The fastest CPU of each vendor we measured, by the geometric mean of its throughput at the ranked size over the four workloads: Intel Xeon 6 (Granite Rapids) for Intel, AMD EPYC Turin for AMD. Both have AVX-512.
The hardware
- Cores
- 2 × 96
- Threads
- 192
- Memory
- 375 GiB
- Peak power (measured)
- 320 W per socket
- Cores
- 2 × 96
- Threads
- 192
- Memory
- 755 GiB
- Power limit
- 600 W per socket
Results
| Workload | AMD EPYC Turin | Intel Xeon 6 (Granite Rapids) |
|---|---|---|
| Swaption portfolio, LMM paths per second at 1,024K paths | 44.0M/s | 41.1M/s |
| Barrier options, Monte Carlo options per second at 16,000 options | 1.04M/s | 991K/s |
| American options, binomial lattice options per second at 512K options | 4.94M/s | 4.35M/s |
| European options, Black-Scholes formula options per second at 128M options | 17.3B/s | 23.2B/s |
Tuned code: tuned for each chip under the same rules. Every size, and every platform, is in the results; each platform name links to its measurement record.
AMD EPYC Turin is 1.07× faster than Intel Xeon 6 (Granite Rapids) on the LMM swaption portfolio; AMD EPYC Turin is 1.05× faster than Intel Xeon 6 (Granite Rapids) on barrier options (Monte Carlo); AMD EPYC Turin is 1.14× faster than Intel Xeon 6 (Granite Rapids) on American options (binomial lattice); Intel Xeon 6 (Granite Rapids) is 1.3× faster than AMD EPYC Turin on European options (Black-Scholes formula).
Energy per result
The LMM swaption portfolio: AMD EPYC Turin uses 59% less energy per result than Intel Xeon 6 (Granite Rapids); Barrier options (Monte Carlo): AMD EPYC Turin uses 52% less energy per result than Intel Xeon 6 (Granite Rapids); American options (binomial lattice): AMD EPYC Turin uses 51% less energy per result than Intel Xeon 6 (Granite Rapids); European options (Black-Scholes formula): AMD EPYC Turin uses 30% less energy per result than Intel Xeon 6 (Granite Rapids).
How these were measured
Each platform ran code tuned for that chip under the same rules, in double precision: five launches per measurement, the middle three averaged, warm calls in every launch until the 95% confidence interval of their mean was within ±1% or until the call limit. Any test spreading more than 5% is repeated on a second machine of the same type, and the steadier of the two is published. A Black-Scholes call makes 100 passes over the whole batch, each counted as work, so its options per second are options × 100 passes over the time of a call. At every problem size, the option price is validated against the reference implementation's before any time counts. The full method is on How we measure.