TSMC Completes 5 nm Design Infrastructure, Paving the Way for Silicon Advancement

[Raevenlord]

TSMC announced they've completed the infrastructure design for the 5 nm process, which is the next step in silicon evolution when it comes to density and performance. TSMC's 5 nm process will leverage the company's second implementation of EUV (Extreme Ultra Violet) technology (after it's integrated in their 7 nm process first), allowing for improved yields and performance benefits.

According to TSMC, the 5 nm process will enable up to 1.8x the logic density of their 7 nm process, a 15% clock speed gain due to process improvements alone on an example Arm Cortex-A72 core, as well as SRAM and analog circuit area reduction, which means higher number of chips per wafer. The process is being geared for mobile, internet, and high performance computing applications. TSMC also provides online tools for silicon design flow scenarios that are optimized for their 5 nm process. Risk production is already ongoing.



View at TechPowerUp Main Site

 

[cryohellinc]

Intel be like

Spoiler

 

[Space Lynx]

Moving faster than I expected... Intel fabs are like dinosaur age now, they can't even get 10nm stable, and struggle with 14nm still...

Ryzen 4xxx series will be 5nm node for sure, holy crap this is great news.

 

[moproblems99]

Moving faster than I expected... Intel fabs are like dinosaur age now, they can't even get 10nm stable, and struggle with 14nm still...

Ryzen 4xxx series will be 5nm node for sure, holy crap this is great news.

Negative. They don't even have 7nm EUV yet. Ryzen 4000 will be lucky to be on that. I wouldn't bet on 5nm until at least Ryzen 7000.

 

[bug]

Moving faster than I expected... Intel fabs are like dinosaur age now, they can't even get 10nm stable, and struggle with 14nm still...

Ryzen 4xxx series will be 5nm node for sure, holy crap this is great news.

completed the infrastructure design for the 5 nm process

But the interesting part is what I've read over time places the physical limit of a Si transistor at or around 5nm. The next decade will get really interesting.

 

[Steevo]

But the interesting part is what I've read over time places the physical limit of a Si transistor at or around 5nm. The next decade will get really interesting.

Quantum tunneling FTW!!

 

[Deleted member 67555]

This is news...I didn't expect yet.

 

[bug]

Quantum tunneling FTW!!

Yeah, maybe someone invents the quantum guard to herd electrons back into the transistor and combat leakage. Who knows?

 

[lexluthermiester]

Yeah, maybe someone invents the quantum guard to herd electrons back into the transistor and combat leakage. Who knows?

Quantum Guard? Do you realize how funny that sounds?

 

[natr0n]

Knowledge/Technology is increasing absurdly.

Only means one thing to the believers.

 

[Space Lynx]

But the interesting part is what I've read over time places the physical limit of a Si transistor at or around 5nm. The next decade will get really interesting.

Not really, they will just move to a chiplet design, well AMD already, have you seen pictures of Ryzen 3700? There is room for another chip on the die. When everyone is on 5nm and maxed out, they will just make the die bigger and add more chiplets and scale it.
IPC will probably be dead though.

 

[windwhirl]

Negative. They don't even have 7nm EUV yet. Ryzen 4000 will be lucky to be on that..

I don't know about 5 nm, but I thought Ryzen 3000 (Zen 2) was already on 7 nm...? Or did I misunderstand something?

 

[krykry]

But the interesting part is what I've read over time places the physical limit of a Si transistor at or around 5nm. The next decade will get really interesting.

>5nm
That's a marketing name of the process. Actual sizes proposed for TSMC's 5nmprocess is 44nm for transistor gate pitch and 32nm for the interconnect. This is nowhere the sizes where quantum mechanics start getting in the way.

 

[moproblems99]

I don't know about 5 nm, but I thought Ryzen 3000 (Zen 2) was already on 7 nm...? Or did I misunderstand something?

I believe it will be on 7nm LP or something of the sort.

 

[notb]

Moving faster than I expected... Intel fabs are like dinosaur age now, they can't even get 10nm stable, and struggle with 14nm still...

Intel's 10nm was meant to be competing with TSMC 7nm EUV. Neither of this technologies is used for mass production yet.

Intel is also working on 7nm (to compete with TSMC 5nm) - it might just be that they'll be first to deliver.
Don't worry too much.

Ryzen 4xxx series will be 5nm node for sure, holy crap this is great news.

Hardly possible if 4xxx series was to be launched next year.
Also, don't unify TSMC and AMD. TSMC is just a supplier - they sell to the highest bidder. If Intel decides to become a TSMC client, there will be no supply left for AMD. ;-)
Well... that's unless TSMC decides to buy AMD...

Not really, they will just move to a chiplet design, well AMD already, have you seen pictures of Ryzen 3700? There is room for another chip on the die. When everyone is on 5nm and maxed out, they will just make the die bigger and add more chiplets and scale it.
IPC will probably be dead though.

Before we get maxed out on 5nm, almost all computing activities will be moved to cloud - gaming included (at least for the people that can accept the idea).
And you'll be looking like a dinosaur with your ever-growing chiplet CPU.

