we live in...

I might be a little selfish for this topic but I need help from you guys to answer this question for my business assignment. What is AMD doing in order to win over the market giants Intel and Nvidia? What strategies or factors are they using to win and how they are using it? Also give examples when possible. I am in quite a pinch so it would be great if you guys could help me out.

http://www.extremetech.com/computing/208618-graphene-sheathed-copper-wires-could-dramatically-speed-future-processors-cut-power-consumption http://phys.org/news/2014-05-physicists-unlimited-graphene.html Stanford scientists have recently made note that in semiconductors, the copper wiring interconnects deep in the bowels of your chips are wrapped in a layer of Tantalum Nitride which serves 2 functions. 1) Ensure the copper does not flow where it is unwanted during the manufacturing process 2) Provide a support and cooling structure around the copper. To put it simply, heat generated in any resistive circuit (anything that isn't a superconductor) is directly proportional to the current^2 * resistance of the circuit. The resistance of a wire is inversely proportional to its diameter. As wires get thinner, resistance grows, heat generation increases, and heat dissipation abilities are lost, hence the Tantalum Nitride layer. All the major semiconductor companies are having issues reducing the size of their metal layers because of these effects which also effect clock rate ceilings. Stanford has found a way to use Graphene to replace this layer, and constructing the graphene sheaths is actually easy from a manufacturing standpoint compared to trying to produce uniform sheets of the stuff, and a perfect lattice isn't required. A single support layer of graphene would be 1/8 the thickness of Tantalum Nitride, but what wasn't covered in the ExtremeTech article is even more interesting. Graphene has a greater than ideal thermal dissipation property. Due to quantum mechanical effects beyond my education level in the field, graphene conducts heat better the longer the path between the source and drain becomes, a behavior contrary to every other material on earth (to my knowledge). In other words, the effort needed to keep a CPU or GPU cool may take a nose dive in the next five years. Furthermore, because of this heat dissipation ability, we should see semiconductor companies being better able to shrink their metal layers again, meaning ever more transistors for you and me. One last bit from the Extremetech article.

Hey all, I'm currently researching new semiconductor materials for (eventual) use in computer and related technology (additional applications include solar cells, LEDs, optoelectronics, etc). However, my knowledge of computers is limited to what an entry level computer technician would know; I can build computers, work with peripherals, and explain how different parts of a computer work about 1/2 as well as Linus can, but that isn't sufficient for my work. What I want to know is, does anyone here bridge the gap on that knowledge base? Going from fundamental semiconductor knowledge to computer semiconductor technology? I really feel like I could improve my direction in research materials science if I knew how my semiconductors applied to RAM/Cache/bus speeds/processors/etc. Because my eventual goal would be to take something like the current cost prohibitive cache systems and make them as cheap as standard RAM (not entirely possible, but it can be much closer than it is). However, it'd be nice to hear someone say that it's actually worth the time. So do you think it is? If I could make a material that made cache 5-10 times cheaper, would we dump that much more memory on processors? Or is there another reason in addition to cost that it's prohibitive to just drop memory on the processor directly? I'm just as happy to hear that it's not going to make a significant improvement in tech, because that means I can move on and spend less time on the topic. But in general, I'd like to start some materials science talks here on what materials go into each computer component and why they're selected! I think that would be an awesome topic here. Cheers! A physics grad student

I just want to give a shout out to the Newegg team, because i first saw this on their channel. Scientists led by Yoke Khin Yap at Michigan State University have developed transistors that do not need to use silicon or any other semiconductor material. they instead used golden balls 3nm in diameter on a nanoinsulator. (no idea what a nanoinsulator is by the way) There are several possible advantages to this system over traditional processors. First of all it can be run at room temperature. Second of all die sizes could be shrunk significantly smaller than silicon. There is also no leakage, which is one of the main reasons modern processors get so hot and consume so much energy. http://www.dailytech.com/New+LeakageFree+Nanotube+Quantum+Dot+Transistor+Uses+No+Semiconductors/article31832.htm What do you guys think? Is this actually a step forward in processor technology? EDIT: whoops i mean Michigan Tech not Michigan state