Looking at the pie chart breakdown, I'm wondering what's special about the 150/180nm processes that they're still 8% of revenue when the newer and smaller 90/110/130nm ones have much smaller residual shares.
Looking at how quickly it's share has imploded since 7nm came out, I'm also wondering if 10nm will be going away entirely in the next year or two.
"what's special about the 150/180nm processes that they're still 8% of revenue when the newer and smaller 90/110/130nm ones have much smaller residual shares."
It's highly likely those geometries are 8 inch wafers vs. others being 12 inch wafers, which means wafer starts for those geometries are low-cost and low-risk. #speculation
A company I worked for a few years ago had an ancient 4-inch wafer fab that made the sensor elements for a sensor we developed. I never asked what size process it was, but it was probably somewhere between huge & enormous....
I don't know what you mean by sensor, but most optical equipment, although it uses semiconductors, and uses some lithography (eg to create ridges and waveguides) is not defined by "nodes" in the way that logic and memory are. The difficulties, and the advances, happen along very different dimensions.
For example right now one dimension is trying to expand wavelengths, to get beyond far infrared to THz lasers. (These can be manufactured today, but tend to be based on non-linear processes, so are extremely low power). A second dimension is improving existing tech (like ICLs and QCLs) for higher efficiency, or higher power, or covering a broader band.
The high end devices are created from artificial materials. ie you lay down a layer (one or two or three atoms high) of one element, then a similarly thin layer of a second element, then a third, and repeat this say fifty to a hundred times. Designing this correctly, and then fabricating it successfully, is where the magic is. Once you have the material, the actual laser fabrication (lithography, etching) is pretty vanilla.
This tech is still being done with tiny wafers (often 2 inches). Laugh if you like --- but there are many ways to be complicated, and atomic layer epitaxy is just as tough in its way as creating 7nm finFETs.
What does this give you? Primarily it gives you a laser that can be used for spectroscopy in the mid-IR, which is probably the best single band for distinguishing different molecules. So you get a gadget that can do things like - non-invasive glucose level detection (just shine it at your skin, no needles) for diabetics - or you can breathe into it, and trace chemicals in your breath can diagnose different issues - or you can mount it on a bridge, with an attached camera, and sense the pollution level of cars driving by. No need for annual smog check, and on-going detection of offenders (no passing smog check, then modifying your engine) - or you can look at scattering through the atmosphere and detect pollution (from factories, or leaks from pipelines or storage) - or less invasive explosives detection at airports and suchlike ...
Pretty much all of this has been demo'ed in labs. Pretty much none of it is yet commercially available, but that's on track. You CAN today, buy things like fancy microscopes that allow you to point at a spot in your sample and get told what chemicals are there; or suitcase sized devices that you can take to a factory and it will tell you all the chemicals around it.
The on-going goal is to get this stuff ever smaller, ever cheaper, so that one day it's built into your iPhone. That may seem optimistic, but there was a time when things like the IR VCSELs in your iPhone (part of FaceID) were way too large and expensive to be used for that task...
This particular sensor was magnetic, specifically looking at the magnetic threads in currency (bills). As far as I remember, the US magnetic thread was 'dumb', either there was one, or there wasn't. The thread in the Euro actually had a barcode pattern that was unique per denomination... ie different between 5/10/20/etc Euro.... so for machines setup for Euro, the barcode pattern was part of the authentication process. If the various other sensors (optical incl.) said one thing and the magnetic thread said another, the machine might call it a counterfeit.
here's an analogy: way back when, I worked in a Swiss screw machine (look it up) shop that still ran mostly cam machines when CNC was taking hold. they had an old Bechler that did nothing but cut brass tubing, with OD chamfer, into about 2 inch lengths. two shifts, 6 days. went into GM carburetors. yeah, so capitalists will milk machines until it costs too much to keep them running.
Leakage current gets worse the smaller the node is. http://ww1.microchip.com/downloads/en/appnotes/014... So for low power applications, like sensors which don't need processing power, a small fabrication process would be counterproductive. So small slow but low power microcontrollers.
It's also likely that some military/space designs use ICs based on these large but robust fabrication nodes. In such cases it's also very expensive to switch the node, due to testing and validation.
