The analogy with floppy drives is flawed. When floppies were being dropped, the replacement was already there - USB mass storage devices - and it was cross platform. Headphone jacks are cross platform.
A lightening headphone is not cross platform.
A USB-C headphone would be cross platform if iPhones and computers all came with USB-C connectors but today they do not.
Tom's Hardware says 30% core performance increase from Ivy Bridge to Skylake, wither lower power consumption by SkyLake. That's not not getting faster.
They summarize..
"Sandy/Ivy Bridge users:
This time it's a yes. Sandy Bridge and Ivy bridge are both getting pretty old by now, and upgrading to Skylake will yeld a 30-45% (45% if your a sandy user) increase in performance. You will also get the latest technologies from Skylake, namely power efficiency, a variety of USB ports, high speed LAN, and probably the biggest upgrade will be high speed storage. Together, all these features makes it a worthy upgrade to Skylake."
My IvyBridge gaming machine with a 680 still works fine. That's 5 years old. But a SkyLake machine with a 980 would be a big step up in performance, although excel would not feel any different.
I was running week long sims on Broadwell machines recently. Dropping it to a day would improve the quality of my life. In effect the sims I can run are bounded by the compute power available to me.
I really don't think you need a SkyLake machine in order to run The Sims. Pretty sure the last game came out in like 2013...
If The Sims is Turing complete, maybe I can use it.
This isn't a surprise. The next Intel chip won't be much different than the last one either. The next computer you own will be very similar to the one you owned 5 years prior. We have really reached an end of the digital road now. You won't likely see huge progress like we have seen in the past due to limitations of physics and power concerns. I know people will scream "Moores Law!!!" but that doesn't apply in 2016. It looks like the Star Trek type future won't be happening.
Nope. The Intel chips keep getting faster. They keep on with the incremental improvement that renders big changes over a few years (the replacement lifetime of a computer). What's changed is that we passed the point where most people's compute tasks are slow enough to present an uncomfortable latency to interactive users. That was fine 5 years ago. My IvyBridge gaming machine with a 680 still works fine. That's 5 years old. But a SkyLake machine with a 980 would be a big step up in performance, although excel would not feel any different.
My simulations would be faster though. I was running week long sims on Broadwell machines recently. Dropping it to a day would improve the quality of my life. In effect the sims I can run are bounded by the compute power available to me.
Well the problem with housing affordability is that property in new york is in demand because it is desirable. All we have to do is make properties less desirable and the demand and consequently prices should come down. Like if there was a nuclear disaster or if the city became infested with deadly snakes or something, that would really help to keep housing affordable.
Yep. WiMax started out as 802.16 and 802.16a, which described a fixed wireless access arrangement - fixed antennas at both ends. WiMax as it was pushed a few years ago (gaining the WiMax name) was 'mobile' WiMax which involved a number of extensions in 802.16e for mobile stations, much like a cell phone protocol with handovers and all that that entails. However it remained an 802 protocol. So it could bridge with ethernet and 802.11 just like any other 802 protocol. It had QoS to carry voice traffic over the top of the data service. Given that data traffic far outstrips voice traffic, this is the right way to do it.
Mobile WiMax is necessarily less efficient that fixed WiMax because of the worse wireless channel properties between a mobile station and base station. However WiMax was way ahead of the 3/4G standards in using OFDM long before it appeared in 4G proposals and achieving superior performance as a result.
There were many fixed wireless solutions and 802.16 set up to agree a common interoperable FWA protocol. So fixed 802.16 achieved its goals. It was around before WiMax and is around today. It replaced many proprietary protocols with a common 802 protocol. The mobile WiMax is the bit that ran headlong into a fight with the carriers, regulators and cellular manufacturers.
Seriously, whatever happened to WiMax? It sure would be nice if phones had three major ways to handle data: cellular network (4g or whatever), WiMax, and WiFi. That way you would have a third option in-between "in my own home or at a cafe" and "nearly anywhere". Supposing WiMax were actually available that is.
It worked fine when I had WiMax. LTE would not be available today if it were not for WiMax lighting a fire under the rear ends of the 3G people. They did an effective job of tying up the carriers in order to kill WiMax (in the US and Europe at least). This is why LTE falls so short of its design targets. It was rushed out to crush the threat from WiMax.
