Slashdot Mirror
Intel Says Chip-Security Fixes Leave PCs No More Than 10% Slower (axios.com)
Intel trying to defuse concern that fixes to widespread chip security vulnerabilities will slow computers, released test results late Wednesday showing that personal computers won't be affected much and promised more information on servers. From a report: The chipmaker published a table of data showing that older processors handled typical tasks 10 percent slower at most, after being updated with security patches. The information covered three generations of processors, going back to 2015, running Microsoft's Windows 10 and Windows 7 computer operating systems. Further reporting: Intel, Microsoft offer differing views on impact of chip flaw
Intel was knowingly breaking security to make their crap seem faster.
Put into context, my US state's sales tax was 5% and it went up to 6% a few years ago. Everyone was pissed by a 1% change (and I still think about it when I make some purchases based on how it increases the price).
So 10% slower for "typical" tasks seems like our computers had a minor stroke. There should be a discount on future Intel chips based on owning a current one, but that won't happen.
Nice try but the only thing which matter is whatever the statement is true or not.
Intel got billions of reasons to want to spread an image where this doesn't matter much.
Others may have less of a reason. But sure both sides could choose their data and try to exaggerate.
These people lied about every aspect of each of these major vulnerabilities. 10% is whatever, that's bad but worse is that NOBODY CAN TRUST INTEL WHATSOEVER, and they are the market leader.
Their obfuscation of the meltdown issue is unreasonably bad management, and their CEO sold a ton of shares right as the company secretly found out a year ago? The combination is absolutely toxic.
Clean house or watch it burn.
According to their chart, he oldest CPU they tested is 2.5 years old. Giving that some more proactive businesses have a 3+ year retention rate on hardware, "older" is hardly the word i'd use.
I.e. the 6700K.
I.e. all the chips have PCID
It's a bit hazy when PCID and INVPCID became supported.
This says PCID was first supported in Westmere
https://www.realworldtech.com/...
Another long overdue improvement to the page tables is the Processor Context ID (PCID). The PCID is a field in each TLB entry that associates a given page to a process. Previously, Intel's TLB could only contain entries from a single process and whenever the CR3 register was written (e.g. a context switch), the TLB was flushed. The PCID lets pages from different processes safely inhabit the TLB together, so that CR3 writes no longer flush the TLB. Whenever a process tries to access a page in memory, the PCID is checked to determine whether the page is actually mapped into the process' address space; if the PCID does not match then a TLB miss occurred. This is very much analogous to Intel's VPID, which enables the TLB to contain pages from different virtual machines and avoid TLB flushes during VM transitions.
The LWN patch says
http://lkml.iu.edu/hypermail/l...
PCIDs are generally available on Sandybridge and newer CPUs. However,
the accompanying INVPCID instruction did not become available until
Haswell (the ones with "v4", or called fourth-generation Core). This
instruction allows non-current-PCID TLB entries to be flushed without
switching CR3 and global pages to be flushed without a double
MOV-to-CR4.
I.e. it'd be interesting to see what happens on a CPU old enough not to support enough of PCID/INVPCID to optimized KPTI.
The claims of >10% hits are all for these old CPUs.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
Is 10 percent "at most" supposed to be reassuring?
This posting is provided 'AS IS' without warranty of any kind, implied or otherwise.
https://newsroom.intel.com/edi...
https://newsroom.intel.com/wp-...
i7 8700K Windows 10 SSD
SYSMark 2014 SE Responsiveness 88%
So even then it's a larger impact than 10%. On the latest processor. But the system used had their 600p SSD which is really slow. How about the 960 Pro or their Optane stuff?
As for what the responsiveness test actually test I don't know (may be possible to google that) but file-performance and virtualization may be worse.
There will be cases where the impact is beyond 10%, a 10% average would be pretty crappy. Mind you that you can get a B350 board and a Ryzen 5 1600 processor for about half the price of a Z370 board and an Intel i7 8700K.
If Intel is admitting a 10% slowdown then it must be much much larger. Because Intel and benchmarks don't live in reality.
after the way, last week, that they put it about that the problems affected all chips from all manufacturers to the same degree. They showed themselves to have better skills at sophistry that chip design.
In reality, you'll be getting 10% support from Intel on that matter.
Then I guess I'm expecting at least 10% of the cost of the processor cost back as a refund.
Velociraptor = Distiraptor / Timeraptor
The only problem is from the chart some of the workflows were even faster by 3% to 5%. I might need to send them some more money if I start using those workflows.