'stupid' ICs, like motor driver,... also need to get produced by someone.
Interesting observation. The 16/20 dropped at the same time 10 peaked. Gives the appearance that some folks jumped to 10nm... then struggled... and jumped back to 16/20. Maybe their 10nm was really hard and difficult to ramp. Intel is existence proof of that possibility. The 16/20 is dropping again as 7 ramps.
Not at all. Two DIFFERENT things are constantly happening.
The leading edge crowd (Apple, Huawei, QC) used 10nm for a year, then moved to 7nm as soon as they could. That's the huge changes you see as each leading node comes on line.
Meanwhile everyone else, on a much less aggressive schedule, moves from one node to another in a way that's much slower and more gradual. They were likely on 28nm for 5 yrs or so, and when they finally needed better specs for their part, 16nm (ie finFETs) was the easiest next node that did the job.
Again the timing is different from what you might think. - Manufacturing (especially for Apple) peaks a quarter or two before devices ship. - Payment does not have to exactly match manufacturing or delivery. Either Apple or TSMC may have particular reasons (tax, revenue smoothing, whatever) for trying to shift payment forward or backward by a quarter. - The other large 10nm players (like Mediatek and Huawei) are on their own particular schedules that might lead or trail Apple by a quarter.
All the numbers show is how the PAYMENTS flowed for 10nm vs 7nm. That's different from (not completely detached from, but different from) how the manufacturing flowed.
amazing that 10nm was such a dog of a process that it only grew in share for TWO quarters! and then even when it was still the "top" node, it proceeded to lose share two quarters before its replacement even came out. Wow.
10nm was basically a phone only process, and q4/q1 are probably peak sales time periods for the SoCs with Q2 the nadir. Q3/4 are new iphones (late q3 is production for the release day stock), Q4 Christmas, and Q1 production for all the new Android phone releases and the end of the quarter.
If it was broken out separately, I suspect you'd see something similar for 20nm; IIRC that was the last short lived phones only process.
The ignorance of some really is something, isn't it?
You have UTTERLY NO clue how TSMC does business, do you? Some questions (a) do you think that TSMC lost money on 10nm? Even now, before considering that it will continue to be used for some purpose for years to come? (b) do you think that TSMC could deliver such reliable execution, year after year (compared to, eg, Intel) if they didn't take these intermediate steps? (c) do you think Apple and Huawei were unhappy with the chips they delivered on 10nm?
You're like some Dilbert middle manager who arrives at a building site and says that "obviously scaffolding is a waste of money, because it's gone by the time the building is finished! Next time we'll save so much money and time by not bothering to install it at all"...
Totally agree on (b) and (c). However (a) is arguable. Their gross margin on 10nm probably never reached corporate average, since it takes a year or so. And not likely to see new tape-outs on 10nm when you have much better 7nm available.
actually I do have a clue - they amortize their extreme capital expenses across when something is an expensive leading node, then make boatloads of cash on the fully depreciated process lines 2+ jumps behind leading edge.
know what that strategy depends on? keeping the line running as long as possible, leading edge or not.
Do YOU think TSMC was happy with their 10nm process being functionally dead a year after introduction? nothing else will tape out on 10nm, and why would anyone use it now?
They'll either go 16nm for cost, or 7nm and live with SADP if they really need it. no arguing whether it was necessary or not (obviously it was) but that it was necessary and that they used it doesn't mean it wasn't also a disappointment overall.
Agree. The node was basically made for Apple. Hisilicon is a bonus. So they probably expected much longer volume commitments from the customer. On the bright side, they could reuse all the equipment for 7nm. So at the end of the day they are even.
And you think TSMC wasn't aware of this dynamic? Especially since they went through the exact same thing with 20nm?
What I see when I read these discussions of "short nodes" is people who desperately want to find SOMETHING to criticize in TSMC's execution, and who latch onto an idiotic point for idiotic reasons.
Are you ever reading/joining their quarterly calls? They openly communicate that not all their expectations always maturalize. It was during the last call, and the same was a year ago. Short nodes are good for the progress, but they are not making much money on those.
Like I said, some nodes are the scaffolding that makes the other nodes possible... Or, if you want a different analogy, some nodes are the loss-leader movie, but they make it up in merch and foreign distribution.