 

[R0H1T]

Intel's 10nm was meant to be competing with TSMC 7nm EUV. Neither of this technologies is used for mass production yet.

Intel is also working on 7nm (to compete with TSMC 5nm) - it might just be that they'll be first to deliver.
Don't worry too much.

Hardly possible if 4xxx series was to be launched next year.
Also, don't unify TSMC and AMD. TSMC is just a supplier - they sell to the highest bidder. If Intel decides to become a TSMC client, there will be no supply left for AMD. ;-)
Well... that's unless TSMC decides to buy AMD...

Before we get maxed out on 5nm, almost all computing activities will be moved to cloud - gaming included (at least for the people that can accept the idea).
And you'll be looking like a dinosaur with your ever-growing chiplet CPU.

Intel's fabs were supposed to 5 years ahead of everyone, according to their initial road-maps. Though tbf Intel's 10nm & 7nm are ahead of TSMC & Samsung nodes by a fair margin.

Pretty sure TSMC is gonna beat 7nm Intel to the market.

Never gonna happen, there's this thing called pride & hubris which Intel's full of. They derided ARM/big little, AMD/glue et al & look where they are now - copying their competition!

I'm not sure what's that supposed to mean

 

[MCJeeba]

This is as hardcore as marketing gets.

 

[notb]

Intel's fabs were supposed to 5 years ahead of everyone, according to their initial road-maps. Though tbf Intel's 10nm & 7nm are ahead of TSMC & Samsung nodes by a fair margin.

And for many years, when Intel was the only party doing advanced CPUs, that might have been true. Not because of Intel's greatness. Semiconductor and processor technology is not limited by some corporations' R&D. What you can buy today is basically what our civilization is capable of at given moment.
GPUs and ARM were few years behind on tech, because they didn't need to be on the edge. They kept using well known, cheaper fab node that has been around for longer.

ARM: because no one knew how to use its potential.
GPUs: because they were used for gaming and no one cared. I mean: we had some GPUs that gave us some fps in some games. No benchmark. We really didn't know if that's the limit of this tech.
But then Nvidia started improving performance by 20% yearly. With CPUs we're getting ~5% yearly because of tech limits. This means gaming GPUs were many years behind.

Pretty sure TSMC is gonna beat 7nm Intel to the market.

Maybe they will, maybe they won't. It's not that important. Intel is making their own CPUs, so they aren't competing with TSMC.
Intel did make a 10nm product as a showcase (a tiny CPU for laptops) before TSMC launched 7nm. They have the tech. It just wasn't profitable.

Never gonna happen, there's this thing called pride & hubris which Intel's full of. They derided ARM/big little, AMD/glue et al & look where they are now - copying their competition!

I don't see this "copying". MCM is a very old idea, which both Intel and AMD (among many other companies) utilized over the years.
Now, what Intel does in marketing (calling competition's product "glued") is something totally separate from what they do in engineering. It's better to make marketing mistakes and good products than other way around.
And MCM is a huge compromise - something that should be seen as the last resort. So yes, Intel tries to avoid it at all cost, but at this moment they didn't manage to compete with EPYC without it. When they move to smaller node with good yield, maybe MCM won't be needed anymore.

I'm not sure what's that supposed to mean

I'm not sure what you meant here (you gave few answers but haven't partitioned my post). The cloud part? I meant exactly what I is written there.
Computing will be covered by cloud in 5 years tops. By "covered" I mean: you won't need a high performing PC at all, for any task.
Today you still need to do some things locally - gaming being the obvious example. But I'm sure you've noticed we're getting awfully close.

And of course cloud will always be priced to compete with intermittent hardware use. So if you game for 2-3h a week, cloud should be cheaper. But if you run a computing node 24/7, hardware will remain the cheaper option.

 

[Wavetrex]

Quantum Guard? Do you realize how funny that sounds?

We're well on our way to invent Quantum Slipstream Drive.

I can bet it will use 1nm chips ;-)

 

[cucker tarlson]

Ryzen 4xxx series will be 5nm node for sure, holy crap this is great news.

Not really, they will just move to a chiplet design, well AMD already, have you seen pictures of Ryzen 3700? There is room for another chip on the die. When everyone is on 5nm and maxed out, they will just make the die bigger and add more chiplets and scale it.
IPC will probably be dead though.

 

[lexluthermiester]

Wow the "fanboy" is strong in this thread..

Negative. They don't even have 7nm EUV yet. Ryzen 4000 will be lucky to be on that. I wouldn't bet on 5nm until at least Ryzen 7000.

Let's review...

Risk production is already ongoing.

Yup, that's what the article says. When TSMC makes such an announcement, they are close to being production ready. Maybe a year out. Ryzen will be on 5nm likely by Q3 of 2020. Those are realistic numbers based on past TSMC performance.