I was one of many authors of the WiMax spec (I'm primarily responsible for the security protocol PKMv2) and worked on the design of WiMax chips, so my views have the jaded properties you might expect from an insider.
The world with wireless data as an 802 protocol with voice running on top of it, rather than the other way around, would have been simply better.
I'm sure this comment will sound monstrously dated before the end of the decade, but in my opinion 4G LTE is fast enough. I'd rather have better network coverage (I'd be truly festive for my cellular data to not cut out when I'm on the subway) and larger/cheaper data caps than better speed.
Smaller cells (think in building/in room) = more cells = greater capacity/m^2.
It's not about the data rates per connection (they could already be faster if they wanted to be). It's about the cell density and killing WiFi which makes no revenue to carriers.
A better analogy. A key under a 10 ton rock. You know the lifting capacity of a sucessful adversary. It's still obfuscation. But it has well defined properties.
Intel has a defined set of policies and procedures, managed by a dedicated team, to actively monitor and respond to vulnerabilities identified in released products. In the case of the Intel Management Engine, there are mechanisms in place to address vulnerabilities should the need arise.
Are they revealing they can turn their "secret chip" off remotely? Or perhaps even nuke the whole CPU?
If that would be the case, I guess some hardware hackers are going to have a field day.
Nope. If an Intel engineer rolls up to your house, breaks in and attaches a probe to your motherboard you should probably ask what the hell he or she is doing.
Security through obscurity is not a particularly successful technique and never has been, as you can tell from the vast number of published exploits against systems that were not actually secure based on vulnerabilities that were discovered despite their obscurity.
Security through obscurity is essential in a lot situations and is highly sucessful, particularly in the systems that don't end up in a paper at some CHES conference.
Undersirable as DRM might be, it by definition is security through obscurity because everything needed to break the DRM is in the hands of the owner of the device. If there's a key, it's in the box and you need to obscure it.
Side channel protections are often security through obscurity. Again everything is in the device and is observable. Methods of obscurity are the primary methods of side channel mitigations in hardware. It's not (necessarily) some weakly thought out practice. You can reason about the properties of your obfuscation in practical terms of things like the equipment needed, or amount of data needed for Bayesian statistics to have any power.
The chip in your credit card engages in obfuscation practices that are well developed but generally inadequate anyway because there's real money involved and governments and criminal enterprises are willing to invest in overcoming card security mechanisms.
Pointing to the systems that failed is confirmation bias. There are many obfuscation based security mechanisms, particularly in silicon, that have stood the test of time.
In spite of the incessant propaganda over the last century from the general relativity camp, gravity is an instantaneous or nonlocal phenomenon, just as Isaac Newton assumed. If changes in gravity traveled at the speed of light, as relativists claim, Newtonian gravity equations would not work at all and all planetary orbits would become unstable. There would be no planetary systems orbiting stars and there would be no galaxies. The reason is that it would take more than eight minutes for changes (caused by its motion around the Milky Way galaxy) in the sun's gravity to reach the earth and even longer for the more distant planets. So the earth's orbit around the sun would depend on where the sun was eight minutes ago, the time it takes changes in the gravitational field to reach the earth, and not on where it is now. This is not observed.
[emphasis added]
If true, this was the most interesting part of your article, Louis. Got an external link to back it up?
It's not true. I don't have a link to back it up, but I did read it somewhere plausible written by a real physicist. The motion of the planets fits the relatavistic model. One of the problems with Newton's laws is that the motions of the planets did not fit the predictions of Newton's laws. Relativity made up for the discrepancy. The motion of the planets is conditionally stable, just like any control system in the real world (as opposed to your simulator).
Slashdot Mirror
The analogy with floppy drives is flawed.
When floppies were being dropped, the replacement was already there - USB mass storage devices - and it was cross platform.
Headphone jacks are cross platform.
A lightening headphone is not cross platform.
A USB-C headphone would be cross platform if iPhones and computers all came with USB-C connectors but today they do not.
How can you tell when Betteridge's law of headlines doesn't apply?
Force the issue. Write a headline "Betteridge's law of headlines doesn't apply".
I drink ice tea instead of coffee.
Monster!
Many places with municipal water distribution ban personal wells.
So I'm not allowed to play my guitar any more?