You have to look at the ENTIRE picture. No-one wants to lose money on a node (and I don't believe they ever do). But not all nodes are there to be cash cows; and as long as they do their job of moving the company onward to the next cash cow, everyone's happy --- including the INTELLIGENT analysts and investors. You can always find some idiot complaining about the financial performance of one node; the question is --- why is that idiot worth listening to?
I keep expecting to see an announcement that they are going to break ground on a new fab or two and open up their process to third parties in an effort to cover R&D costs and generate additional revenue.
it's worth noting that 4/17 10+7 is not much smaller than they became in 4/18. also, isn't 7nm generally considered just 10nm over most of the real estate on a SoC?
A big one is analog. A lot of that stuff is very high- profit and there's little need to shrink it to newer nodes. And analog as a market is booming right now, especially in automotive https://www.mybkexperience.xyz/www-mybkexperience-...
Fine, I was just heading home was there something I could do for you? As she leans down showing me her nice looking titties I simply tell her Doug has been keeping me informed of your training and I was hoping to see if I could give you a quick test to see how you are progressing. Reaching into her dress and giving her right tit a good squeeze and all she does was smile with a verbal yes. Good let’s go back to your classroom and give you this pop quiz. Once back in her class I head for her chair and take a seat https://bigboobsgifs.com/ . As she approaches I just lift my hand telling her to just lift her dress and to lay back on her desk. Seeing her matching panties with her dress just makes my pussy wet for some reason. Without a word she lays back pulls her panties to the side and spreads her legs wide with her feet pointing to opposite sides of the room. Doug has trained her well, leaning in I know I don’t have much time so I start with a trick Frank does on me and just seal place my lips around Mrs. Rodrigues bald pussy and shove my tongue straight into heaven. Sucking and licking her clit, her pussy folds inside and out rimming her ass and finger fucking too. I was in heaven between these legs, it took nearly twenty minutes for her to have that amazing orgasm that Doug so often spoke about and the taste of her pussy juices running down my throat, chin and smeared all over my fingers as she shoots her load of women cum. I stand up, as she sits up asking if she pass. With flying colors as I lean in and give a very passionate kiss, tongue and all. As I head out to meet the boys I turn and tell her we need a girls night soon https://boobsgifs.com . She smiles, nodding yes as I head back to the car. I arrived just as the boys were coming out of the school, perfect timing on my part, smiling like I just ate a piece of heaven.
As we head home it did not take long for Frank to smell Mrs. Rodrigues on me. Asking mom who’s pussy are we smelling. As if this was a normal conversation I reply. It is Mrs Rodrigues’s pussy you smell, I saw her coming out of school wearing that dress thinking how hungry I was for some of her. Your dad does not stop talking about how good her pussy is, so I had time and wanted her to know as well what she and Doug were doing and that she will be servicing me from time to time as well. You know mom Mrs. Rodrigues is are last class of the day and I would just love to watch her eat your pussy and for you to eat her as well of course. Thinking it is Wednesday tomorrow and nothing was planned I quickly agree. We will have your father text not to wear panties tomorrow or bra William adds, I smile and agree to no bra tomorrow for Mrs. Rodrigues. As I drive the boys home I ask if they need to be dropped off anywhere, they just shake their heads no.
With a quick dinner the booths head upstairs to do their homework and get ready for school tomorrow. I head into the living room were Doug is watching a action movie, which brought up memories of that night just a couple of months ago https://bigtitsgifs.com/ . The boys finger fucking my pussy and ass, sucking my tits I was in sexual heaven that night and almost every night since than, shaking my head thinking how good my life is right now. I told Doug what the boys and I are going to do tomorrow with his latest pet and to text her letting her know what is expected of her and to be naked under her dress. I also told Doug that I will be dress the same tomorrow all day as well. With that said all he could do was ask for a blow and fuck before bed. I was so charged up if he did not ask I was going to take it from him anyways. I do love my husband.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
44 Comments
Back to Article
SydneyBlue120d - Thursday, January 17, 2019 - link
Astonishing achievement, kudos to TSMC.DanNeely - Thursday, January 17, 2019 - link
Looking at the pie chart breakdown, I'm wondering what's special about the 150/180nm processes that they're still 8% of revenue when the newer and smaller 90/110/130nm ones have much smaller residual shares.Looking at how quickly it's share has imploded since 7nm came out, I'm also wondering if 10nm will be going away entirely in the next year or two.
