Opteron vs. EPYC Benchmarks & Performance-Per-Watt: How AMD Server Performance Evolved Over 10 Years

[edwaleni]

2.5 inch NVMe "drives" are a compromise to some degree. Datacenter service managers who didn't want to lose access to storage from the front and to manufacturers who didn't want to make wholesale changes in their rack designs.

Also an issue is signal loss using the kluge M.2 to U.2 cable adapters. Having 2 physical connectors between the NVMe storage and the signal bus raises interference and can cause signal loss. Seems to be a compromise in design.

We have recommended that they instead use NVMe trays for the front bays of modern servers that hinge down and expose the M.2 slots for insertion and replacement. Someone (we think Intel) doesn't like that as they say it exposes the planar to random static shock from DC floor staff performing maintenance or replacement. They prefer "drive" looking NVMe hardware that is shock and static resistant and provides a better cooling profile. We think it wastes space and provides no better density or cooling than M.2.

All that said, it appears there is room for innovation for NVMe storage formats, especially for hardware that can accommodate several NVMe devices in a single chassis.

[torsionbar28]

Holy crap I didn't know Opterons sucked that hard. No wonder AMD wasn't making any money at all.

Uh, No, you're forgetting that these are 10 year old models from a 12 year old platform. Compare to intel Xeon of similar vintage and you'll find roughly the same performance. AMD was reasonably competitive with intel back then. Kids these days.... SMH.

This article is comparing an ancient DDR2 platform to the latest DDR4 platform. The three generations of DDR3 Opteron that came in between are missing from this comparison.

[torsionbar28]

All that said, it appears there is room for innovation for NVMe storage formats, especially for hardware that can accommodate several NVMe devices in a single chassis.

For sure, there's still no workstation solution. As you said, several NVMe devices in a single chassis is required. Until then, workstations and small servers are in limbo.

[oleyska]

Uh, No, you're forgetting that these are 10 year old models from a 12 year old platform. Compare to intel Xeon of similar vintage and you'll find roughly the same performance. AMD was reasonably competitive with intel back then. Kids these days.... SMH.

This article is comparing an ancient DDR2 platform to the latest DDR4 platform. The three generations of DDR3 Opteron that came in between are missing from this comparison.

I challenge you to test a 2700K vs a 7700K where 2700K has highest memory speeds possible and 7700K the lowest.
Where's the IPC improvements gone.. ohh well, most lies in memory.
I should done it more scientific benchmark with proper notes but I tell ya it's scary how little have happened in raw core performance

[M@yeulC]

No. It's been used on Phoronix... That "EP.C" in NPB is short for embarrassingly parallel, C data set size.

Oh, my bad, then. I apologize. I certainly discovered that term with their article, though. And as a frequent phoronix reader (2011-ish, almost daily), I was surprised to see it only now. Well, maybe it is that I only now pay more attention to this term?
In any case, not a big deal. And thanks for your articles

[nomadewolf]

I've got an idea: Intel shouldn't release anything new in the server space for 10 years. Then in 2027 we can be all shocked & amazed when their 2027 Xeon parts beat up on their 2017 parts and act like it's a big deal.

It's an interesting comparison, that's all.
We all know the real comparison is against Intel.
But we all know that AMD will beat Intel in performance per watt and performance per dollar on the server market. It's just a matter of find out by how much...

[cjcox]

Uh, No, you're forgetting that these are 10 year old models from a 12 year old platform. Compare to intel Xeon of similar vintage and you'll find roughly the same performance. AMD was reasonably competitive with intel back then. Kids these days.... SMH.

This article is comparing an ancient DDR2 platform to the latest DDR4 platform. The three generations of DDR3 Opteron that came in between are missing from this comparison.

Those ancient generation Opterons held up pretty well against their Intel P4 competition (perhaps that puts things into better perspective).

[audi.rs4]

I challenge you to test a 2700K vs a 7700K where 2700K has highest memory speeds possible and 2700K the lowest.
Where's the IPC improvements gone.. ohh well, most lies in memory.
I should done it more scientific benchmark with proper notes but I tell ya it's scary how little have happened in raw core performance

I think you meant "7700K the lowest"?

[Adarion]

Sometimes it is just worth to look back and compare. So you can value what you have today.

[schmidtbag]

Holy crap I didn't know Opterons sucked that hard. No wonder AMD wasn't making any money at all.

You do realize these Opterons were designed to compete against Xeons based on Core2, right? Nehalem meanwhile was released shortly after; the hyper-threading quickly obsoleted the performance gain the Opterons had.