Isn't that in SystemD already?
On Microsoft Linux, it's svchost.exeD
Tom's Hardware says 30% core performance increase from Ivy Bridge to Skylake, wither lower power consumption by SkyLake.
That's not not getting faster.
They summarize..
"Sandy/Ivy Bridge users:
This time it's a yes. Sandy Bridge and Ivy bridge are both getting pretty old by now, and upgrading to Skylake will yeld a 30-45% (45% if your a sandy user) increase in performance. You will also get the latest technologies from Skylake, namely power efficiency, a variety of USB ports, high speed LAN, and probably the biggest upgrade will be high speed storage. Together, all these features makes it a worthy upgrade to Skylake."
My IvyBridge gaming machine with a 680 still works fine. That's 5 years old. But a SkyLake machine with a 980 would be a big step up in performance, although excel would not feel any different.
I was running week long sims on Broadwell machines recently. Dropping it to a day would improve the quality of my life. In effect the sims I can run are bounded by the compute power available to me.
I really don't think you need a SkyLake machine in order to run The Sims. Pretty sure the last game came out in like 2013...
If The Sims is Turing complete, maybe I can use it.
This isn't a surprise. The next Intel chip won't be much different than the last one either. The next computer you own will be very similar to the one you owned 5 years prior. We have really reached an end of the digital road now. You won't likely see huge progress like we have seen in the past due to limitations of physics and power concerns. I know people will scream "Moores Law!!!" but that doesn't apply in 2016. It looks like the Star Trek type future won't be happening.
Nope. The Intel chips keep getting faster. They keep on with the incremental improvement that renders big changes over a few years (the replacement lifetime of a computer). What's changed is that we passed the point where most people's compute tasks are slow enough to present an uncomfortable latency to interactive users. That was fine 5 years ago. My IvyBridge gaming machine with a 680 still works fine. That's 5 years old. But a SkyLake machine with a 980 would be a big step up in performance, although excel would not feel any different.
My simulations would be faster though. I was running week long sims on Broadwell machines recently. Dropping it to a day would improve the quality of my life. In effect the sims I can run are bounded by the compute power available to me.
Well the problem with housing affordability is that property in new york is in demand because it is desirable. All we have to do is make properties less desirable and the demand and consequently prices should come down. Like if there was a nuclear disaster or if the city became infested with deadly snakes or something, that would really help to keep housing affordable.
Snakes on a Cable Train!
Yep. WiMax started out as 802.16 and 802.16a, which described a fixed wireless access arrangement - fixed antennas at both ends. WiMax as it was pushed a few years ago (gaining the WiMax name) was 'mobile' WiMax which involved a number of extensions in 802.16e for mobile stations, much like a cell phone protocol with handovers and all that that entails. However it remained an 802 protocol. So it could bridge with ethernet and 802.11 just like any other 802 protocol. It had QoS to carry voice traffic over the top of the data service. Given that data traffic far outstrips voice traffic, this is the right way to do it.
Mobile WiMax is necessarily less efficient that fixed WiMax because of the worse wireless channel properties between a mobile station and base station. However WiMax was way ahead of the 3/4G standards in using OFDM long before it appeared in 4G proposals and achieving superior performance as a result.
There were many fixed wireless solutions and 802.16 set up to agree a common interoperable FWA protocol. So fixed 802.16 achieved its goals. It was around before WiMax and is around today. It replaced many proprietary protocols with a common 802 protocol. The mobile WiMax is the bit that ran headlong into a fight with the carriers, regulators and cellular manufacturers.
Be fearful! There might be strangers sleeping somewhere in a property near you.
I bet the hotels are lobbying for this. Airbnb is one thing that is pushing the cost of visiting New York down.
Seriously, whatever happened to WiMax? It sure would be nice if phones had three major ways to handle data: cellular network (4g or whatever), WiMax, and WiFi. That way you would have a third option in-between "in my own home or at a cafe" and "nearly anywhere". Supposing WiMax were actually available that is.
It worked fine when I had WiMax. LTE would not be available today if it were not for WiMax lighting a fire under the rear ends of the 3G people. They did an effective job of tying up the carriers in order to kill WiMax (in the US and Europe at least). This is why LTE falls so short of its design targets. It was rushed out to crush the threat from WiMax.