deepsubmicronwarrior - Thursday, January 17, 2019 - link
"what's special about the 150/180nm processes that they're still 8% of revenue when the newer and smaller 90/110/130nm ones have much smaller residual shares."It's highly likely those geometries are 8 inch wafers vs. others being 12 inch wafers, which means wafer starts for those geometries are low-cost and low-risk. #speculation
drexnx - Thursday, January 17, 2019 - link
that makes a lot of sense, .13u was around when most fabs went to 300mm waferse1jones - Thursday, January 17, 2019 - link
A company I worked for a few years ago had an ancient 4-inch wafer fab that made the sensor elements for a sensor we developed. I never asked what size process it was, but it was probably somewhere between huge & enormous....woggs - Thursday, January 17, 2019 - link
That's puts it at ~10um and maybe 5um at best.name99 - Thursday, January 17, 2019 - link
I don't know what you mean by sensor, but most optical equipment, although it uses semiconductors, and uses some lithography (eg to create ridges and waveguides) is not defined by "nodes" in the way that logic and memory are. The difficulties, and the advances, happen along very different dimensions.For example right now one dimension is trying to expand wavelengths, to get beyond far infrared to THz lasers. (These can be manufactured today, but tend to be based on non-linear processes, so are extremely low power). A second dimension is improving existing tech (like ICLs and QCLs) for higher efficiency, or higher power, or covering a broader band.
The high end devices are created from artificial materials. ie you lay down a layer (one or two or three atoms high) of one element, then a similarly thin layer of a second element, then a third, and repeat this say fifty to a hundred times. Designing this correctly, and then fabricating it successfully, is where the magic is. Once you have the material, the actual laser fabrication (lithography, etching) is pretty vanilla.
This tech is still being done with tiny wafers (often 2 inches). Laugh if you like --- but there are many ways to be complicated, and atomic layer epitaxy is just as tough in its way as creating 7nm finFETs.
What does this give you? Primarily it gives you a laser that can be used for spectroscopy in the mid-IR, which is probably the best single band for distinguishing different molecules. So you get a gadget that can do things like
- non-invasive glucose level detection (just shine it at your skin, no needles) for diabetics
- or you can breathe into it, and trace chemicals in your breath can diagnose different issues
- or you can mount it on a bridge, with an attached camera, and sense the pollution level of cars driving by. No need for annual smog check, and on-going detection of offenders (no passing smog check, then modifying your engine)
- or you can look at scattering through the atmosphere and detect pollution (from factories, or leaks from pipelines or storage)
- or less invasive explosives detection at airports and suchlike
...
Pretty much all of this has been demo'ed in labs. Pretty much none of it is yet commercially available, but that's on track. You CAN today, buy things like fancy microscopes that allow you to point at a spot in your sample and get told what chemicals are there; or suitcase sized devices that you can take to a factory and it will tell you all the chemicals around it.
The on-going goal is to get this stuff ever smaller, ever cheaper, so that one day it's built into your iPhone. That may seem optimistic, but there was a time when things like the IR VCSELs in your iPhone (part of FaceID) were way too large and expensive to be used for that task...
e1jones - Thursday, January 17, 2019 - link
This particular sensor was magnetic, specifically looking at the magnetic threads in currency (bills). As far as I remember, the US magnetic thread was 'dumb', either there was one, or there wasn't. The thread in the Euro actually had a barcode pattern that was unique per denomination... ie different between 5/10/20/etc Euro.... so for machines setup for Euro, the barcode pattern was part of the authentication process. If the various other sensors (optical incl.) said one thing and the magnetic thread said another, the machine might call it a counterfeit.FunBunny2 - Thursday, January 17, 2019 - link
here's an analogy: way back when, I worked in a Swiss screw machine (look it up) shop that still ran mostly cam machines when CNC was taking hold. they had an old Bechler that did nothing but cut brass tubing, with OD chamfer, into about 2 inch lengths. two shifts, 6 days. went into GM carburetors. yeah, so capitalists will milk machines until it costs too much to keep them running.UpSpin - Thursday, January 17, 2019 - link
Leakage current gets worse the smaller the node is.http://ww1.microchip.com/downloads/en/appnotes/014...