After an OSX update a real world compile of a project that takes around two minutes to complete, too almost exactly the same amount of time, or slightly faster... since compilation involves lots of small files and system calls I would expect it to be harder hit than most tasks. However because they had a partial Meltdown patch in around December, not sure if we would see much of an effect... no-one in December complained about slowdowns from the OSX update at that time though either.
I don't think things most people do will be that affected by the patch.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I'm still not clear on if the slowdown is due to the per-OS workaround, or if Intel is talking about their eventual fix to the hardware/firmware problem causing the slowdown...TFA seems to indicate a "fix" to Windows OS' specifically, which would imply the per-OS workaround.
Anyone?
On this machine (i5-5250U in an NUC5i5RYK) performance is fucking AWFUL after the Windows 7 patch and BIOS update. Webpages like YouTube peg a CPU core somehow. So does SSMS.
My main machine is fine, because it's so old (2600k) that there is no BIOS update available for my motherboard. Allegedly you can download the microcode patch and shove it in yourself if your odn't get an official BIOS release. But fuck it. I'll be upgrading to the next Ryzen revision in a couple of months, hopefully. But FUCK current RAM prices.
now it's 10%. Also, define "Personal Computer". My bro regularly runs 3-4 VMs in a virtual computer lab on his i7 while learning new tech or testing our scripts. And it's a 4th Gen i7 to boot.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
I'm covered in the dust of the leader. He favors me!
I am even dustier -- dustier than thou!
-- Sometimes you have to turn the lights off in order to see.
...what about the $15,000 servers I bought 6 months ago?
CPU cost $999 so $99.90 refund coming my way - sweet
Sounds to me like you could get faster performance by building Linux AMD blade servers for a heck of a lot less money and they would perform better.
At least, that's my takeaway.
-- Tigger warning: This post may contain tiggers! --
I despise Intel's ME sleazyness as much as anyone, but ME has nothing to do with Meltdown or Spectre.
If you wanted to argue that ME is tangentially related because it's a backdoor that might allow someone to then use Meltdown to attack a machine then that sounds very plausible to me, but again, that's only tangentially related and also has nothing to do with performance of Intel's chips.
I don't think there is a small amount of Intel chippery prior to 2015 running around. I'm probably an outlier, but mine is from 2008, (c) 2007.
Slashdot: Playing Favorites Since 1997
Does any other website have this problem?
Maybe time to down-mod this website?
just set the flag and get your speed back
I don't claim to be any kind of semiconductor engineer, but I am a customer that paid for something, and post-facto have a choice between insecure, or less performance than I paid for.
If you think this is fine, then you are either a paid shill or a deluded fanboy.
Slashdot still doesnâ(TM)t support Unicode after it was added to the HTML standard in 1997.
So when are we customers going to get ten per cent of our money back?
I am sure that there are many other solipsists out there.
Moore's Law has been dead for over 10 years now.
Well, I'm no shill (and am no longer on Intel's payroll) but am a fanboi.
That said, the position you and Intel are in is commonly referred to by its acronym:
F.U.B.A.R.
There is no realistic recourse for Intel to offer that would satisfy the majority of the install base.
Full replacement at cost is likely to leave a lot of people very angry, and devastate Intel's Fiscal Year, but it is likely the best possible outcome.
Trying to make Intel replace everything affected for free (like with the FDIV bug) is a non starter. Intel can't likely even fab the old chips any more, and even if they could it would still require a redesign, so it's a non-starter. Giving everyone new chips would not be like for like, so you have issues where old software won't run, but is still required, also a non-starter. Additionally, both those options would likely bankrupt the company entirely, meaning people *still* wouldn't get replacements, and you'd have 100K freshly unemployed.
The most likely outcome I see is a rebate/coupon towards the purchase of any system containing a new Intel CPU from any vendor where the dollar value of the rebate is tied to the age/sku of the old CPU, with no or soft requirements to return the old CPU.
What would you (as a consumer) expect?
I likely won't get squat, since all my CPUs are samples that employees were given at various times, or bought via employee purchase channels.
whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
If one of my products turned out to do only 90% of advertised, I'd offer a rebate or return for credit (customer's choice). Maybe that's why I'm a small company, not a mega-corp.
I agree replacement is not an option - hell many of the chips are soldered in place. I think a refund rather than coupon is the way to go. I'd rather have $50 than a coupon for $50 on their next defective product. Financial penalties should nearly bankrupt companies otherwise there is no incentive to not release defective products. Limiting top execs salaries to 500k for the next 5 years would be a good gesture as well.
