Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Intel publishes misleading benchmarks against AMD

servethehome.com

Post reply on HN
Print view View original

171–180 of 209 posts

Re: Intel publishes misleading benchmarks against AMD

#171

Post by lifty »

darkcoderrises wrote: post #158
outworlder wrote: post #99

Earlier quoted context omitted.

According to this: https://www.techrepublic.com/article/spectre-and-meltdown-ex... > Presently, 13 Spectre variants and 14 Meltdown variants have been identified. Initially, AMD processors were thought to be immune to Meltdown, though one variant has been successfully demonstrated on AMD systems. ONE variant. So technically true. But the situation on the Intel side is far worse. Intel got way too greedy with their op…

Yeah, but isn't being greedy with optimizations good?

Not when it comes at the expense of security. Perhaps there are contexts where security is not important and this rule does not apply, but it clearly is a problem for CPUs.

Re: Intel publishes misleading benchmarks against AMD

#172

Post by bjornjaja »

NicoJuicy wrote: post #39

They wouldn't have an answer to AMD untill 2021: https://newsroom.intel.com/news/2019-intel-investor-meeting/... Ps. Long AMD, perfect execution the last two years. It's crazy, even their late firmware fixes improve performance as it had bad press for a moment for false benchmarks advertisement ( it was not false so it seems, the firmware was just not ok yet. Not everything is fully ok now, it's the only minor nitpic…

AMD bringing the competition forces Intel to up their game and bolster R&D—they may come with some beast chips as a result.

Such is the way things work, scales tip one way then the next.

What about NVIDIA??

Also isn’t Amazon working on their own chips?

Re: Intel publishes misleading benchmarks against AMD

#173

Post by roblabla »

icedchai wrote: post #161
computerex wrote: post #52

Earlier quoted context omitted.

You are not supposed to disable the mitigations... Especially on servers and workstations... And keep in mind that the usual mitigations aren't even holistic. The only truly 100% secure fix for Spectre/Meltdown is to disable hyperthreading altogether.

For servers that aren't running arbitrary third party code, like PaaS or IaaS, why not?

If your server is running baremetal, no problem. If it's running on a shared hosting environment, a co-tenant VM can exfiltrate your data. See L1TF/Foreshadow.

Re: Intel publishes misleading benchmarks against AMD

#174

Post by lolc »

greatpatton wrote: post #144
paulmd wrote: post #45

Earlier quoted context omitted.

Also, the difference in sub-NUMA configuration actually favors AMD as well. NPS=4 is optimal for AMD, as shown by ServeTheHome's own work earlier. Both systems are tested in their proper (fastest) configuration. Despite ServeTheHome's own previous work showing this, they're whining about it being different. But if Intel hadn't tested AMD in the proper configuration they would have complained about that too. Yeah, you…

oh really one thread per core is the optimum for EPYC server? I don't know why AMD care about hyper threading then.

We're not talking about the general case. One thread per core might be optimal for that specific workload.

Re: Intel publishes misleading benchmarks against AMD

#175

Post by mcdevilkiller »

londons_explore wrote: post #62
thaumaturgy wrote: post #59

Earlier quoted context omitted.

They can't fake public PassMark results though. They currently hold four of the top five results for high end processors there ( https://www.cpubenchmark.net/high_end_cpus.html ) and a lot of the top price/performance results ( https://www.cpubenchmark.net/cpu_value_available.html ). A lot of folks use charts like these to guide buying decisions or make recommendations to others.

If I were Intel I'd have whole teams of people trying to invent new benchmarking tools and apps. And all those apps would be compiled with options that favor Intel and include only tests where Intel looks good...

Nah, they prefer having all the reviewers and online press in the pocket, you just need money for that, and if there is one thing Intel has, is money.

Re: Intel publishes misleading benchmarks against AMD

#176

Post by icedchai »

roblabla wrote: post #173
icedchai wrote: post #161

Earlier quoted context omitted.

For servers that aren't running arbitrary third party code, like PaaS or IaaS, why not?