So for low power applications, like sensors which don't need processing power, a small fabrication process would be counterproductive. So small slow but low power microcontrollers.
It's also likely that some military/space designs use ICs based on these large but robust fabrication nodes. In such cases it's also very expensive to switch the node, due to testing and validation.
'stupid' ICs, like motor driver,... also need to get produced by someone.
woggs - Thursday, January 17, 2019 - link
Interesting observation. The 16/20 dropped at the same time 10 peaked. Gives the appearance that some folks jumped to 10nm... then struggled... and jumped back to 16/20. Maybe their 10nm was really hard and difficult to ramp. Intel is existence proof of that possibility. The 16/20 is dropping again as 7 ramps.name99 - Thursday, January 17, 2019 - link
Not at all. Two DIFFERENT things are constantly happening.The leading edge crowd (Apple, Huawei, QC) used 10nm for a year, then moved to 7nm as soon as they could. That's the huge changes you see as each leading node comes on line.
Meanwhile everyone else, on a much less aggressive schedule, moves from one node to another in a way that's much slower and more gradual. They were likely on 28nm for 5 yrs or so, and when they finally needed better specs for their part, 16nm (ie finFETs) was the easiest next node that did the job.
woggs - Thursday, January 17, 2019 - link
That makes sense. If the plot went further back in time, might see it the 28 to 16/20 transition, too.woggs - Thursday, January 17, 2019 - link
Still... 10nm peaked and dropped long before 7 was ready. That's weird.Dvgeniy - Thursday, January 17, 2019 - link
Probably this is not the first year when Apple suddenly cuts their orders. Looks like they did the same in q1 2018.name99 - Thursday, January 17, 2019 - link
Again the timing is different from what you might think.- Manufacturing (especially for Apple) peaks a quarter or two before devices ship.
- Payment does not have to exactly match manufacturing or delivery. Either Apple or TSMC may have particular reasons (tax, revenue smoothing, whatever) for trying to shift payment forward or backward by a quarter.
- The other large 10nm players (like Mediatek and Huawei) are on their own particular schedules that might lead or trail Apple by a quarter.
All the numbers show is how the PAYMENTS flowed for 10nm vs 7nm. That's different from (not completely detached from, but different from) how the manufacturing flowed.
Dvgeniy - Friday, January 18, 2019 - link
>Either Apple or TSMC may have particular reasons (tax, revenue smoothing, whatever) for trying to shift payment forward or backward by a quarter.If Apple were prepaying, TSMC would not be complaining now about high inventory levels. Mos likely they pay on delivery.
drexnx - Thursday, January 17, 2019 - link
amazing that 10nm was such a dog of a process that it only grew in share for TWO quarters! and then even when it was still the "top" node, it proceeded to lose share two quarters before its replacement even came out. Wow.DanNeely - Thursday, January 17, 2019 - link
10nm was basically a phone only process, and q4/q1 are probably peak sales time periods for the SoCs with Q2 the nadir. Q3/4 are new iphones (late q3 is production for the release day stock), Q4 Christmas, and Q1 production for all the new Android phone releases and the end of the quarter.If it was broken out separately, I suspect you'd see something similar for 20nm; IIRC that was the last short lived phones only process.
iwod - Thursday, January 17, 2019 - link
Are there any SoC using 10nm other Apple A11? Qualcomm and Samsung SoC are all using Samsung Foundry for 10nm.The 6% revenue suggest iPhone 8 / X are still selling pretty well so far.
name99 - Thursday, January 17, 2019 - link
Huawei Kirin 970 is TSMC 10nm.Mediatek Helio X30
Qualcom Centriq and Snapdragon 700 are on Samsung 10nm
https://en.wikichip.org/wiki/10_nm_lithography_pro...