Yeah, except that Intel admits this is a flaw.
Why are you trying to shift blame that even Intel is (reluctantly) accepting to someone else?
Exaggeration isn't necessary. Even taking Intel at their word, that 10% differential has been their selling point for years. That means the very reason people bought Intel over the competitor is now bunk. At the very least, consumers are due a rebate, and that only (barely) addresses the lost value to those consumers, to say nothing about the damage Intel caused to its competitors in peddling this lie.
The 10% figure sure sounds negligible... until you give it more than five seconds of critical thought. Any way you slice it, Intel reaped illegitimate profits. And instead of making it right, they're busy trying to discredit the very people who have supported them over the years: their customers. This is the gold standard of how not to handle a situation like this.
The most likely outcome I see is a rebate/coupon towards the purchase of any system containing a new Intel CPU from any vendor where the dollar value of the rebate is tied to the age/sku of the old CPU, with no or soft requirements to return the old CPU
That rebate is extremely unlikely to cover more than the value of the CPU, what about the secondary cost of replacing the rest of the computer it was attached to. If it uses a different socket, or is soldered directly or more likely isn't even compatible anyway, then it has been rendered worthless. I don't buy a new computer every year, nor do I want to: consumer looses, Intel wins.
I assume that Intel will be offering their flawed processors for 10% LESS across the board?
Even faster if you don't use rigged benchmarks and compilers!
Twinstiq, game news
Intel does this because Intel gets away with this.
Intel will continue doing this because Intel will continue getting away with it. Nothing you post, nor any vote you cast, nor any action you can take will change this.
The world is not a just place. It never has been.
Intel gave the programmers this "feature" to use or abuse. They allowed this to be done, and knew the dangers, yet continued anyway designing CPUs like that for many more years.
Obviously just one test does not mean a lot, but a full build of an application (especially a very large one) does make use of a lot of different system resources, so it's a pretty good indicator of general issues.
That article is really good because it does something I did not have time to do - try out something between 10.12 and 10.13 (I only measured the impact before and after the latest security patch on 10.13). So thanks for providing a link... it looks like system calls do indeed have large performance impacts, but that when you are talking about real cases (like compiling, as I did and they did also as a test) the real performance loss is under 10% - when there is any, and on some systems moving to High Sierra still has an overall gain on performance for some tasks from Sierra. Not all work people do on a computer will see an impact.
I'm also thinking over time we'll see more efforts put in to mitigate the speed losses, all OS makers are just being extra cautious to start with.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
This page over here, using actual benchmarking software before and after the Meltdown and Spectre patches, shows iPhone performance losses around 40% after applying the patches: https://melv1n.com/iphone-perf...
Given that most of the last week's media spiel has been saying that "ARM's CPUs are supposed to be largely unaffected by these things" I doubt Intel's CPUs would behave much differently, certainly not better and certainly not "only 10% impact at worst".
> going back to 2015
> older CPUs
A good CPU from 2015 (Haswell) is a pretty new CPU. High-end CPUs from 2012 (Ivy Bridge) are still perfectly capable, especially in laptops. If you don't need stuff like USB 3.0, you can easily end up today with a pretty beefy Ivy Bridge/Haswell laptop by just doing RAM and storage updates (and maybe WiFi). This is just a carefully designed PR piece to make the issue look less bad, nothing to see here.
fair, no?
So says the intel shill.
Intel can't likely even fab the old chips any more, and even if they could it would still require a redesign, so it's a non-starter.
I disagree. They would only need to do one redesign, because the architecture is the same across many different chip families. And why the flaw stretches back to 1995 in the first place.
The bigger question would be to repackage the chips to the various sockets (and voltages) where they may no longer have the specific packaging machines in place to do so.
But this is Intel with the world's best engineers. If they say they can't do it, it will give the appearance that they are being lazy and dragging their feet because they need to protect the stock sell their CEO had back in November.
On the other hand, if they really did have the best engineers eager to get the job done, they might see this as a way to create new technology that they could turn around and sell that could perhaps also extend the lifetime of legacy systems.
People should be upset for being forced to be on an upgrade mill anyway!
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
with higher SSD i/o load it can easily reach 50%, and on pre-Haswell or so CPUs it can also be more, well done. get an AMD today ;-)
It's very hard to put a price on it.
Right now I use the Samsung EVO 850 SSD and there it was just 4K performance which was impacted 10-12%. So not an average for my system.
