Upgrade to Killer Wireless-N WiFi Card?

I currently have an Intel 5300 wifi card, is it worth the upgrade to a killer wireless-n 1102 card?

Will I feel a step up in performance/range of the card? Anybody using a killer wireless card?

 

I would read this article: AnandTech - Bigfoot?s Killer-N 1102 Wireless Networking vs. the World

Basically, if all yuo want is range, i don't see a reason to upgrade. If you want better ping, then by all means go ahead.

 

I've had several issues with this AnandTech review- I'm too lazy to read that again and look for them but for starters throughput for several cards is way too high i.e. Netgear WNR3500L is not capable of 170-190mbps real-life throughput on Wi-Fi. If it was no one would ever complain about Wi-Fi as it would be twice as fast as 100mbps Ethernet which it isn't.

Second- latency on LAN is almost always 1ms. I have no idea how did they come up with 4-6ms (or more) on other cards. As I said I haven't bothered reading it thoroughly but the only way to get this kind of score is to do some heavy downloading at the same time.
Also remember that most of the latency comes from your Internet connection- you may have the fastest Wi-Fi card in the world but BigFoot won't help you at all once you're out of the LAN.

So assuming your current Internet ping with 5300 is 12ms and when you ping your router you get 4ms with 5300 and 1ms with 1102 your overall Internet ping with 1102 will drop to 9ms. Not as impressive.

To sum it up- if you're going to upgrade anyway think big and go for Big Foot 1103 but there's really no sense doing it hoping for an improvement in gaming.

Your router is much more important than the card as it does the routing and handles multiple connections that are needed while gaming. It would be much more beneficial to leave the 5300 as it is and look closely at you router rather than upgrade the Wi-Fi card and still use a crappy router (that's assuming yours is )

 

wow, very detailed information. Much more clear now, I'll stick with the 5300 then, thanks

 

recently went from the Intel 5300 to the Bigfoot N1103 and all I can say is wow!
My router is downstairs in a fairly decent sized house: my M15 would get 60% signal at an average of 90-120 Mbps on the 5300 using a Dlink 655 router.
Same location, same laptop and same router but the Killer N1103 gives me 80% signal and 300 Mbps.
In simple terms, better connection/range and three times the speed. Worth it in my opinion.

 

The DIR 655 is notorious for fluctuating signals and speeds. Your performance increase is based almost entirely on the improved signal strength and nothing more. Had you changed your router to something better, you would have seen an equal change in performance. Also, while your connection says 300, what is the actual throughput? Packet loss and fluctuations mean you are probably not getting what you think. Yes, you connected at that rate, it doesn't mean you are transferring at that rate. It coudl be that the card just has a wider tolerance than the Intel, allowing you to connect at the faster speed regardless of true connection strength.

The 655 has some good hardware, but was crippled by bad choices (power supply system) and software by D-Link. I have a long and sordid history with this device, if you want better performance, change your router, you will probably see an equal or better increase than you saw by changing your card. At least the Killer Wireless is cheap (their Nic is outrageously priced).

Oh, and remember the Killer wireless, can only do 300. It lacks the necessary 3rd ant. for 450. Personally, I wouldn't limit myself since 450 is here and moving down to lower range models.


I have to agree with Downloads here about the card.
Their claims are silly. They made similar claims about their nic, which most will tell you wasn't worth the money either. Granted if I was building a $3000 desktop, why not, but realistically, the cards will not get you what they claim unless you are simply using brute force to overcome a networking problem that could have been fixed another way.


To the OP,
Keep your 5300, it's a good card.

 

I can tell you right off the top of my head, there are some differences in how the Atheros drivers (ie: AR9280/AR9380) report speeds to the Win7 Networking subsystem, versus that of the Intel hardware (ie: 4965/5300/etc.).

You shouldn't use the Windows-reported connection speed as a benchmark between the Intel cards versus the Atheros.

 

Oh, and remember the Killer wireless, can only do 300. It lacks the necessary 3rd ant. for 450.

strange, because on the card I connected the N1103 I connected 3 antenna -- could be you are referring to the N1102 which has 2 antenna.

As for the Dlink 655 is had been problem free and connects to 5 laptops in my 3600 sq. ft. house just fine.

So, no, I don't think the increased performance is a placebo effect: the N1103 is a much better card than the Intel 5300 it replaced.

 

Original question concerned 5300 and 1102 so that's why the "doesn't do 450mbps" part was there- it was in response to the original question about 1102.

As for the performance- good to hear that you're happy with an upgrade but keep in mind that people have the right to be skeptical if someone uses values like % of signal or "bars".
Also no wireless settings were ever mentioned (Intel requires you to manually set the channel to 40MHz unlike Atheros so you might have been running @ 20MHz channel width in spite of router being set to 40MHz).

All in all it's another vote in favor of Atheros' new chips which is good for consumers. For a while we've had a choice of "come" Intels.

 

The 655 has worked well for some, for me, it worked great, when it worked. It took them 2 years to make it stable and had to remove features to do that. I have a post on here describing my experience with that router, it's quite a read. Dlink's forums used to be full of horror stories and people fed up with it until they culled 7000(!) threads from it. It still remains one of the busiest forums on their support forum. Yes, they sold a lot of them, but in my opinion and many others, they should have been forced to recall it. There were times when the uptime was measured in minutes. We were used as beta testers on a nearly 2 year old product that still wasn't ready for prime time.

I never said the 5300 to Killer was only placebo, I said the Killer was masking other issues. As for a better card, it depends on what you determine that by. The 5300 is a pretty energy efficient little card and I regularly manage 300mb on it. Is the Killer as battery efficient? Are the transfer rates really better? We have no hard numbers to prove any of it. but chances are it isn't better in all areas, so for some, one card, may be better than the other.

The original discussion as Downloads said was based on the 1102, which only has 2 Ant. and therefore is limited to 300.

 

appreciate the comments, sorry if I came off hard , didn't mean offend, just a knack I have!
I too have spent the time over on the dlink forums, so I do recognize the "issues" the router has had -- dspite that it has been robust for me, again, once I got it and the intel to talk to each other nicely (the 20/40 band thing) -- the Killer seems way easier from that standpoint and although I don't have the hard number, I will sya the actual laptop download etc. is more consistent.

One of my frustrations was never getting the 5300 to operate at 300 other than when is was next to the router or in the next room -- sadly for me in our house that is the basement and the media room, not where I actually work, the office. So although the card functioned, it never performed up to its capabilities -- the killer seems by the numbers and the actual access speed/downloads to be closer to its optimum.

OK, always willing to learn: what numbers should I be collecting/reporting to better measure throughput etc. regarding actual performance??

 

If you are only accessing the internet, your throughput isn't going to matter much. 300 is probably beyond any internet connection you have, but if by chance you do, most servers aren't really equipped to send that much data to just a single user.

If you have a home server, copy a large file over and see what kind of speed you get. The faster you go though, the less difference a few megs will make. Theoretical throughput on 300 wireless is somewhere around 40MBps (I think), but you will need a gigabit connection to the server (which can theoretically flow 125MBps), otherwise the other end will bottleneck you.

In other words, unless you have a nice server or world class internet speeds, I wouldn't worry about your throughput.