@Pete-S said in laptop slow after installing Spectre/Meltdown mitigations:

Wasn't the bios also involved in these mitigations?

In so much as a micro-code patches for the processors are concerned - yes. My processor is updated to the most current, Oct 2019. Doesn't matter - performance still sucks!

I applied these reg changes, rebooted - then applied a script I've done many times before. It was amazing how slow the script ran compared to other machines. I took a video, I'll see about posting it later.

Machine was replaced. Looking into how to test the old one further. Might start that tonight.

Couple thoughts...

It's spelled EPYC...

There were some weird NUMA considerations when trying to saturate 100Gbps Ethernet NIC's. I saw some firmware improvement/workarounds (basically stick TCP to a core so it wouldn't span NUMA nodes). 99% of people are not going to care about this, and the work around seems to be holding from what I've seen.

For Single socket it's really the best bang for buck out there. I've seen some impressive vSAN numbers on it if anyone's wondering. It doesn't lock out 1/2 the PCI-E or Memory lanes like Intel in a single socket config.

I've got a RYZEN at home, and quite happy with it.

I'm hearing Intel made some optimizations for multi-threaded and cloud native workloads in Skylake that cause issues for .NET and single threaded apps. I haven't done any bake offs but I'm curious if AMD went down the same pass (especially with spin locks). If you have older .NET code with lots of spin locks you want to get it updated or get ready for potential (heavy) regressions on performance vs. Broadwell. We are entering a weird situation where CPU vendors are starting to NOT optimize for enterprise applications but for the mega clouds.

While I think ARM is awesome, there also is a LOOOOOOOT of optimization to be done with compilers before it gets taken seriously outside of edge cases, or homogenous workloads where people will invest the time like CloudFlare.

It requires a outage to migrate between Intel and AMD at a virtualization layer. Intel also has QuickAssit and some other interesting offload, as well as Apache Pass technology.

@emad-r said in Ryzen 7 1700 + GeForce 8400 GS = no display:

@harry-lui

It does but only for specific CPU models, I think only 2. The rest is all CPU only, meaning you need discrete GPU to power up the system which is stupid in 2018.

That's what Ryzen processors are. CPU only, no GPU. Know your hardware.

@scottalanmiller said in Seeking an AMD AM4 MicroATX Mobo That Supports Dual HDMI Output:

Cybertron

I don't even know who this is (yes I googled them and found them, still doesn't mean I know who they are).

Unless you have some stupid weird use case, you order a basic system from Dell or HP that fits your needs and you walk away done.

Whitebox is absolutely the wrong way to go.

@kelly said in Eight new Spectre Variant Vulnerabilities for Intel Discovered - four of them critical:

@tim_g said in Eight new Spectre Variant Vulnerabilities for Intel Discovered - four of them critical:

Wow....

But how practical is it to not only first exploit the Spectre vulnerability, but then to get any useful data from most likely other unknown shared VMs on the same box? (really, only on a shared hosting provider host is where 99.9% of the threat is)

This stuff is discarded speculative cached data... maybe a thumbnail you won't be viewing (if it goes that big), or maybe a few bits leading in that direction... maybe credentials (that are encrypted anyways)...

It seems like all you can do is "fish" for unknown discarded speculative data... it doesn't really sound like a huge practical threat, however, I do see the severity and horrible potential of it, just not the practicality.

Here is how I see it playing out in the larger world. Bad actors will be spinning up VMs on hosting providers' hardware, and then trolling for data of cohosted VMs. It isn't a large problem in a secure environment where the list of people who can spin up VMs also have the credentials necessary to make a Spectre-ng attack a waste of time and energy.

Also worth noting, in a cloud environment the data that can be caught this way is essentially random and ephemeral. What works today won't work tomorrow, and whose data you are getting is normally unknown. The scale and anonymity of cloud computing makes these attacks more possible, but less effective, almost to the point of useless.

Here is the CTSLabs "Proof of Concept" video: https://www.youtube.com/watch?v=RrhVhFHTe9o. I think they're trying to demonstrate that they can flash the UEFI with a BIOS whose hash doesn't match a valid one. I'm not entirely sure. They lost a ton of credibility when they stated that a typical way for an attacker to hack a server is using the BUILTIN\Administrator account and then copy over the BIOS file to the C$.

@dustinb3403 said in AMD Sixteen Core Ryzen Threadripper CPU for $1K:

That would be one expensive windows license bill right there. lol

For server, yeah. But this is for desktops, where MS is not looking at cores.

Okay - I swear it gave me an error when I typed

$ sensors

before.

asus-isa-0000 Adapter: ISA adapter cpu_fan: 0 RPM radeon-pci-0008 Adapter: PCI adapter temp1: -1.0°C (crit = +120.0°C, hyst = +90.0°C) k10temp-pci-00c3 Adapter: PCI adapter temp1: +6.0°C (high = +70.0°C) (crit = +70.0°C, hyst = +69.0°C)

@Harry-Lui said in AMD Radeon RamDisk Utility for Windows:

@dafyre

Windows 98 boot disk creates a RAM disk to store all its files which also happens to be my first experience with RAM disk. You can also copy the need files off the boot disk to create your own RAM disk in windows.

I've been using RAM disks since 1987 on the Commodore Amiga where we would use it for all kinds of things. Our memory was many times the size of our storage making it a no-brainer. Especially as the storage was insanely slow floppies.

My answer to the OP. I mean I am not saying MS is good or bad, or Linux OS's are good or bad. But the OP was just a whiny bitch.

0_1489771473077_upload-b9cb86b3-5bb0-409a-bbe8-3b4008d3a94f

@NDC said in Did AMD Just Stage a Comeback to the Server Market?:

Is there actual info available for server parts yet? All that I've seen so far is data on the Ryzen 7 line which they seem to be positioning as high end desktop components. Not that it will stop people from running servers with them or anything.

Not yet. That is supposed to come later this year.

@scottalanmiller said in AMD, trying to mount a comeback.:

@hobbit666 said in AMD, trying to mount a comeback.:

wow 32cores that's going to be expensive for licensing when everyone starts doing it based on cores not cpu's 🙂 Like Microsoft SQL and Datacentre

Only a few places do that, most are not based around cores. Most enterprise OSes don't need licensing at all. Remember that even 32 cores is tiny for big iron where they are way past these numbers long ago.

Oh, I'm reliving memories from 1997-2002 here, and the licensing associated with IDEAS, CATIA, and Pro-E. Most of those didn't have per-core licensing, but I remember one did. Just thinking about all the licensing servers in use at that place can cause me to have convulsions.

@brianlittlejohn said in Sony PS4 Neo Announced:

@scottalanmiller RIght now Xbox 360 and PS3, but they are all talking about getting the new ones.

I got a PS2 and PS3 too, just when the PS3 and PS4 came out 😉 I get them a generation behind when they are effectively free.