VOGONS

Common searches


First post, by 386SX

User metadata
Rank l33t
Rank
l33t

Hi,

I'd like to know if there're benchmark results comparing "old" vs modern cpu on a single core and on watts requirements plus on max frequency basis. So to understand how much efficient actual single cores are compared to their freq/cache and power usage beside what the new process size can help reducing power usage and increasing freqs.

Thank

Reply 1 of 2, by agent_x007

User metadata
Rank Oldbie
Rank
Oldbie
386SX wrote:

Hi,
I'd like to know if there're benchmark results comparing "old" vs modern cpu on a single core and on watts requirements plus on max frequency basis. So to understand how much efficient actual single cores are compared to their freq/cache and power usage beside what the new process size can help reducing power usage and increasing freqs.
Thank

Every benchmark that has "Single Threaded" part does it 😀
Max. Frequency is impossible to do (since every CPU clockes different).
Watt/power usage... there isn't really much point.
Faster CPU = less time needed to do the same work = lower power bill (regardless of peak/short power usage).
Efficiency at certain code path (test used "fits" into registers/L1 cache) :
Flops_CPU.png
Source : LINK [PL]

157143230295.png

Reply 2 of 2, by 386SX

User metadata
Rank l33t
Rank
l33t
agent_x007 wrote:
Every benchmark that has "Single Threaded" part does it :) Max. Frequency is impossible to do (since every CPU clockes different […]
Show full quote
386SX wrote:

Hi,
I'd like to know if there're benchmark results comparing "old" vs modern cpu on a single core and on watts requirements plus on max frequency basis. So to understand how much efficient actual single cores are compared to their freq/cache and power usage beside what the new process size can help reducing power usage and increasing freqs.
Thank

Every benchmark that has "Single Threaded" part does it 😀
Max. Frequency is impossible to do (since every CPU clockes different).
Watt/power usage... there isn't really much point.
Faster CPU = less time needed to do the same work = lower power bill (regardless of peak/short power usage).
Efficiency at certain code path (test used "fits" into registers/L1 cache) :
Flops_CPU.png
Source : LINK [PL]

Thanks. 😀