However it's bought with the intent of gaming, streaming, virtual machines and security/encryption (and yeah, disk access isn't irrelevant too then again to get the good stuff HEDT / ThreadRipper is the way to go.)
I've basically paid twice the price over B350 + Ryzen 5 1600 for Z370 + i7 8700K for what isn't twice the performance but just a bit more performance because I wanted to have that extra performance. Ryzen 2 and this bug kinda removes that.
In a professional scenario whatever streaming, virtual machines, rendering, databases or what not the value of the product you may be much more valuable than the equipment you use to produce it and there a 20% performance difference may mean much more than a 20% price of the tool you use.
Also of course the risk / it's likely the case that running virtual machines and running faster storage (I do own an Optane stick too) will have more of an impact than those 10-12%.
You are so full of shit.
Gaming calls for pushing the edges everywhere possible.
Go to the forums, look at the convos and don't bother reporting back here.
It little behooves the best of us to comment on the rest of us.
This is the time for all good chip makers to come to the aid of new customers and emphasize the 10% penalty of Intel vs clean design.
It little behooves the best of us to comment on the rest of us.
Be apathetic. I don't care.
It little behooves the best of us to comment on the rest of us.
i spent $260 on an i7-8700 a month ago, partly because intel had the better performance single core in gaming. If intel's patch decreases 10% in performance, and amd's upcoming spectre patch doesn't, then you could say that i wouldn't have made the same purchase today. so i don't believe it's crazy to get a 10% refund, or at the very least, a rebate in that amount for a future cpu.
Open Source Java Web Forum with LDAP authentication
https://www.computer.org/csdl/proceedings/sp/1995/7015/00/70150211.pdf
I hope those class actions tear Intel a butthole so big it bleeds.
Intel CEO decided to sell off shares after Intel were warned but before the public were told: https://www.bloomberg.com/news/articles/2018-01-09/u-s-senators-urge-government-to-probe-intel-ceo-s-share-sales
but when you are that rich laws don't apply yes???
I just spent a lot of money building a gaming rig, and overclocking it to get another 20% boost. Every bit matters in games. So, -10% is still huge to me. Intel should offer some compensation. Maybe not a whole new chip, but maybe at least a voucher of say $100 on a new CPU.
That's a shame as I'd really rather hoped for more than 10% of the costs of my Intel-based computer refunded.
Intel *will* be offering partial refund for fault goods, right?
Requiem for the American Dream
Okay, then give me a 10% refund on your defective product because this isn't what I paid for. I literally JUST GOT my Nvidia compensation check for their single lane 512MB block of GDDR5 that they pretended had full speed access.
Intel made this upgrade problem themselves.
If you have a DDR2 or DDR3 system, you would need to upgrade to DDR4, and replace the CPU,MB, RAM and even the Chasis.
And if you bought something like an iMac, laptop or other ITX or NUC system, you have to throw away the entire thing.
As much as I want to fault Intel for this, it really is not a problem with the CPU, it is a problem with how operating systems are designed with naive assumptions that out of order execution is garbage collected. The fix at the ASIC level would require something like a 'kernel bit' which would prevent user space from reading/writing to it, or doing it the same way AMD does at a higher performance penalty.
Ultimately people are being way too stupid about this. The largest performance impairment will be Amazon Web Services, and similar cloud/VPS. I would short all these companies around earnings for the next two quarters as they are forced to buy 20% more hardware just to retain customers.
Don't consider the Spectre patch. All the major CPUs are vulnerable to Spectre. It's Meltdown where there is a significant difference.
Also, it's not yet clear to me that Spectre can be patched in the current chip designs (any of them) without disabling speculative execution. That's more than the 10% penalty. (How much more? I've no idea.)
This is quite annoying as I'd been thinking it was time to start considering buying a new computer, but now it looks as if buying one with one of this or the upcoming generation of chips is a bad mistake with no way around it. The chips that I need are the ones that haven't yet been designed.
I think we've pushed this "anyone can grow up to be president" thing too far.
Yes, it is definitely Intel's intention to not give you more than 10% of the money you've spent on their products. (0% *is* less than 10% isn't it?)
I think we've pushed this "anyone can grow up to be president" thing too far.
That doesn't matter! People pay good money for these PCs and should be entitled to a full speed machine, not a ten percent decrease in performance. If that's Intel's attitude, they won't last much longer in this market.
I charge forward recklessly, leaving chaos in my wake.
