SemiAccurate has found another real-world AMD vs Intel server TCO example, this time it is Rome in the hot seat. Like the Naples vs Cascade Lake example we showed you earlier, this one comes with pretty unimpeachable numbers behind it too.
If you recall from the Rome launch, AMD had some pretty hefty claims about TCO but can you recall a launch not filled with rosy numbers? Getting to things like real world TCO values are hard to determine even for your own hardware and facilities, calculating much more than hardware and possibly electrical costs for others is nigh on impossible.
Luckily we can bring you another set to back up our last Cascade vs Naples numbers which showed a pretty clear 10% overall system TCO for deployed Naples vs Cascade. Bear in mind that the top Cascade is about 10-15% faster than Naples and consumes roughly the same energy per socket so that win is based on more than just CPU performance. What do you think will happen when we compare the newest AMD part with 50-100% more performance per socket than Cascade while using only a little more energy? To be fair this comparison is actually Skylake-SP vs Rome so Intel is at about a 5-10% performance deficit versus their newest offerings so scale up to Cascade as you see fit. Enjoy
Note: The following is analysis for professional level subscribers only.
Disclosures: Charlie Demerjian and Stone Arch Networking Services, Inc. have no consulting relationships, investment relationships, or hold any investment positions with any of the companies mentioned in this report.
Latest posts by Charlie Demerjian (see all)
- Qualcomm buys Nuvia for $1.4 Billion - Jan 13, 2021
- Pat Gelsinger is the best possible choice for CEO of Intel - Jan 13, 2021
- AMD’s CES keynote is a disclosure own goal - Jan 12, 2021
- Intel has a blizzard of offerings at CES 2021 - Jan 11, 2021
- What is Intel doing about process and outsourcing? - Jan 11, 2021