Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Fame Douglas
Nov 20, 2013

by Fluffdaddy
Well, it's also the first Monday/weekday of the month.

Adbot
ADBOT LOVES YOU

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Well, Kaby Lake is due very soon.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

DrDork posted:

Kaby is also looking like it's no faster on a clock-to-clock basis than Skylake--they're basically just higher clocked Skylakes with some added video decoding hardware thrown in. So for most of the chips you're looking at a 10% at best speed bump, and it's an open question as to how far they'll be able to overclock.

Waiting for Zen is an interesting choice just to see how that shakes things out and if it does much to Intel's pricing scheme. But you're not really gonna get much extra performance from Kaby vs Skylake.

While not a huge upgrade or strictly necessary, the extra video decoding capabilities seem marginally useful (Netflix 4k, YouTube). Kaby Lake is likely out very soon. At the same time, Skylakes are probably going to drop in price. Seems like a good idea to just wait for a bit. If you can hold out for a few months you should probably do it.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

DrDork posted:

The decoding hardware is really only useful if you've got a laptop, though--massively decreasing the power-use needed to run a 4k video is really great when you're working within a limited power profile, but frankly doesn't matter in a desktop unless you're trying to do something really unusual, like play half a dozen 4k videos at once.

While this is true, Netflix specifically will only support 4k playback on Kaby Lake machines. Some dumb DRM thing.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Boiled Water posted:

And presumably all the bandwidth in the world.

Netflix require a minimum of 25 Mb/s, which doesn't seem like all that much these days. Guess this could be a problem for people living in rural areas?

Being required to use Edge seems more heinous to me.

Fame Douglas fucked around with this message at 19:14 on Nov 25, 2016

Fame Douglas
Nov 20, 2013

by Fluffdaddy
AMD should bring back the "All-In-Wonder" brand.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

silence_kit posted:

This is how all computer chips have worked for some time now. The set-up cost for production of computer chips is incredibly high, so to get around this, computer chip companies manufacture only one type of chip, and disable different amounts of its functionality to create the different product categories.

If this bothers you, I hope you are also railing against software companies for daring to charge money for their software when they also let you download the free, limited functionality version from their website. They could be giving you the full featured version at no additional cost.

At least in theory, chips vary in production quality enough for binning to make sense. Selling scratch-off cards to unlock processor functionality just seems scammy, even if it isn't functionally different from regular processor branding.

Fame Douglas fucked around with this message at 17:43 on Dec 18, 2016

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Paul MaudDib posted:

Another fun example though: the Raspberry Pi has various media decoders on its GPU that are locked down until you pay $5 per codec and put a serial-specific license key into a text file on the boot disk. Pretty hefty money considering the thing only costs $35 in the first place.

I assume you are making an argument against software patents?

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Malloc Voidstar posted:

If Intel is cooperating with the NSA/GCHQ like that they wouldn't need this backdoor, they could just sign code.

Well, not if you presume they do this type of cooperation "under the table". But we'll never know the difference, anyways.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Watermelon Daiquiri posted:

idk, i would love it if usb-c replaced almost every port. Need video? usb-c-. Flash drive? usb-c. power? usb-c for up to a certain wattage.

If only finding suitable USB cables weren't such a nightmare: Had some troubles finding a cable capable of driving my 4K display via USB-C DP alternate mode (monitor input and computer output are USB-C).

Fame Douglas
Nov 20, 2013

by Fluffdaddy

cinci zoo sniper posted:

In the grand total, looks like Microsoft fix is out, and nearly nothing might change for consumer CPUs starting with Skylake. There goes the pre-embargo hysteria, I guess, albeit there admittedly are quite a bit of people with pre-Skylake CPUs. Windows fix appears to be out, so time to wait for someone to bench some 20-30 CPUs.

Isn't this what everyone expected? Some workloads are heavily affected, others pretty much not at all.

Fame Douglas fucked around with this message at 12:27 on Jan 4, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy
The Intel CEO filed his planned stock sale on October 30. Wonder if this is going to become an issue for him https://www.sec.gov/Archives/edgar/data/50863/000112760217033679/xslF345X03/form4.xml

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Paul MaudDib posted:

I mean, the truth is that in 6 months nobody will care about this patch. Fixups happen, as long as there is actually a fixup it's fine. 0-5% performance impact is nothing at all as these things go. If there was a fixup for the Ryzen segfault bug with 5% impact I wouldn't care, the problem is there isn't.