These are compiles on a laptop with an SSD (a bit older, but still). Pretty much all time is spent in the CPU on the build. It's around 700 something files to compile, so it does have a decent amount of data to fetch but most of the time does seem to be spent in the compiler... it may impact other tasks more but if compiler impact is negligible I'm pretty happy as that's most of my day-to-day load.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
We're still getting 90% (or better, supposedly) of what we asked for.
I think it would be reasonable to expect a 10% discount on the next Intel CPU, though, and Intel should eat the hit to their profits that would result.
What about folks who can't currently afford anything new, and have an Intel based system that just manages to play the games they like to play? Well 90% of what they paid for might not cut it. Is a 10% discount on something new any good to them? Not when they can't afford the other 90% to get something new. A 10% rebate won't be that helpful as their computer stutters through their games.
That last 10% performance edge costs a hell of a lot more than 10% of the total price, and you know it. Everyone has always known it.
Well, some of us have always known the reason was bunk. Now it's just obvious to everyone else, too.
No, the 10% of cost part is not fair at all. If you want to be fair they would have to refund the difference in cost over the equivalent AMD offering these customers would otherwise have purchased at the time. Remember, that last 10% of performance is a lot more than 10% of the total cost. The actual number is going to be closer to 50% of the retail value of the chip in most cases.
Turn it off? Surely this is a case for nuking from high orbit!
Sent from my ASR33 using ASCII
or
I like your suggestion. I went with Intel solely based on the performance specs. Now I find out that I will have to loose that plus some. And I paid a premium for it. So of course I expect something. If they offer a rebate/coupon then I would expect the coupon to offer significantly more value. When you factor the motherboard in (because intel refuses to adopt a standard socket design) then the loss of value us even greater.
Now if I factor in all of that then my "older" system (3-4 years but still keeps up with benchmarks) I still expect at least $100-$200 back to help me rebuild with non-affected tech. And that assumes I can find a compatable motherboard that takes my now older ddr3 ram. (Note: I am not expecting them to pay for everything, just help me upgrade my system so that I am no longer affected by this.)
Ya this does not look good for Intel. But as a consumer, should I not expect something back? Why should I have to take a "loss" at no fault of my own.
I'm looking for cash so I can spend it on an AMD CPU, not a voucher for more Intel crap. The law backs me up here - the UK Sale of Goods Act require Intel to either fix the problem quickly (impossible) or give me cash in compensation.
I've had about 50% of the expected life out of my systems with Intel CPUs. Of course, writing the CPU off (server is now unable to cope with the workloads I require) means also replacing the motherboard, RAM and OS (Windows 8 doesn't support modern CPUs). Unfortunately I'll need to replace the PSU as well (long story), and spend several hours working on the change over.
I already opened a support case with Intel about this. After that it's small claims court.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Except that's bullshit.
Until the last year or two quad-cores did fine but now quad-cores without hyper-threading is limiting games and more cores are better for some.
Also the i7 8700K can be overclocked to ~4.9 GHz without delid and 5-5.2 or so with.
The i7 8600K maybe you can reach 5.1 GHz or so without delid so you gain 200 MHz or 4% but you lose hyper-threading for 6 cores ..
I can stream game-play with 1.5-7% processor load (1.5-6 mbps x264), so I guess the best would be to put OBS on one core and the game on the other five, anyway in that scenario of course the i7 8700K is better than say an i5 7600K (i3 8350K by now.)
I'll be asking for 10% of my money back from intel.
For people who have bleeding edge top of the line stuff, the last 10% performance probably cost them 50% more. (That's a bit of a random guess, but you get the idea. Performance/cost is not linear.)
Additionally, both those options would likely bankrupt the company entirely, meaning people *still* wouldn't get replacements, and you'd have 100K freshly unemployed.
Isn't this what should happen in a capitalist system? Those who are unfit to survive, die. General Motors should have been allowed to fail; instead, nothing changed and we keep getting crappy vehicles from them. The same will happen here with Intel.
Privatize gains and socialize losses is a winning strategy!
Someone needs to make processors. Those 100k employees will not stay unemployed forever. Besides, they have unemployment to rely on, right? (lol)
"Someone needs to talk to the tree of liberty about its ghoulish drinking problem." by ohnocitizen
Moore's Law says nothing about clock speeds, it's about transistor counts.
Some of the extra transistors in the next generation of CPU's will go toward fixing these flaws, instead of toward additional cores, or additional cache, or faster integrated graphics.
The software fixes that cause the 10% slowdown will then no longer be necessary.