NIC card made for gamers

[Jammrock]

Actually, it's not really a waste at all - yeah, SATA has higher bandwith than PCI...but which consumer HDD can approach either of them?

A single HDD you are correct. But what happens when you run multiple or a run RAIDs? Then PCI gets saturated and the entire bus gets bogged down. Same thing with putting Gb on PCI. Gb can saturate the PCI bus easily. That's why both Intel and nVidia both made proprietary buses on the motherboard chipsets (pre-PCIe, though nV still does) specifically so Gb could run max speeds without saturating the PCI bus.

[Nowhere]

Well, using PCI SATA controller is/was typically and intermediate measure...I doubt many use more than one HDD this way (and anyway...two should be quite fine). Note that I was commenting only about SATA - the NIC is obviously dumb idea (but OTOH...who has HDDs to saturate it?...anyway, nevermind)

[Nowhere]

Uh-huh, some fascinating details emerge

http://hardware.slashdot.org/hardware/06/08/08/193237.shtml

[Jammrock]

Yeah, and they want $250 for it. I think not.

[rylan]

Well that explains it... the idea/company was formed by people getting their MBAs.

[Taz]

Killer is the only network card that runs an open-source version of embedded Linux and allows users to write and download their own FNapps to the card.

And the first to have a virus installed onto it to trash your machine

It's a product for people with more money than sense.

[Mehen]

Grr. At a lower price it had potential, at $250 - "I think not."

[Helevitia]

I want to see a review or tests or something...

[Nowhere]

To mock it? You can do that already

[Mehen]

hardocp article http://enthusiast.hardocp.com/articl...50aHVzaWFzdA==

"At a recent trip to the Bigfoot offices in Austin, TX, we saw the Killer NIC working and just how the team there was collecting data about its performance. In real world comparable scenarios we saw pings drop 10ms to 20ms on a broadband connection playing World of Warcraft and Counter Strike Source. We also saw frame rates positively impacted anywhere between 3 to 10%."

[az]

10 to 20 ms ping drops? I've had connections with a ping of 40 or so to the net, and nobody can tell me 20 ms of that is from my crappy NIC.

[Helevitia]

Very interesting....

Over the course of five 15-minute benchmarking sessions, the non-Killer NIC machine averaged 15.41 fps, and 46ms of latency. Minimum frames per second sank as low as 9 and reach a maximum of 23. The general experience at this level of performance borders on unplayable.

The Killer-NIC machine trounced the performance of its lacking sibling. Over the course of the same 15 minute testing blocks, the Killer NIC'ed box averaged 23.5 frames per second, a more than 65% performance gain. Minimum and maximum fps were also better, split between 12 on the low end and 35 on the high. Pings were relatively similar to the standard box, though we did notice latency spikes much less often on the Killer NIC'ed machine.

http://gear.ign.com/articles/729/729733p1.html

[Helevitia]

Another quote:

The card bumped the framerate on the bundled F.E.A.R. from 74 to 86 fps, and the ping went from an average of 67ms to 45ms.

[Mehen]

Very cool - maybe in a few years some of the features of it will make its way into more mainstream cards, or I suppose the price will drop as well. "Dedicated online gamers with a bit of cash for experimentation are encouraged to give the Killer NIC a shot."

[Wombat]

I have a very hard time believing this.

First, the machines weren't identical. That invalidates the test right there.

Second, notice that their copy of F.E.A.R. was apparently supplied by Bigfoot (notice the Killer on the CD). I automatically don't trust it.

Third, where's the standalone numbers? No single-player benching? If you can't compare single-player to networked to networked-via-KillerNIC, then it's also useless.