Shots have been fired at AMD’s extremely successful EPYC line up!

Talk about gas lighting:

Intel Sapphire Rapids Xeon CPUs Show Surprising Results Versus AMD EPYC Genoa In AVX-512

So, who’s doing the shooting? Jason R. Wilson a writer for wccftech.com has that very dubious honor of severely distorting the test results with his misleading headline.

Sure, the new Saphphire Rapids do show considerable improvements on AVX-512 workloads but that really only goes to show that Intel has not given up the ghost in terms of applicability for a slew of applications such as, AI, HPC and ML.

There are however definite positives to several points in the article those are particularly apparent where the testing suite is brought to the forefront. Michael Larabel editor in chief of phoronix.com has put together a suite that really ups the game.

We are particularly impressed by how using Cpuminer-Opt in benchmarking on a range of algorithms clearly shows the massive gains in CPU productivity across the board. This is solid information that any sys admin who has been charged with keeping machines hot and ready for on demand HPC can use to optimize their productivity.

Unfortunately we again come back to wccftech.com and their somewhat dubious writer who seems to claim that the original data also contains data on cpuminer-gr testing. That is the algorithm used by Raptoreum one of the more interesting cryptocurrencies in the space at the moment. “Yes, Jason data sources matter, they matter a whole damn lot!”

So we went ahead and rented a pair of Sapphire Rapids from AWS and mined some RTM. The results were frankly underwhelming, since even with using the AVX-512 instruction set, it only reached a fraction of the performance of Genoa EPYCs due to the lack of L3 cache. The only data we can use from this test is very simply that Sapphires did even more abysmally on GR hashing for RTM with the mining software being free to choose the instruction set.

That leaves us to ponder the question of why a reputable tech site like wccftech.com would use such a second rate writer who feels the need to embellish the facts as well as distort them into something of very dubious value?

Is that writer trying to rekindle the fire in the Raptoreum community and get them hyped again? Similarly to what happened at the end of 2021 with them getting enough tech news coverage that they at one point were being blamed for the collapse of Nvidia stock?

Or is Jason just another sad Intel stock holder who feels that the only way for him to recover from their underwhelming performance is to help gaslight the sysadmin community?

We have a lot of respect for a project that sets out to build up an ecosystem without any start up funding, not even and ICO or premine that actually gets stuff done and keeps at it! Combine that with honest no bullshit management and you have a clearly winning long term bet.

Jason, upsetting a rather rowdy community like the RTM one by trying to sneak repumpish gaslighting into otherwise unrelated writing seems ill-advised!

If you are happy with our work, we like:

RTM: RCbyq5ey3wVcshCe9wsgbJuVLdWEbMeuwi
XMR: 8BmXBq2KSs6is4L8W2TSHBFGPBLWGwnJ44J5p1PXQ1FPTEVh4PKdtoi5phNhA2mnxNKBqEGcgMtbf5Hk7giU5xxK6gr1MyA
Firo: aPwWaw5aacY2kCFMfVccqtsXiEHrxgpHpT