AMD 32-Core EPYC "Milan" Zen 3 CPU Fights Dual Xeon 28-Core Processors

[trparky]

Do you know how much money intel has it would take years for them to go bankrupt, they do more than just processors. You don't get to be that big a company without diversifying your revenue stream.

That may very well be true but it's the management at Intel that needs to get out of the way of the engineers to let them do their jobs. So far management hasn't done that and until that happens they've got their hands tied.

 

[Makaveli]

That may very well be true but it's the management at Intel that needs to get out of the way of the engineers to let them do their jobs. So far management hasn't done that and until that happens they've got their hands tied.

Yup there management has been terrible for the last few years so that is a major issue for them. However they will not be bankrupt and gone in the next 5-10 years even if they fail on the next few products.

 

[FinneousPJ]

You are better offbaquiring ryo. Monero is flawed as it isnt private as has been thought. If you want what monero is suppose to be, get ryo. Follow at https://ryo-currency.com. chances are you allready own an asic for mining ryo in your possesion. People have named it gpu.


There are lot of other cpus on the market arm, tachyum, etc. We dont need intel.

My point was that Monero mining is extremely efficient on Ryzen. If Ryo is susceptible to ASIC mining then it is quite bad for mining on this chip.

 

[trparky]

However they will not be bankrupt and gone in the next 5-10 years even if they fail on the next few products.

I agree. They're going to be burning money for the next five to ten years until they let the engineers do their jobs.

 

[Vayra86]

pretty sure we are all going arm anyway sooo yeah

Sure, just give it a decade at least.

x86 is like oil companies, it WILL keep itself relevant somehow. Its too big to fail until all avenues to succeed are gone.

 

[Xuper]

Holy Molly!!! I thought 32 cores vs 28 cores. wow damn that hurts so much for Intel. one CPU is equal to 57% of Dual Xeon 28 cores on other hand one Intel CPU is 47.3% slower than Zen3.
Xeon = 117171 / 2 = 58,585.5 points , Zen 3 = 111379 so

 

[silentbogo]

I'm the last person to defend Intel, but why wasn't the same kernel used for testing both platforms? 4.2.2 is pretty old, like over 5 years old.

It's Parrot 4.2.2 with kernel 4.18.0 and some patches, which is still 2 years old. Though, I'm having the same issue with these numbers - this version of Parrot OS isn't even in mainstream anymore, and there were a ton of performance improvements in linux kernel since then (to the point where even Clear Linux or any Zen optimized distros don't really give any advantage these days).
Plus, a glaring issue of single-socket vs dual-socket. I'm sure EPYC still wins, but that comparison is not just skewed, it's off the rails.

Geekbench is a better benchmark, stresses a larger number of workloads, and cinebench is just rendering

Geekbench is the worst benchmark in existence. Their error margin is so huge, you can't even get stable results on the same machine back to back (especially GB4).

 

[ratirt]

Oh. AMD is not stopping. I'm just so amazed how they are pushing forward but.... get it in the market and then we will talk

 

[InVasMani]

Did you roll need before greed on that EPYC loot!!?

 

[ZoneDymo]

Sure, just give it a decade at least.

x86 is like oil companies, it WILL keep itself relevant somehow. Its too big to fail until all avenues to succeed are gone.

I mean sure, but oil has Tesla, and x86 has Apple going against it, thats not nothing

 

[Makaveli]

I mean sure, but oil has Tesla, and x86 has Apple going against it, thats not nothing

My opinion but I don't really see apple vs x86.

People that buy apple products love that wall garden and will stay in it. They were probably never going to buy a x86 product. Intel and Amd are clearly competitors Apple kinda competes with itself.

 

[Vayra86]

I mean sure, but oil has Tesla, and x86 has Apple going against it, thats not nothing

Tesla is an excellent example. It offers electric cars, while at the same time, hydrogen fueled vehicles are shaping up to serve another segment. The consensus is that we can now diversify our propulsion engines, so to speak, but its certainly not decided and set on electric alone. ARM and x86 are much the same. There are still big segments of the market where x86 is just essential, or offers advantages causing users not to switch over.

Place Tesla in the position of Apple and its a perfect fit, too. Tesla made electric simple and accessible, but still covers barely or less thsn 10% of the market, but still guides the other fossil based car builders forward, and makes it too big to ignore.

 

[Aquinus]

My opinion but I don't really see apple vs x86.

People that buy apple products love that wall garden and will stay in it. They were probably never going to buy a x86 product. Intel and Amd are clearly competitors Apple kinda competes with itself.

When you go with Apple, you're buying into the ecosystem and all of the benefits that come with having so much of the product being in-house (even more so with Apple Silicon.) Apple makes good products for the general user if you're willing to pay the price premium. There are people who don't think the price is worth it and there are people that are looking for something that Apple doesn't actively compete with, such as PC gaming. It depends on what's important to you. For me, my MacBook Pro is everything I wanted in a laptop. I just had to sell a kidney to afford it.