Dvgeniy - Thursday, January 17, 2019 - link
HiSillicin aka Huawei used TSMC 10nm.ilt24 - Thursday, January 17, 2019 - link
It seems their numbers are now heavily dependent on iPhone sales. In Q4 2017 Apple sold 77M iPhones, followed by 52M in Q1'18 and 41M in Q2'18cpkennit83 - Thursday, January 17, 2019 - link
Samsung's 10nm had more volume i think. It got both the 835 and 845, while tsmc got apple and kirin.eva02langley - Thursday, January 17, 2019 - link
They have more than 50% of the world business. At this point, Samsung is like an AMD vs Intel.Dvgeniy - Thursday, January 17, 2019 - link
Apple is more volume than 835 and 845 combinedname99 - Thursday, January 17, 2019 - link
The ignorance of some really is something, isn't it?You have UTTERLY NO clue how TSMC does business, do you?
Some questions
(a) do you think that TSMC lost money on 10nm? Even now, before considering that it will continue to be used for some purpose for years to come?
(b) do you think that TSMC could deliver such reliable execution, year after year (compared to, eg, Intel) if they didn't take these intermediate steps?
(c) do you think Apple and Huawei were unhappy with the chips they delivered on 10nm?
You're like some Dilbert middle manager who arrives at a building site and says that "obviously scaffolding is a waste of money, because it's gone by the time the building is finished! Next time we'll save so much money and time by not bothering to install it at all"...
Dvgeniy - Thursday, January 17, 2019 - link
Totally agree on (b) and (c). However (a) is arguable. Their gross margin on 10nm probably never reached corporate average, since it takes a year or so. And not likely to see new tape-outs on 10nm when you have much better 7nm available.drexnx - Friday, January 18, 2019 - link
actually I do have a clue - they amortize their extreme capital expenses across when something is an expensive leading node, then make boatloads of cash on the fully depreciated process lines 2+ jumps behind leading edge.know what that strategy depends on? keeping the line running as long as possible, leading edge or not.
Do YOU think TSMC was happy with their 10nm process being functionally dead a year after introduction? nothing else will tape out on 10nm, and why would anyone use it now?
They'll either go 16nm for cost, or 7nm and live with SADP if they really need it. no arguing whether it was necessary or not (obviously it was) but that it was necessary and that they used it doesn't mean it wasn't also a disappointment overall.
Dvgeniy - Friday, January 18, 2019 - link
Agree. The node was basically made for Apple. Hisilicon is a bonus. So they probably expected much longer volume commitments from the customer. On the bright side, they could reuse all the equipment for 7nm. So at the end of the day they are even.name99 - Friday, January 18, 2019 - link
And you think TSMC wasn't aware of this dynamic?Especially since they went through the exact same thing with 20nm?
What I see when I read these discussions of "short nodes" is people who desperately want to find SOMETHING to criticize in TSMC's execution, and who latch onto an idiotic point for idiotic reasons.
Dvgeniy - Saturday, January 19, 2019 - link
Are you ever reading/joining their quarterly calls? They openly communicate that not all their expectations always maturalize. It was during the last call, and the same was a year ago. Short nodes are good for the progress, but they are not making much money on those.Dvgeniy - Saturday, January 19, 2019 - link
*materializename99 - Saturday, January 19, 2019 - link
Like I said, some nodes are the scaffolding that makes the other nodes possible...Or, if you want a different analogy, some nodes are the loss-leader movie, but they make it up in merch and foreign distribution.
You have to look at the ENTIRE picture. No-one wants to lose money on a node (and I don't believe they ever do). But not all nodes are there to be cash cows; and as long as they do their job of moving the company onward to the next cash cow, everyone's happy --- including the INTELLIGENT analysts and investors.
