News for hackers. Stuff that matters.

A comment on AMD Zen 3/Ryzen 5000 announcement [video]. Parent comment

Re:AMD Zen 3/Ryzen 5000 announcement [video] by Jap2-0 on Thursday October 08, 2020 @04:59PM (#24723891)
Although everyone is saying that Intel is dead in single thread (which is definitely far more true than previously), I don't see 5.4GHz as infeasible for them - tenth gen goes up to 5.3GHz at the very high end and I would be surprised if they didn't reach at least 5.4GHz.

Looking historically, the 6700k reached 4.2GHz, the 7700k 4.5GHz, the 8700k 4.7GHz, the 9900k 5.0GHz, and the 10900k 5.3GHz - which would imply 5.5 or 5.6GHz for what will (presumably) be the 11900k.

[ Parent | on HN ]
Re:AMD Zen 3/Ryzen 5000 announcement [video] by systemvoltage on Thursday October 08, 2020 @11:55PM (#24726649)
Everyone here is missing the most important concept in semiconductor manufacturing: binning and yield [1]. They're interconnected.

It's not like Intel can make 5.4 GHz chips for days on. It doesn't work that way. Ofcourse they can manufacture processors capable of some ridiculous 6.0 GHz (not accounting for power consumption)... but that would be 0.1% yield. You cannot run a factory like this.

[1] https://www.techspot.com/article/2039-chip-binning/

[ Parent | on HN ]
Re:AMD Zen 3/Ryzen 5000 announcement [video] by effie on Thursday October 08, 2020 @07:27PM (#24725059)
Power consumption will probably curb that frequency hike above 5GHz.
[ Parent | on HN ]
Re:AMD Zen 3/Ryzen 5000 announcement [video] by Jap2-0 on Thursday October 08, 2020 @08:07PM (#24725352)
Haven't people been saying that for a long time?
[ Parent | on HN ]
Re:AMD Zen 3/Ryzen 5000 announcement [video] by read_if_gay_ on Friday October 09, 2020 @04:13AM (#24727878)
We've also been stuck at ~4GHz for a long time. The 10th gen Core only manages more because of their ridiculous 300W power consumption.
[ Parent | on HN ]