I was indeed mistaken in the peripheral interconnect being distinct from the processor interconnect, but I cannot see anything that relates AMD's Infinity Fabric to Infiniband.
Sounds like possible confusion with Hypertransport (or functional equivalent) buses; the inter-processor links are typically discrete from the peripheral interconnects. Also worth noting that AMD CPUs of the previous generations used HT to bond the groups of cores within a single module, as well as the inter-socket links, and a dual socket had 4 NUMA nodes to deal with.
Back when I had to interact with servers running those platforms, there were some very strange interactions between MySQL and the NUMA arrangement, leading to its performance being utterly crippled. I also encountered many circumstances where the servers were configured incorrectly, leaving half of the nodes in the system without local RAM (i.e. the RAM was attached only to one node per socket).
My understand is that the NES system was based on timing of scan lines on the CRT; most LCD monitors/TVs i've seen refresh the panel more or less all at once, which results in no possibility of determining the timing.
... and we have some breaking news, a man in a Guy Fawkes mask has been spotted stumbling through rush hour traffic with no pants, and his shoe laces tied together...
In the ongoing battle against the effects of Comcast's iffy hardware/network, I've been running speed tests every ten minutes for several days.
The actual issue being diagnosed in this case is random bouts of upstream constriction; so far, these don't seem to be tied to any activity inside of the network.
The nominal speed of this network is ~117 mbit/s downstream; times are local to the server. http://pastebin.com/zMUncLei
Pretend that the downstream speed is your line voltage, and anything 100 + is equivalent to nominal. Also pay attention to the times at which the depressions occur; it seems to know no rhyme or reason. During the day, many of these can be attributed to heavy user activity.
Also, a warning about static IPs and Comcast: If you have a static IP through them, you are required to rent one of their routers. They both have issue; the SMC D3G (which we use) has hopelessly broken IPv6 support. The Cisco DPC3941, which we used previously, had bizarre issues with latency that it would impart upon any traffic that crossed it; it also had an unacceptably high rate of random (cable) signal drop-outs and packet loss. We went through two of those before switching to the SMC. I was told at one point that they'd be getting some newer Netgear routers, but I'm not very willing to beta test anything for them.
From what I've read, the IoT stuff tends to be mainly an issue with default/hard coded credentials more than any issues with the underlying software. The end result is, depressingly, the same.
While I don't know of any incidents within the US, I have seen at least one case where a drone collision actually occurred: http://avherald.com/h?article=...
Call me crazy, but isn't the more logical solution to the issue of battery life to make the phone consume less power? I know people want their apps, but I'm not convinced that people want to carry around a laptop battery in their pocket. Also, my smartphone usually lasts about 6-7 days on a charge as-is, primarily because it does very little aside from phone stuff. Turn on the WiFi and that drops to around 3 days in normal use (simply being connected, without actually using the internet).
Back then, things probably seemed novel and exciting to a broader array of people. These days, I get the feeling that the people who were once excited about those things (myself included) now, often, see them as little more than faster and more complicated versions of the things they replaced. Another issue is people's preconceived notions about how the latest and greatest tech ought to be; I suspect that they feel let down by the slow progress towards the things they believe to be the way of the future. If these things are true, it's not surprising that people in all circles would tend to have a pessimistic outlook on technology.
The rate at which they're throwing out these major updates reminds me of the Windows NT 4.0 service packs... two within the first year (1996), one each in '97 and '98, and then 3 in '99...
As that mess was largely before my time, does anyone have any recollection of the breakdown of good vs harm caused by deployment of the older NT 4.0 service packs? Is it even a good analogy?
On day 91, Clarke logged back into his account to find that it was still active. Unbeknownst to him, logging back in simply to check would reset the clock back to zero. "Yahoo confirmed via email yesterday if you access your account it resets the timer," he told me. "So, if you login to ensure your account has been deleted and it hasn't, you have to wait at least another 90 days."
So, if an account was being accessed by way of cookie (or other) exploits, I assume this would also reset the timer?
Have there been any notable cases in which it was ruled, in any court system, that Uber drivers were NOT employees and/or entitled to some level of benefits?
Laserjet 4 Plus here. I rebuilt the upper feed roller and replaced the laser scanning module after the laser diode cooked a few years ago. 25000 pages so far. Manufactured in May 1994.
Slashdot Mirror
I think what they were driving at ("capacity") is that the Intel supports ~1.5 TB RAM, whereas the AMD supports 2TB.
I was indeed mistaken in the peripheral interconnect being distinct from the processor interconnect, but I cannot see anything that relates AMD's Infinity Fabric to Infiniband.
Sounds like possible confusion with Hypertransport (or functional equivalent) buses; the inter-processor links are typically discrete from the peripheral interconnects. Also worth noting that AMD CPUs of the previous generations used HT to bond the groups of cores within a single module, as well as the inter-socket links, and a dual socket had 4 NUMA nodes to deal with.
Back when I had to interact with servers running those platforms, there were some very strange interactions between MySQL and the NUMA arrangement, leading to its performance being utterly crippled. I also encountered many circumstances where the servers were configured incorrectly, leaving half of the nodes in the system without local RAM (i.e. the RAM was attached only to one node per socket).
My understand is that the NES system was based on timing of scan lines on the CRT; most LCD monitors/TVs i've seen refresh the panel more or less all at once, which results in no possibility of determining the timing.
Is it going to require a persistent internet connect + named account to use it as anything but a basic brick?
We call this "clod backup".
... and we have some breaking news, a man in a Guy Fawkes mask has been spotted stumbling through rush hour traffic with no pants, and his shoe laces tied together...
In the ongoing battle against the effects of Comcast's iffy hardware/network, I've been running speed tests every ten minutes for several days. The actual issue being diagnosed in this case is random bouts of upstream constriction; so far, these don't seem to be tied to any activity inside of the network.
The nominal speed of this network is ~117 mbit/s downstream; times are local to the server.
http://pastebin.com/zMUncLei
Pretend that the downstream speed is your line voltage, and anything 100 + is equivalent to nominal. Also pay attention to the times at which the depressions occur; it seems to know no rhyme or reason. During the day, many of these can be attributed to heavy user activity.
Also, a warning about static IPs and Comcast: If you have a static IP through them, you are required to rent one of their routers. They both have issue; the SMC D3G (which we use) has hopelessly broken IPv6 support. The Cisco DPC3941, which we used previously, had bizarre issues with latency that it would impart upon any traffic that crossed it; it also had an unacceptably high rate of random (cable) signal drop-outs and packet loss. We went through two of those before switching to the SMC. I was told at one point that they'd be getting some newer Netgear routers, but I'm not very willing to beta test anything for them.
Are there any good arguments to suggest that these cable/telco mega-mergers are anything but consumer-hostile? I have yet to find one.
From what I've read, the IoT stuff tends to be mainly an issue with default/hard coded credentials more than any issues with the underlying software. The end result is, depressingly, the same.
But, it's a direct admission that they were basically gouging for want of competition.
Good catch, I hadn't re-read it since it was first posted. Thanks!
While I don't know of any incidents within the US, I have seen at least one case where a drone collision actually occurred:
http://avherald.com/h?article=...
Call me crazy, but isn't the more logical solution to the issue of battery life to make the phone consume less power? I know people want their apps, but I'm not convinced that people want to carry around a laptop battery in their pocket. Also, my smartphone usually lasts about 6-7 days on a charge as-is, primarily because it does very little aside from phone stuff. Turn on the WiFi and that drops to around 3 days in normal use (simply being connected, without actually using the internet).
Sometimes it's convenient to be a fluid. Not always, but sometimes.
Worse, I'm pretty sure the AC's just regurgitating something they found on the ground...
Back then, things probably seemed novel and exciting to a broader array of people. These days, I get the feeling that the people who were once excited about those things (myself included) now, often, see them as little more than faster and more complicated versions of the things they replaced. Another issue is people's preconceived notions about how the latest and greatest tech ought to be; I suspect that they feel let down by the slow progress towards the things they believe to be the way of the future. If these things are true, it's not surprising that people in all circles would tend to have a pessimistic outlook on technology.
The rate at which they're throwing out these major updates reminds me of the Windows NT 4.0 service packs... two within the first year (1996), one each in '97 and '98, and then 3 in '99...
As that mess was largely before my time, does anyone have any recollection of the breakdown of good vs harm caused by deployment of the older NT 4.0 service packs? Is it even a good analogy?
On day 91, Clarke logged back into his account to find that it was still active. Unbeknownst to him, logging back in simply to check would reset the clock back to zero. "Yahoo confirmed via email yesterday if you access your account it resets the timer," he told me. "So, if you login to ensure your account has been deleted and it hasn't, you have to wait at least another 90 days."
So, if an account was being accessed by way of cookie (or other) exploits, I assume this would also reset the timer?
Have there been any notable cases in which it was ruled, in any court system, that Uber drivers were NOT employees and/or entitled to some level of benefits?
That this discusssion will get Godwin'd very quickly
Does suggesting this count?
Contextually relevant usernames are always in style :3
Glad to see this is paying off for them. This was covered here just over 9 years ago: https://slashdot.org/story/07/...
You missed the part where they don't actually let you turn it off.
Laserjet 4 Plus here. I rebuilt the upper feed roller and replaced the laser scanning module after the laser diode cooked a few years ago. 25000 pages so far. Manufactured in May 1994.