TSMC 5nm Node Increases Transistor Density by up to 84-87%

Earlier this month, we reported how TSMC estimated that, all going well, they should begin full-blown 5nm chipset production by April. Now, admittedly, that date is entirely open for speculation now following the Coronavirus outbreak, it seems though that when the chip is released, we’re going to see some pretty huge improvements on the existing 7nm designs.
TSMC 5nm Designs
In a report via TechPowerUp, it is believed that the smaller node design is going to see a huge increase in transistor density when compared to the existing 7nm designs. While TSMC themselves have already suggested a figure in the region of 84%, an independent source (who has presumably gotten hold of one of their 5nm N5P silicon fabrication nodes) has suggested that it might be as high as 87%.
What does this mean for us consumers? Well, in a nutshell, significantly faster processors. Well, at least in theory.

What Do We Think?
In terms of desktop 5nm processors, we don’t expect to see any consumer-level releases until at least 2021. With the Coronavirus pandemic, however, even this may see a delay until 2022. Specifically, as manufacturers look to catch up on production line delays and disruption. The chances are, however, that this will first be debuted within AMD’s Zen 4 architecture. Why? Well, it’s well known that Intel is lagging behind (and struggling to catch up) with their smaller node designs.
So, the short version is, for PC users, this technology is still a fair way in the distance. If accurate, however, it’s certainly looking set to bring some pretty huge improvements.
What do you think? Are you excited about 5nm node design processors? – Let us know in the comments!
intel is dead. TSMC has no capacity for Intel. INtel is so dead as many going forth more more cores
“Significantly faster processors” doesn’t really sum it up. Higher density means that more cache can be crammed in, or more cores, but per core performance for example won’t necessary see a huge jump in performance. It does sound promising for GPUs, which tend to scale better with the number of cores. Also for mobile chips, where size typically matters more, so it’d be possible to cram a significantly better CPU and GPU on the same die.