If they didn't state it that way, then someone could hold down 3 16 hr./week jobs that paid $27/hr., work 10 extra hours at each and effectively subvert the overtime rule.
The immediate benefit to legacy clients is that they can be partitioned into 3 channel groups, each with its own full allotment of 802.11g bandwidth.
You would only need a hardware upgrade if you wanted each client to be able to make use of multiple channels simultaneously and reach that 50Mbps throughput figure quoted in the article.
Otherwise it's a solution for reducing bandwidth contention in heavily trafficked networks (and protecting 802.11g users from bandwidth degradation by 802.11b clients, as mentioned).
I had no idea Gb Ethernet switches had dropped so much in price. If I was buying a new switch today I'd definitely be buying one of those $100 Linksys switches. Considering the cost is so cheap why even bother with 100MB if you think you'll be using bandwidth hungry apps?
The caveat here as I might have hinted in my question is that you might get what you pay for. To the point, the Linksys EG008W workgroup gigabit switch won't do jumbo frames and between two 64-bit/66MHz gigabit XP servers (one with an Intel server NIC, the other on-board Tyan gigabit) I can only muster about 100Mbps sustained between the two! (with standard 1500 MTU Ethernet anyway).
Not to be a weenie but can you quantify the improvement you saw?
My experience with early gigabit-equipped motherboards was underwhelming. Granted at the time the available options were mostly 33MHz PCI motherboards that seemed to shoe-horn in a Broadcom gigabit chipset as an afterthought, but doing the math, it should have been obvious these on-board NIC's were going to be bus bandwidth-limited and underperform a NIC sitting on a beefier (66MHz, 64-bit) PCI bus. Sure enough my experiments bore this out.
I think you're only considering a flow involving two machines. I'm looking at multiple high bandwidth streams on the same segment. And though today's DVD-quality video isn't a problem over 100BaseT, I want max headroom - think serving a couple high-definition video streams.
To your point about disk though, my two main servers are 64-bit/66MHz PCI boxes with striped arrays that seem have enough sustained throughput to saturate a gigabit link with just a few clients' worth of video demand. And at this point my equipment is virtually antiquated - I suspect the striped WD Raptor SATA arrays that are typical of many power users' servers could do the job just as easily.
He was very easy to track down. Apparently, a red flag gets raised at Google whenever anyone actually clicks on those ads. So, they eliminated the guy who needed ink jet cartridges and sent the police in.
It was the 1.13GHz Coppermine, not Tualatin, that were faulty on introduction. The Tualatin.13u core's introduction followed the 1.13GHz Coppermine debacle nearly immediately.
This is why it's darn rare to find an original 1.13GHz FCPGA chip (that will fit in a Coppermine era motherboard without a converter.
Everyone also knows that the average human only ever uses 10% of her potential brain power, with the exceptional folks only using a few percentage points more.
This statistic has always bothered me.
How reliable could an estimate of human potential be, if calculated by humans who by their own admission are utilizing only 10% of their brain potential?
This is a nice hack but I fail to see the evidence of "abuse" on the part of either manufacturer. Maybe the more expensive brand has a better warranty that the parts costs subsidize? Maybe the cartridges are nearly the same form factor but one brand goes through a more rigorous quality assurance process?
The lack of compatibility certainly gnaws at the engineers in us but it's hasty to assume that the cost to make them compatible would have been zero, especially when you take into account intangibles such as warranty, service, support, etc. Maybe it's just MuVo 2 (4GB compact flash)-type opportunism but the article doesn't bear that out on its own. More research is due before simply calling it "abuse".
This is immensely bad news for Metallica cover bands - if you're any good, you're doubly fucked because you're going to get sued by Lars and James and then your fans are going to get sued by the RIAA.
Slashdot Mirror
I have a copy of "CATALOG: The Commodore 1541's Greatest Hits" sitting here. Tracks include:
- Drive - The Cars
- Step By Step - New Kids On The Block
- You Spin Me Round - Dead Or Alive
- Crash - Dave Matthews Band
I can't make out much of the label after that, can anyone help me here?I'd like to see a few hundred comments about the issue; who's with me?
did this happen to be a certain pizza place on el camino in sunnyvale?
Nevermind, I don't know what I'm talking about, I'm about to be outsourced.
If they didn't state it that way, then someone could hold down 3 16 hr./week jobs that paid $27/hr., work 10 extra hours at each and effectively subvert the overtime rule.
You would only need a hardware upgrade if you wanted each client to be able to make use of multiple channels simultaneously and reach that 50Mbps throughput figure quoted in the article.
Otherwise it's a solution for reducing bandwidth contention in heavily trafficked networks (and protecting 802.11g users from bandwidth degradation by 802.11b clients, as mentioned).
My experience with early gigabit-equipped motherboards was underwhelming. Granted at the time the available options were mostly 33MHz PCI motherboards that seemed to shoe-horn in a Broadcom gigabit chipset as an afterthought, but doing the math, it should have been obvious these on-board NIC's were going to be bus bandwidth-limited and underperform a NIC sitting on a beefier (66MHz, 64-bit) PCI bus. Sure enough my experiments bore this out.
To your point about disk though, my two main servers are 64-bit/66MHz PCI boxes with striped arrays that seem have enough sustained throughput to saturate a gigabit link with just a few clients' worth of video demand. And at this point my equipment is virtually antiquated - I suspect the striped WD Raptor SATA arrays that are typical of many power users' servers could do the job just as easily.
He was very easy to track down. Apparently, a red flag gets raised at Google whenever anyone actually clicks on those ads. So, they eliminated the guy who needed ink jet cartridges and sent the police in.
Busting a nut to Nocona is like jerking off to Juli Ashton when you could have had Jenna Jameson if you'd waited another couple scenes.
How would you feel about doing that for every THG article? What's your KarmaPal address?
This is why it's darn rare to find an original 1.13GHz FCPGA chip (that will fit in a Coppermine era motherboard without a converter.
There's a word for that: Macromediocrity
How reliable could an estimate of human potential be, if calculated by humans who by their own admission are utilizing only 10% of their brain potential?
(Answer: 10% give or take)
Where do you want to "go" today?
The lack of compatibility certainly gnaws at the engineers in us but it's hasty to assume that the cost to make them compatible would have been zero, especially when you take into account intangibles such as warranty, service, support, etc. Maybe it's just MuVo 2 (4GB compact flash)-type opportunism but the article doesn't bear that out on its own. More research is due before simply calling it "abuse".
Seriously though, try all recipes if you want something a little less generic.
This is immensely bad news for Metallica cover bands - if you're any good, you're doubly fucked because you're going to get sued by Lars and James and then your fans are going to get sued by the RIAA.