0-30%

And anyone operating server infrastructure is going to care.

Paul MaudDib posted:

nah, it's pretty clear that the host was patched, not a particular service.

"CPU usage after our host was patched".

edit: like how would that even work, kpti but for only a couple processes a single process? yeah no.

Services running on multiple servers?

Fame Douglas fucked around with this message at 19:07 on Jan 6, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy

GRINDCORE MEGGIDO posted:

What's the expectations for patched bios releases like, anyone expecting to get screwed by manufacturers not releasing them?

My 2015 Asus laptop didn't get any Management Engine fixes, not holding my breath for any updates.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
https://www.phoronix.com/scan.php?page=article&item=linux-more-x86pti&num=1

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Craptacular! posted:

30% seems like an outlier. I don't have the link, but Red Hat's testing hit 10-12%, and I tend to trust that they know real-world scenarios for enterprise.

Even that's terribad. That's like two to three CPU generations wiped out at a time when performance increases are hard to come by.

Fame Douglas fucked around with this message at 01:10 on Jan 7, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy
You guys are missinformed.

The Spectre fix requires a microcode update, an OS-level patch isn't sufficient

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Kazinsal posted:

Let's go over this one more time.

Code running in ring 0 can load microcode updates. Windows and Linux have historically done this when BIOS patches have been unavailable or when there's something serious that might need to be loaded like a major CPU errata fix.

The mechanism by which ring 0 code can load microcode updates is completely documented in the IA-32 and Intel 64 Software Developers Manual -- see volume 3A, chapter 9, section 11 - Microcode Update Facilities.

Let's go over this one more time.

No one knows whether Microsoft will patch the microcode using an OS update. So far, they haven't. Microsoft themselves suggest people update their BIOS. They also tend to be very reluctant about distributing any microcode updates. This means that, at the moment, a BIOS update is the only way to be safe. This may or may not change in the future (hopefully, it will).

Fame Douglas fucked around with this message at 10:44 on Jan 9, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Intel's CEO has promised patches for all processors introduced in the past five years to be released till the end of January. This means anything pre-Haswell is out.

Fame Douglas fucked around with this message at 13:34 on Jan 10, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Mr Shiny Pants posted:

I thought it was a microcode update and the mobo vendor would not be needed for it to update. Kinda like how Intel makes drivers for their onboard nics.

The Microcode is saved on the BIOS chip and loaded into the CPU at startup.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Combat Pretzel posted:

Yea, we should keep repeating this one, so that it gets ignored some more.

Why would we repeat this as long as we don't know whether Microsoft will even release a microcode update? So far, the only fix is a Bios update. That's what we should repeat, not the hope for a better future that may or may not come.

Combat Pretzel posted:

Because updating a system with tons of files live and sending a signed closed microcode package to the CPU via specific means, both that Intel supplied, is so much of the same.

lol if you think there aren't tons of weird edge-cases and incompatibilities with patching the microcode through the OS. https://wiki.debian.org/Microcode#Updating_CPU_microcode_within_Debian_.28Intel_or_AMD.29

Fame Douglas fucked around with this message at 16:54 on Jan 11, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy
DigitalFoundry has a video on gaming performance after these patches: Seems pretty noticeable at times, especially with games like The Witcher 3. https://www.youtube.com/watch?v=LC1WuKdPVCQ

Intel has some PR benchmarks up as well that show a decrease in performance of around 5-10% for office workloads: https://newsroom.intel.com/editorials/intel-security-issue-update-initial-performance-data-results-client-systems/

Fame Douglas fucked around with this message at 17:40 on Jan 11, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy

feedmegin posted:

I'm not sure why you think that link proves that. BIOS or OS, each will be doing exactly the same thing to update the microcode. There's nothing magic the BIOS does and the OS doesn't to make that happen, it just does it earlier.

I'm not sure you know what you're arguing. I assume a microcode update delivered using a BIOS update is tested before release. It's impossible for Microsoft to test all motherboard/CPU combinations for compatibility issues. I assume that's why they're being so conservative with releasing microcode updates through Windows Update.

But, let's hope they do it this time.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

DrDork posted:

The before-and-after benchmarks of games and normal office applications have been noted, though, as having pretty minimal impact unless your CPU is hilariously old. So, yeah, people complaining bitterly about losing 1-3% or whatever aren't being taken particularly seriously.

The rest of it--what's affected, how bad is it really, when are we getting fixes that work, etc., you're absolutely right to be concerned about, though.

