How to Save $6000 on a 28-core Flagship Intel Xeon: Platinum 8280 vs Gold 6258R
by Dr. Ian Cutress on August 7, 2020 8:00 AM ESTTest Bed and Benchmarks
For this test, we’ve run through our updated suite of benchmarks, as part of our #CPUOverload project. As this isn’t a strict review of the processors, more of a comparison article to see if they perform the same, then each benchmark is relatively binary– yes it performs the same, or no they don’t (and which one is better). For these tests, we fired up our single socket LGA3647 testbed.
| AnandTech LGA3647 Test Bed | |||||
| AnandTech | COLUMN | ||||
| CPU | Intel Xeon Platinum 8280 Intel Xeon Gold 6258R |
28C/56T 28C/56T |
2.7-4.0G 2.7-4.0G |
205W 205W |
$10009 $3950 |
| Cooling | Asetek 690LX-PN (500W) | ||||
| Motherboard | ASUS ROG Dominus Extreme (0601) | ||||
| DRAM | SKHynix 6 x 32 GB DDR4-2933 | ||||
| SSD | Crucial MX500 1TB | ||||
| PSU | EVGA 1600T2 | ||||
| GPU | Sapphire RX460 | ||||
| Chassis | Anidees Crystal XL | ||||
Both processors were tested on 192 GB of SK-Hynix DDR4-2933 RDIMMs, and a sufficient 500W liquid cooling configuration.
For non-performance benchmark related data, we saw both CPUs score the same average core-to-core latency (8280 was 45.8 ns, 6258R was 45.6 ns), both CPUs get to turbo from idle to max in 35-38 milliseconds, and power consumption was almost identical.
There is a slight variation here, though this could just be down to the specific voltage characteristics of the chips I have. The 6258R hits nearer the 205 W TDP that both chips have.
For the performance benchmarks, don’t get too excited all at once. We’ll mark any performance difference as significant where a >4% change is observed.
| Intel Xeon Scalable 2nd Gen Shootout | |||
| AnandTech | Platinum 8280 |
Gold 6258R |
Performance 6258R vs 8280 |
| Office | |||
| Agisoft 1.3 | 1867 sec | 1797 sec | 103.8% |
| AppTimer GIMP | 54.1 sec | 55.0 sec | 98.4% |
| Science | |||
| 3DPMavx | 54280 pts | 56177 pts | 103.5% |
| yCruncher 2.5b | 47.00 sec | 46.20 sec | 101.7% |
| NAMD ApoA1 | 4.42 ns/day | 4.56 ns/day | 103.2% |
| AIBench 0.1.2 | 523 pts | 521 pts | 99.6% |
| Simulation | |||
| DigiCortex 1.35 | 2.47x | 2.48x | 100.4% |
| DwarfFortress S | 124 sec | 124 sec | = |
| Dolphin 5.0 | 329 sec | 329 sec | = |
| Rendering | |||
| Blender 2.83 | 224 sec | 224 sec | = |
| Corona 1.3 | 13.30 Mray/sec | 13.64 Mray/sec | 102.6% |
| POV-Ray 3.7.1 | 10370 pts | 10461 pts | 100.8% |
| V-Ray | 36899 Kray/sec | 38366 Kray/sec | 103.98% |
| CB R20 ST | 391 pts | 393 pts | 100.5% |
| CB R20 MT | 11539 pts | 11851 pts | 102.7% |
| Encoding | |||
| Handbrake 1.3.2 4K | 74 fps | 74 fps | = |
| 7zip Combined | 183k MIPS | 189k MIPS | 103.2% |
| AES Encode | 15.9 GB/s | 16.4 GB/s | 103.1% |
| WinRAR 5.90 | 30.52 sec | 30.17 sec | 101.2% |
| Legacy / Web | |||
| CB10 ST | 8183 pts | 8185 pts | 100.02% |
| CB10 MT | 66851 pts | 66198 pts | 99.0% |
| Kraken | 929 ms | 929 ms | = |
| Speedometer | 90 rpm | 90 rpm | = |
| Synthetic | |||
| GB4 ST Overall | 4739 pts | 4737 pts | 99.95% |
| GB4 MT Overall | 65039 pts | 66274 pts | 101.9% |
| DRAM Read | 124 GB/s | 126 GB/s | 101.6% |
| DRAM Write | 102 GB/s | 102 GB/s | = |
| DRAM Copy | 115 GB/s | 116 GB/s | 100.9% |
| sha256 8k ST | 486 MB/s | 487 MB/s | 100.2% |
| sha256 8k MT | 12452 MB/s | 12833 MB/s | 103.1% |
| LinX 0.9.5 | 1484 GFLOPs | 1528 GFLOPs | 103.0% |
| SPEC (Geomean of tests, Estimated)* | |||
| SPEC2006 ST | 45.8 | 45.8 | = |
| SPEC2017 ST | 6.0 | 6.0 | = |
| SPEC2017 MT | 109.4 | 111.1 | 101.6% |
| *SPEC results not submitted to SPEC.org have to be labelled as 'Estimated' as per SPEC press licensing rules. | |||
Well, that was a whole lotta nothing.