To stay on topic though: For me, I hate this comparison because of the huge difference in linux kernel versions. Multi-threaded scores on the Xeon setup with mitigations turned off is going to benefit a lot from all of the scheduling and power management changes in the Linux kernel over the last 5 years. This is why I say that this comparison is dumb.

 

[thesmokingman]

I wonder if this Patrick is Turmania's alt account.

Either way, pretty desperate and doesnt change reality soooo yeah, fail.

They all seem to repeat the same mantra don't they?

Anyways on topic, with the ARM onslaught, we have only AMD to thank for fighting back that scourge!

 

[1d10t]

They all seem to repeat the same mantra don't they?

Anyways on topic, with the ARM onslaught, we have only AMD to thank for fighting back that scourge!

That leave Intel holy endeavor is got squished in middle between AMD performance and highly efficient ARM /s

 

[ScaLibBDP]

Why in the world would someone get their hands on a brand new Zen 3 32-core CPU and the first thing that comes to mind is run Geekbench instead of Cinebench?

The problem is actually in a significant misunderstanding that it Does Not make Any sense to compare directly two systems with Very different numbers of cores!

56 Cores vs. 32 Cores, that is the Dual CPU system vs. Single CPU system!

For many-many years Intel and AMD are using Composite Theoretical Performance ( CTP ) values to evaluate Peak Processing Power of CPUs ( also known as Rpeak in Computer Science and HPC communities ):

CTP values ( Non-Normalized ) for Single CPU configurations:

Intel Xeon Platinum 8280 - 2.7 GHz / 28 cores / AVX-2 / FMA support - CTP = 2.419 TFLOPs ( ~15.6% slower than AMD EPYC 7543 Zen 3 )
AMD EPYC 7543 Zen 3 - 2.8 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.867 TFLOPs

CTP values ( Normalized ) for Single CPU configurations:

Intel Xeon Platinum 8280 - 2.7 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.765 TFLOPs ( ~3.6% slower than AMD EPYC 7543 Zen 3 ) ( 2.765 = 2.419 TFLOPs * 32/28 )
AMD EPYC 7543 Zen 3 - 2.8 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.867 TFLOPs

Note: Intel's CPU number of Cores is multiplied by 32 / 28 ~= 1.1428571 to Normalize number of Cores to 32 and in that case we're comparing 32-Core Intel system vs. 32-Core AMD system

For Dual CPU configurations multiply TFLOPs numbers by 2.

As you can see Intel Xeon CPU is only 3.6% slower when compared to the latest Gen 3 AMD CPU if Normalization of number of Cores is taken into account.

PS: I'm a C/C++ Software Engineer and I'm Not in favour of Intel or AMD in that case. However, "Apples-must-be-compared-to-Apples" and Not to "Lemons"...

 

[owen10578]

The problem is actually in a significant misunderstanding that it Does Not make Any sense to compare directly two systems with Very different numbers of cores!

56 Cores vs. 32 Cores, that is the Dual CPU system vs. Single CPU system!

For many-many years Intel and AMD are using Composite Theoretical Performance ( CTP ) values to evaluate Peak Processing Power of CPUs ( also known as Rpeak in Computer Science and HPC communities ):

CTP values ( Non-Normalized ) for Single CPU configurations:

Intel Xeon Platinum 8280 - 2.7 GHz / 28 cores / AVX-2 / FMA support - CTP = 2.419 TFLOPs ( ~15.6% slower than AMD EPYC 7543 Zen 3 )
AMD EPYC 7543 Zen 3 - 2.8 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.867 TFLOPs

CTP values ( Normalized ) for Single CPU configurations:

Intel Xeon Platinum 8280 - 2.7 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.765 TFLOPs ( ~3.6% slower than AMD EPYC 7543 Zen 3 ) ( 2.765 = 2.419 TFLOPs * 32/28 )
AMD EPYC 7543 Zen 3 - 2.8 GHz / 32 cores / AVX-2 / FMA support - CTP = 2.867 TFLOPs

Note: Intel's CPU number of Cores is multiplied by 32 / 28 ~= 1.1428571 to Normalize number of Cores to 32 and in that case we're comparing 32-Core Intel system vs. 32-Core AMD system

For Dual CPU configurations multiply TFLOPs numbers by 2.

As you can see Intel Xeon CPU is only 3.6% slower when compared to the latest Gen 3 AMD CPU if Normalization of number of Cores is taken into account.

PS: I'm a C/C++ Software Engineer and I'm Not in favour of Intel or AMD in that case. However, "Apples-must-be-compared-to-Apples" and Not to "Lemons"...

Why in the world would you normalize the cores when they are not the same in the real world? That's just single core performance comparison with extra steps.