If your server is running baremetal, no problem. If it's running on a shared hosting environment, a co-tenant VM can exfiltrate your data. See L1TF/Foreshadow.

Haven't all the big cloud providers have patched their hypervisors for that?

Re: Intel publishes misleading benchmarks against AMD

#177

Post by kllrnohj »

snoman wrote: post #143
kllrnohj wrote: post #100

Earlier quoted context omitted.

> So, we're talking about desktops not used to browse the internet, and bare metal servers whose network stacks aren't exposed to hostile actors? To be fair the context here is HPC. Nobody is installing Chrome on something like IBM's Summit supercomputer. Nor is it running "untrusted" code. It's configured to go as fast as possible, nothing more. Anyone in cloud hosting or desktop would be insane to not install spect…

Is it though? There's a significant security bug that has a known fix, on your supercomputer. The second anything remotely security-related happens and it gets out that you've opted to hedge your bets that no untrusted code will ever execute on that machine... and you could be in some seriously hot water. I'm not really in the HPC space, but I can't really see any vp/cto/cio reasonably putting that asterisk on a secu…

What do you think is the actual risk? Who is attacking whom, and how? If you're running on HPC you basically rented the hardware. It's a time-share system. The entire thing is yours, all yours, for the hours you managed to procure. So who are you attacking? And what vp/cto/cio are you even talking about? The ones who own the hardware, or the ones who are renting it?

And why do you think a security policy is needed here at all instead of just, you know, laws & contracts? You can't just anonymously rent time on these systems even if you have the money - you need your usage approved, too, and it's a big whole process. If you then go and successfully launch a security attack, why does that require technical safeguards instead of just the FBI knocking down your door and hauling your ass to jail?

Re: Intel publishes misleading benchmarks against AMD

#178

Post by mcdevilkiller »

NicoJuicy wrote: post #65

And if anyone from AMD is reading this. I'm truelly awaiting AMD's NUC in multiple price ranges of 150-700€, if possible :)! I'm currently building a DIY RPi 4 cluster for microservices/kubernetes setup and would like to replace my DIY home streaming setup. http://www.fanlesstech.com/2019/10/exclusive-amds-nuc-is-com...

Yes, please! And if possible, a lower frequency, higher thread count processor version plus a low core count, high frequency processor one.

Re: Intel publishes misleading benchmarks against AMD

#179

Post by mcdevilkiller »

loeg wrote: post #133
eyegor wrote: post #117

Earlier quoted context omitted.

Desktop? Not only will that be huge, but I doubt such a system would work within the typical 1500-2000w power budget of a household electrical run.

Rome is nominally 225W/socket on the very high end of the price range; a 4P config puts you at 900W for the CPUs. Depending on other peripherals you might be at 1200-1500W total but I don't think you necessarily go much past that unless you're loading it up with power hungry graphics cards or something. You'd definitely want an efficient power supply! You're right to express some concern -- practical residential limi…

Wait, in the US you use 15A at 120v? Here in Europe (Spain at least) the typical is 16A at 240v. 1500W isn't that much of a problem. Oven breakers are even 20 or 32A.

Re: Intel publishes misleading benchmarks against AMD

#180

Post by thaumaturgy »

Someone1234 wrote: post #88
thaumaturgy wrote: post #59

Earlier quoted context omitted.

They can't fake public PassMark results though. They currently hold four of the top five results for high end processors there ( https://www.cpubenchmark.net/high_end_cpus.html ) and a lot of the top price/performance results ( https://www.cpubenchmark.net/cpu_value_available.html ). A lot of folks use charts like these to guide buying decisions or make recommendations to others.

Why can't you fake a PassMark? I'm legitimately asking. Seems like an aggregated public data set would be particularly prone to intentional cheating.

You're right, they could, but they don't seem to be, and it would be easily discovered (since anybody can run the same software on their system and compare the results for their specific CPU), and probably make quite a splash in the tech media.
Post reply on HN