Intel fabs are like dinosaur age now

This is silly. Intel may have fallen behind, but "dinosuar" is a grossly and highly dubious statement. IC development has reached a place where the laws of physics start doing funny things on such small scales. Global Foundaries, TSMC, Intel and other IC fabs each have their own unique ways of working the chemistries and the difficulties involved. Intel will find the way through the problems of it's own process's. Let's face facts, Intel is doing more(higher IPC and faster clocks) with 14nm than anyone else is doing on smaller production nodes.

Before we get maxed out on 5nm, almost all computing activities will be moved to cloud - gaming included

We've been through this in other threads.. Never gonna happen! While the cloud has it's uses, it can not and will not replace standard operation methodologies for the common/general user.

 

[cucker tarlson]

I don't know about 5 nm, but I thought Ryzen 3000 (Zen 2) was already on 7 nm...? Or did I misunderstand something?

current 7nm is just the prelude to 7nm EUV that's the real deal.Nvidia is waiting for 7nm EUV (7nm+) cause current 7nm that we see on RVII and Zen 3000 is not really that much of a game changer.
TSMC's 12nm (refined 16nm) that it's supplying for turing works almost as good as 7nm on R7.You can see R7 clocks pretty badly given how vega liquid could do 1700mhz stock on 14nm and R7 needs water cooling capable of almost 500W to reach 2100mhz, that 2080Ti comes close to doing on air on refined 16nm.
process number is not the whole story.

 

[lexluthermiester]

cause current 7nm that we see on RVII and Zen 3000 is not really that much of a game changer.

It's enough of a game changer to make a difference. Positive progress is always positive.

TSMC's 12nm (refined 16nm) that it's supplying for turing works almost as good as 7nm on R7.

That's an assumption on your part as there are no 7nm Turing parts to make an actual comparison to.

 

[R0H1T]

And for many years, when Intel was the only party doing advanced CPUs, that might have been true. Not because of Intel's greatness. Semiconductor and processor technology is not limited by some corporations' R&D. What you can buy today is basically what our civilization is capable of at given moment.
GPUs and ARM were few years behind on tech, because they didn't need to be on the edge. They kept using well known, cheaper fab node that has been around for longer.

ARM: because no one knew how to use its potential.
GPUs: because they were used for gaming and no one cared. I mean: we had some GPUs that gave us some fps in some games. No benchmark. We really didn't know if that's the limit of this tech.
But then Nvidia started improving performance by 20% yearly. With CPUs we're getting ~5% yearly because of tech limits. This means gaming GPUs were many years behind.

Maybe they will, maybe they won't. It's not that important. Intel is making their own CPUs, so they aren't competing with TSMC.
Intel did make a 10nm product as a showcase (a tiny CPU for laptops) before TSMC launched 7nm. They have the tech. It just wasn't profitable.


I don't see this "copying". MCM is a very old idea, which both Intel and AMD (among many other companies) utilized over the years.
Now, what Intel does in marketing (calling competition's product "glued") is something totally separate from what they do in engineering. It's better to make marketing mistakes and good products than other way around.
And MCM is a huge compromise - something that should be seen as the last resort. So yes, Intel tries to avoid it at all cost, but at this moment they didn't manage to compete with EPYC without it. When they move to smaller node with good yield, maybe MCM won't be needed anymore.


I'm not sure what you meant here (you gave few answers but haven't partitioned my post). The cloud part? I meant exactly what I is written there.
Computing will be covered by cloud in 5 years tops. By "covered" I mean: you won't need a high performing PC at all, for any task.
Today you still need to do some things locally - gaming being the obvious example. But I'm sure you've noticed we're getting awfully close.

And of course cloud will always be priced to compete with intermittent hardware use. So if you game for 2-3h a week, cloud should be cheaper. But if you run a computing node 24/7, hardware will remain the cheaper option.

Technically they aren't but they are competing with AMD+TSMC as a whole & previously TSMC+ARM in the mobile space. Guess what they lost the latter with billions of dollars down the drain, even when the competition was using inferior nodes.

It's not just copying, Intel prides itself as a leader in all sorts of stuff but frankly outside x86 & to an extent fabs, they aren't even second best in most other categories! Some of the major innovation in x86 space in the last 2 decades have in fact come from AMD - x64, IMC, APU (concept?) besides HBM, chiplets. The last 2 aren't limited to CPU but just to count a few things which tiny AMD has scored over Intel. In the meantime chipzilla spent billions on Itanic, P4, Larabee, Atom (in mobile space) just to name a few. The amount of time & money spent in promoting most of these products is staggering, in some cases they bribed, threatened even contra revenued the competition. Intel is a lot of things but the one thing they're not is a (tech) leader that you'd look upto.

Cloud will never replace local computing, not unless Google, MS, Amazon start subsidizing their cloud.

 

[cucker tarlson]

It's enough of a game changer to make a difference. Positive progress is always positive.

That's an assumption on your part as there are no 7nm Turing parts to make an actual comparison to.

cut down vega die (60 cu) at 1750mhz with 300w tdp
full vega die (64 cu) at 1670mhz with 345w tdp
not a game changer.