You can always find some idiot complaining about the financial performance of one node; the question is --- why is that idiot worth listening to?
eva02langley - Thursday, January 17, 2019 - link
Intel is really in trouble... there is no way in hell they will keep their fabs.SquarePeg - Thursday, January 17, 2019 - link
I keep expecting to see an announcement that they are going to break ground on a new fab or two and open up their process to third parties in an effort to cover R&D costs and generate additional revenue.FunBunny2 - Thursday, January 17, 2019 - link
it's worth noting that 4/17 10+7 is not much smaller than they became in 4/18. also, isn't 7nm generally considered just 10nm over most of the real estate on a SoC?Dvgeniy - Thursday, January 17, 2019 - link
A12 have 60% more transistor than A11. And all with a smaller die size!tommybobberson - Friday, January 18, 2019 - link
What about Intel, what share of the market does it hold?Kvaern1 - Friday, January 18, 2019 - link
https://www.statista.com/statistics/294804/semicon...Kvaern1 - Friday, January 18, 2019 - link
Obviously the Anandtech article is about a subset of the total semiconducrtor market.Kvaern1 - Friday, January 18, 2019 - link
Pretty amazing.Pre-Apple every TSMC roadmap was a joke and every launch vaporware as I recall it.
Also, it annoys me 16 and 20nm are grouped as the same node, I'd imagine 20nm on it's own would look a lot more like 10nm.
Deniehy654 - Thursday, January 31, 2019 - link
A big one is analog. A lot of that stuff is very high- profit and there's little need to shrink it to newer nodes. And analog as a market is booming right now, especially in automotive https://www.mybkexperience.xyz/www-mybkexperience-...wiviruqe - Wednesday, December 23, 2020 - link
Fine, I was just heading home was there something I could do for you? As she leans down showing me her nice looking titties I simply tell her Doug has been keeping me informed of your training and I was hoping to see if I could give you a quick test to see how you are progressing. Reaching into her dress and giving her right tit a good squeeze and all she does was smile with a verbal yes. Good let’s go back to your classroom and give you this pop quiz. Once back in her class I head for her chair and take a seat https://bigboobsgifs.com/ . As she approaches I just lift my hand telling her to just lift her dress and to lay back on her desk. Seeing her matching panties with her dress just makes my pussy wet for some reason. Without a word she lays back pulls her panties to the side and spreads her legs wide with her feet pointing to opposite sides of the room. Doug has trained her well, leaning in I know I don’t have much time so I start with a trick Frank does on me and just seal place my lips around Mrs. Rodrigues bald pussy and shove my tongue straight into heaven. Sucking and licking her clit, her pussy folds inside and out rimming her ass and finger fucking too. I was in heaven between these legs, it took nearly twenty minutes for her to have that amazing orgasm that Doug so often spoke about and the taste of her pussy juices running down my throat, chin and smeared all over my fingers as she shoots her load of women cum. I stand up, as she sits up asking if she pass. With flying colors as I lean in and give a very passionate kiss, tongue and all. As I head out to meet the boys I turn and tell her we need a girls night soon https://boobsgifs.com . She smiles, nodding yes as I head back to the car. I arrived just as the boys were coming out of the school, perfect timing on my part, smiling like I just ate a piece of heaven.As we head home it did not take long for Frank to smell Mrs. Rodrigues on me. Asking mom who’s pussy are we smelling. As if this was a normal conversation I reply. It is Mrs Rodrigues’s pussy you smell, I saw her coming out of school wearing that dress thinking how hungry I was for some of her. Your dad does not stop talking about how good her pussy is, so I had time and wanted her to know as well what she and Doug were doing and that she will be servicing me from time to time as well. You know mom Mrs. Rodrigues is are last class of the day and I would just love to watch her eat your pussy and for you to eat her as well of course. Thinking it is Wednesday tomorrow and nothing was planned I quickly agree. We will have your father text not to wear panties tomorrow or bra William adds, I smile and agree to no bra tomorrow for Mrs. Rodrigues. As I drive the boys home I ask if they need to be dropped off anywhere, they just shake their heads no.
With a quick dinner the booths head upstairs to do their homework and get ready for school tomorrow. I head into the living room were Doug is watching a action movie, which brought up memories of that night just a couple of months ago https://bigtitsgifs.com/ . The boys finger fucking my pussy and ass, sucking my tits I was in sexual heaven that night and almost every night since than, shaking my head thinking how good my life is right now. I told Doug what the boys and I are going to do tomorrow with his latest pet and to text her letting her know what is expected of her and to be naked under her dress. I also told Doug that I will be dress the same tomorrow all day as well. With that said all he could do was ask for a blow and fuck before bed. I was so charged up if he did not ask I was going to take it from him anyways. I do love my husband.