If we retain that a 4% difference might be more than just statistical noise, then none of these benchmarks come close. A slightly blurry eye with these results might concede that the 6258R actually has the upper hand, which might go in line with the slight variation in power consumption we saw in the power test. But by and large, these chips are essentially identical in performance.
Breakdowns of most of the benchmarks and sub-tests can be found by looking at our benchmark comparison database, Bench. To get the best experience when comparing products on Bench. I find it best to increase the browser zoom and reduce the browser window width, so it looks like this:
Click on the image to go to the section in Bench that compares these two CPUs.
Facebook
Twitter
RSS
81 Comments
View All Comments
Santoval - Friday, August 7, 2020 - link
Virtualization is tricky for Ryzen, but it is not tricky for Threadripper and it is *certainly* not tricky for Epyc. AMD need to fix their shit in order to successfully compete in this market because it's an utter shame if workstation and server customers prefer Intel due to its "robustness" alone.Santoval - Friday, August 7, 2020 - link
p.s. Though, come to think about it, there is nothing "robust" about Intel's multiple security flaws. These matter to server buyers because if things go sideways due to these flaws they are liable to lawsuits by their customers and they can also suffer damage to the reputation and trustworthiness of their company.Santoval - Friday, August 7, 2020 - link
p.s. By the way, the "strange performance behaviors that don't happen on Intel (no resolution in sight)" bit was way too vague and arbitrary. As stated it means nothing, so could you please clarify?ZoZo - Friday, August 7, 2020 - link
I did some benchmarks under Linux, and some of them (a small minority) showed some drastic performance drops when run in virtual machines under KVM (kernel 5.7), as much as 80% in one case. That did not happen when I tested with a Xeon W-2295 before. I spent hours trying to find what was wrong, tinkering with many variations of settings for the VMs. Turns out switching from KVM to ESXi 7.0 kind of fixed the problem. So most probably not a problem with the silicon, but it still showed me that software support can be shaky in some areas.eek2121 - Friday, August 7, 2020 - link
Honestly, I tend to look past a lot of the defenses for Intel, but claiming perf/watt doesn't matter? At least I know you aren't in IT. I deal with a very large amount of companies where the difference of a single watt can determine whether dozens of people have jobs or not. Why? because a single server may not matter much, but thousands? they matter a lot .ZoZo - Saturday, August 8, 2020 - link
I'm not trying to defend Intel, what is it with all you binary-thinking people. I just bought AMD for myself and after some struggle am finally happy with it. I'm only telling what I suspect could be a reason to still pick Intel today, after what I've experienced with an AMD platform.Where did I say that performance/watt didn't matter? Please, point it out.
I said it's not just about that, meaning that it's not the only thing that might need to be considered.
It's getting annoying to have to explain to compensate for poor reading skills.
WaltC - Friday, August 7, 2020 - link
Yes, and how do we know it hasn't changed already?...;) I think it has. Hallelujah, it's about time! This article reminds me of the Intel monopoly Halcyon days when Intel had no high-end x86 competition--the years after the Operon/A64 peak but prior to the AMD Zen debut with Zen 2 a year ago. Let's hope that Intel would iron out most problems in its architectures after essentially rehashing them for many years...;) Milking the cow until the cow ran dry...! Time for something new, Intel..chop, chop...!yankeeDDL - Friday, August 7, 2020 - link
6000 premium for more socket support, seems precisely one of those tricks played by a market monopolist that I am glad to see ending.Spunjji - Friday, August 7, 2020 - link
Indeed. One wonders what the point of setting the price that much higher really is if, in fact, none of their customers ever actually pay that price. Perhaps it's merely to make giving "discounts" that much easier, and/or to provide an opportunity to milk the wealthiest customers.MrVibrato - Friday, August 7, 2020 - link
You are probably right. It is very likely a sales tactic. Those CPUs are not sold in retail (at least not in quantities), but as part of enterprise/complex sales, where contracts (including prices) are negotiated. It is good negotation tactic for a seller to start with a high(er) price. Everyone who ever visited a bazar and haggled with a merchant knows how it goes.And despite having no evidence or experience of Intels approach to sales and negotation, i doubt that anyone who buys more than a handful of those puppies would pay anything close to the listed price...