Even Intels PR benchmarks show impact much more severe than 1-3%: https://newsroom.intel.com/wp-content/uploads/sites/11/2018/01/Blog-Benchmark-Table.pdf

And check out the Witcher 3, tested by DigitalFoundry: Pretty severe framerate impact in some complex scenes: https://www.youtube.com/watch?v=LC1WuKdPVCQ

Fame Douglas fucked around with this message at 21:31 on Jan 13, 2018

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Microsoft has released a new version of their optional (only available via the Update Catalog) microcode update. It now contains updates for more CPUs:

https://support.microsoft.com/en-us/help/4090007/intel-microcode-updates

Fame Douglas
Nov 20, 2013

by Fluffdaddy
I love the idea of the Intel superfan "just setting things right" on the AMD subreddit.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Just checked it out: The "cheap" Mac Mini is $500 - for a Haswell i5-4260U, a 500 GB spinning drive and 4 GB of RAM. That's almost at the level of being a criminal rip-off.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
There's no way that thing isn't slow as gently caress, even for "bargain" (it's not a bargain) computer users.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
I really like the Fractal Define R6

Fame Douglas
Nov 20, 2013

by Fluffdaddy
You don't want your friends laughing behind your back when they find out you have a SATA in your computer.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
"Intel Advanced Snake Oil"

Fame Douglas
Nov 20, 2013

by Fluffdaddy

LRADIKAL posted:

Who said anything about ethical? Keep telling yourself that you're the smart one and Elon is the stupid one, that'll get you far.

Musk has plenty of dumb ideas in the transportation sector, at least.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
I am very offended

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Intel shouldn't have hired Musk to do their fab work. No wonder 10 nm is delayed again.

Fame Douglas
Nov 20, 2013

by Fluffdaddy
I'm excited to see the discussion about Elon Musk when Apple announces their switch to ARM processors.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

Ok Comboomer posted:

Can’t imagine that they’re not looking at the Surface line or portables from Dell, et al. with similar ideas.

The Surface Pro X already exists, it even has an x86 compatibility layer just like Rosetta.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

repiv posted:

Emphasis on x86, Microsoft's emulator can't run x86_64 apps which kinda sucks. Tons of stuff is 64-bit only nowadays.

They're working on 64 bit compatibility. But emulation is always going to be comparatively slow, something I'm very dubious even Apple could really solve.

Ok Comboomer posted:

Eh, this one honestly looks like it’s going to be substantially less painful than the Intel Transition and Rosetta 1. Idk that they would’ve shown off Maya running via Rosetta 2 if they didn’t expect people to try it for themselves and look for holes—but also Apple totally worked extensively with Autodesk behind the scenes to make that successful demo happen.

I haven't seen the demonstration, but if it's rendering on the GPU, it would make sense that the emulation penalty wouldn't be as large. But that doesn't mean regular more CPU-based programs are going to run amazingly well.

Fame Douglas fucked around with this message at 21:59 on Jun 22, 2020

Fame Douglas
Nov 20, 2013

by Fluffdaddy
There have been attempts at producing ARM servers for years, Apple won't be an important factor in server production. At the moment, they're only interesting for some specialized applications.

Fame Douglas
Nov 20, 2013

by Fluffdaddy

DrDork posted:

And then there's the compatibility issue. I would not bet any money on Microsoft being able to replicate Apple's claimed success in emulation layer software to allow you to run x86 on ARM. Not just because Microsoft moves slower, but because they have an enormously heavier lift there: Apple is basically designing the translation/emulation software to target a single, limited family of chips, which they control, on a specific, limited set of motherboards, which they also control. Microsoft controls none of that, and would have to take into account all the hardware compatibility issues that come with using a limitless combination of random poo poo. That ain't easy, so performance is likely going to suffer heavily. And in the meantime, while Apple can basically strong arm their entire ecosystem into jumping to ARM because they say so, everyone else is going to continue developing for the platforms their customers already have.

Windows for ARM already has an x86 emulation layer, and 64 bit is being developed right now. It does seem to be slow, though, but who knows how it actually stacks up to Apple's solution.

Adbot
ADBOT LOVES YOU

Fame Douglas
Nov 20, 2013

by Fluffdaddy
Why would Intel have to change. They're making bank, technology leadership doesn't actually matter.

Fame Douglas fucked around with this message at 00:42 on Jul 29, 2020

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply