News AMD 128-core Zen 5-based EPYC 9755 'Turin' tested: 2X faster than 128-core predecessor

[Admin]

AMD's 128-core EPYC 9755 is two times faster than 128-core EPYC 9754 processor in a compression/decompression benchmark.

AMD 128-core Zen 5-based EPYC 9755 'Turin' tested: 2X faster than 128-core predecessor : Read more

 

[Stomx]

When we will finally see these long awaited Turin processors with 3nm versus older 5nm in Genoa?

By the way when they run 7-Zip how exactly they set up the test just to get impression on the speed numbers of our own computers? And was it Windows or Linux based ?

 

[TCA_ChinChin]

I mean the Zen 5 part has 140 watts higher TDP and the Bergamo part is running AMD's small cores instead of the full Zen4 cores. Still pretty impressive for what its worth.

 

[Stomx]

The small cores when there will be 192 of them on a single chip like it was promised will also probably have TDP 500W

 

[stuff and nonesense]

I’ll wait till there are real benchmarks, speculation based on one unverified “data point” is pretty meaningless.

 

[bit_user]

When we will finally see these long awaited Turin processors with 3nm versus older 5nm in Genoa?

The launch event is rumored to be October 10th.

I think the Zen 5 chiplets (powering CPUs up to 128 cores) will be the same ones made on TSMC N4P that we've seen in the Ryzen 9000 CPUs launched thus far. The N3 dies should be the ones featuring the Zen 5C cores and powering the 192-core CPUs.

By the way when they run 7-Zip how exactly they set up the test just to get impression on the speed numbers of our own computers? And was it Windows or Linux based ?

Good question. The article said "@9550pro did not disclose where his 7zip compression/decompression benchmark numbers come from, so one should take them with a grain of salt, or two.", so we really don't know.

I basically treat such leaks as a positive sign (from the AMD camp's perspective), but put very little weight behind them, unless/until many emerge that all seem to tell a consistent story.