4 ms·
Would be interesting to compare transistor count without L3 (and perhaps L2) cache. 16-core Zen 5 CPU achieves more than 2 TFLOPS FP64. So number crunching per
by kvemkon 1y ago
Would be interesting to compare transistor count without L3 (and perhaps L2) cache.
16-core Zen 5 CPU achieves more than 2 TFLOPS FP64. So number crunching performance scaled very well.
It is weird, that the best consumer GPU can 4 TFLOPS. Some years ago GPUs were an order of magnitude and more faster than CPUs. Today GPUs are likely to be artificially limited.
- kvemkon 1y agoE.g. AMD Radeon PRO VII with 13.23 billion transistors achieves 6.5 TFLOPS FP64 in 2020 [1]. [1] https://www.techpowerup.com/gpu-specs/radeon-pro-vii.c3575 https://www.techpowerup.com/gpu-specs/radeon-pro-vii.c3575
- zozbot234 1y ago> 16-core Zen 5 CPU achieves more than 2 TFLOPS FP64. So number crunching performance scaled very well. These aren't realistic numbers in most cases because you're almost always limited by memory bandwidth, and even if memory bandwidth is not an issue you'll have to worry about thermals. Theoretical CPU compute ceiling is almost never the real bottleneck. GPU's have a very different architecture with higher memory bandwidth and running their chips a lot slower and cooler (lower clock frequency) so they can reach much higher numbers in practical scenarios.
- kvemkon 1y agoSure, not for BLAS Level 1 and 2 operations. But not even for Level 3?
- layla5alive 1y agoHuh, consumer GPUs are doing Petaflops of floating point. FP64 isn't a useful comparison because FP64 is nerfed on consumer GPUs.
- kvemkon 1y agoEven recent nVidia 5090 has 104.75 TFLOPS FP32. It's useful comparison in terms of achievable performance per transistor count.