The M1 Chip feels 38,026 years every 5 minutes
1–4 of 4 posts
Re: The M1 Chip feels 38,026 years every 5 minutes
#2Ignoring the personification of the M1 which is not even an AI system, so it doesn't 'feel' anything, I'm trying to understand these numbers. The article claims that a cycle in the M1 is 0.25 nanoseconds. The M1 has a clock rate of 3.2Ghz, which is 3200 Hz. A Hz is 1/s, so a cycle would be 1/3200, which is 0.0003125 seconds, or 312.5 nanoseconds. Based on that, the rest of the numbers in the article are wrong.
Also, not sure what would be so special about making these calculations about the M1 and saying that it 'feels' time, when there are other CPUs with clock rates as high as 5.5Ghz, and I believe the highest clock rate achieved with overclocking is over 8Ghz.
A fun, albeit misguided, fanboy fiction :)
Re: The M1 Chip feels 38,026 years every 5 minutes
#3Ignoring the personification of the M1 which is not even an AI system, so it doesn't 'feel' anything, I'm trying to understand these numbers. The article claims that a cycle in the M1 is 0.25 nanoseconds. The M1 has a clock rate of 3.2Ghz, which is 3200 Hz. A Hz is 1/s, so a cycle would be 1/3200, which is 0.0003125 seconds, or 312.5 nanoseconds. Based on that, the rest of the numbers in the article are wrong. Also,…
All due respect, 3.2GHz is not 3200Hz. That’s kHz.
3.2GHz is 3,200,000,000Hz. But yes, you're right; M1 is 3.2 GHz, not 4.0
Re: The M1 Chip feels 38,026 years every 5 minutes
#4Ignoring the personification of the M1 which is not even an AI system, so it doesn't 'feel' anything, I'm trying to understand these numbers. The article claims that a cycle in the M1 is 0.25 nanoseconds. The M1 has a clock rate of 3.2Ghz, which is 3200 Hz. A Hz is 1/s, so a cycle would be 1/3200, which is 0.0003125 seconds, or 312.5 nanoseconds. Based on that, the rest of the numbers in the article are wrong. Also,…
All due respect, 3.2GHz is not 3200Hz. That’s kHz. 3.2GHz is 3,200,000,000Hz. But yes, you're right; M1 is 3.2 GHz, not 4.0
You're right, my bad, my calculations are off by a factor of 10^6, that means 0.0000000003125 seconds, which is 0.3125 nanoseconds, which is not 0.25 nanoseconds but is on the same order of magnitude.