I was one of many authors of the WiMax spec (I'm primarily responsible for the security protocol PKMv2) and worked on the design of WiMax chips, so my views have the jaded properties you might expect from an insider.
The world with wireless data as an 802 protocol with voice running on top of it, rather than the other way around, would have been simply better.
I'm sure this comment will sound monstrously dated before the end of the decade, but in my opinion 4G LTE is fast enough. I'd rather have better network coverage (I'd be truly festive for my cellular data to not cut out when I'm on the subway) and larger/cheaper data caps than better speed.
Smaller cells (think in building/in room) = more cells = greater capacity/m^2.
It's not about the data rates per connection (they could already be faster if they wanted to be). It's about the cell density and killing WiFi which makes no revenue to carriers.
Patiently awaiting the APPS guy.
I predict it will be the SNAPPS guy this time.
A better analogy. A key under a 10 ton rock. You know the lifting capacity of a sucessful adversary. It's still obfuscation. But it has well defined properties.
Intel has a defined set of policies and procedures, managed by a dedicated team, to actively monitor and respond to vulnerabilities identified in released products. In the case of the Intel Management Engine, there are mechanisms in place to address vulnerabilities should the need arise.
Are they revealing they can turn their "secret chip" off remotely? Or perhaps even nuke the whole CPU?
If that would be the case, I guess some hardware hackers are going to have a field day.
Nope. If an Intel engineer rolls up to your house, breaks in and attaches a probe to your motherboard you should probably ask what the hell he or she is doing.
Obscurity is a perfectly acceptable security tool.
No, it is not. Your key being blank #23 and key pattern 8442332 isn't "obscurity". But the spare key being kept under a rock by the back door is.
Yes it is. Pointing out a bad example doesn't mean there aren't good obfuscation techniques.
Security through obscurity is not a particularly successful technique and never has been, as you can tell from the vast number of published exploits against systems that were not actually secure based on vulnerabilities that were discovered despite their obscurity.
Security through obscurity is essential in a lot situations and is highly sucessful, particularly in the systems that don't end up in a paper at some CHES conference.
Undersirable as DRM might be, it by definition is security through obscurity because everything needed to break the DRM is in the hands of the owner of the device. If there's a key, it's in the box and you need to obscure it.
Side channel protections are often security through obscurity. Again everything is in the device and is observable. Methods of obscurity are the primary methods of side channel mitigations in hardware. It's not (necessarily) some weakly thought out practice. You can reason about the properties of your obfuscation in practical terms of things like the equipment needed, or amount of data needed for Bayesian statistics to have any power.
The chip in your credit card engages in obfuscation practices that are well developed but generally inadequate anyway because there's real money involved and governments and criminal enterprises are willing to invest in overcoming card security mechanisms.
Pointing to the systems that failed is confirmation bias. There are many obfuscation based security mechanisms, particularly in silicon, that have stood the test of time.
As a competent security professional you cannot be unemployed right now in the current job market. Security jobs cover a lot more than 'IT'.
> representing a 29 percent increase since 2013, according to a report by Pokemon Institute (...) to destroy the hackers
Catch. Gotta catch them all. Not destroy them!
Times change.
You don't know what the fuck you're talking about.
Oh Mr (or Mrs, or Ms) Coward. You seem to be scientifically illiterate.
> representing a 29 percent increase since 2013, according to a report by Pokemon Institute.
In they past they would have sent out Pikachu and a Sqirtle to destroy the hackers. These days they sit around in an institute writing studies. Sad.
[emphasis added]
If true, this was the most interesting part of your article, Louis. Got an external link to back it up?
It's not true. I don't have a link to back it up, but I did read it somewhere plausible written by a real physicist. The motion of the planets fits the relatavistic model. One of the problems with Newton's laws is that the motions of the planets did not fit the predictions of Newton's laws. Relativity made up for the discrepancy. The motion of the planets is conditionally stable, just like any control system in the real world (as opposed to your simulator).
It's easy if you just think of it in terms of kilopounds per megainch.
No km is kibibytes per millimeter.
A real physicist on the radio called it "swimming in jello" (That's jelly for the rest of the world).
That seems a pretty good analogy. Radiating gravity waves seems fine too. People are so picky.