Bigfoot KillerNIC Review
Jeremy Starley takes an in-depth look at Bigfoot Network's Killer Network Interface Card from the perspective of an MMORPG gamer.
Hardware Reviews By Jeremy Star on December 13, 2006
 | ||||||||||||||||||
Testing Procedure - You did what with it? Here's how I did what I did, why I did it, and what I did it on. My test platforms were my very own PC, and my fiancée's PC. Here is the basic info, for those of you who like to know these things. My PC
Running games at 1920x1200 resolution. My Fiancée's PC
Running Games at whatever they will chug along on best, usually 1024x768 resolution. All tests were run using driver version 2.1.0.6. If you know anything about computers, you've figured out by now that I upgrade mine quite often, and my fiancée, well, doesn't. The purpose of using both computers as test platforms was to allow me to see how Killer influences performance on a higher end PC as well as on a lower end PC. I also tried to slip one over on my fiancée and see if she could notice the difference without me telling her I had done anything. That did not work out so well, but more on that later. As for testing, I played some of the bigger MMOs, both with and without the Killer installed, and benchmarked some of my time on my PC with FRAPS. I tried to do the same exact thing both times, so that the results would be semi-accurate. I also tested some of the games on the slower PC, and benchmarked WoW on that one as well. I played to see if I could actually notice a difference, if I could quantify it with FPS benchmarks, and if I could see a reduction in ping times. And now, a note on testing: Network lag is really unpredictable. There is almost no way to replicate the exact conditions I experienced on the first run through the second time I did a run through. There could be more or less people on the server, more or less network traffic on the internet in general, and more or less people on my cable internet hub in my town. All of these factors plus many more make ping almost completely random and therefore it is extremely difficult to test reliably for any sort of ping reductions. The only way to do something even close to reliable would be to have two PC that are exactly the same, one with the Killer, and one without. If you then connected to the same server, and ran pretty much side by side while watching the pings, you should be able to tell the difference. I don't have two duplicate PCs. Truth be told, I don't bother looking at pings while I am playing MMORPGs, because unless they are atrocious, you don't notice so much. So I did what any responsible journalist in my position would do: I made the ping section really, really short. On to the testing! Performance - Is this thing on? WoW So, the first game I played was World of Warcraft. You knew that was coming, so I don't want to hear about it. With the built in NIC card on my Nforce 4 motherboard, WoW runs like it always does. I have all the graphics goodies and such cranked up as high as they will go, and WoW chugs along pretty smoothly. Until I hit Ironforge at peak time, or go on a raid when the server is busy, I don't notice too many slow downs except for the occasional lag spike here and there. So of course, I wanted to push this and go on a raid in Molten Core at 8pm EST. My performance varied, to say the least. Wait, did the server crash? No, that was just a giant lag spike that wiped half the raid, it's ok! Ping? Well, the little bar showed green... NO! RED! No, green. Yelllllllow, RED! Oh, you kept moving, you're out. No surprises from the built in NIC, so I popped in Killer and tried again. The game was smooth as always in most areas. Ironforge at peak times still slowed down, but I thought it was maybe a tad smoother. Then I hit MC again. Holy ghost, Scooby Doo! Game play was noticeably smoother. I seemed to experience less lag spikes pausing me momentarily, and it felt faster overall. Sure, we still wiped on a big lag spike, and there were still times when it felt like I was pausing, but they seemed to come fewer and farther between. The last time I noticed a performance increase that felt this big, it was when I went from 1GB of RAM to 2. Ping was still a game of Red-light/Green-light. Non-Killer Performance
Killer Performance
|
Bummer I missed the previous comments.
My only concern is that having any software operate outside the boundaries of the windows kernel (i.e. any software talking directly to the hardware) can have disasterous results. I wonder if it will be windows "approved".
I question its usefulness given lag usually is coming from beyond your nic card as pointed out in the WoW example.
I guess if you go out and buy a physX card, you're going to get this too.
I didn't know what to expect from the article considering how much advertising I see for on this site, but as a network professional I can tell you the article pretty much fell into line with what I expected the Killer NIC to do..
And not do.
Hehe,
Now, if I feel that a game that was not made by SoE/Sigil has a lot of lag and I want to reduce it, I will know there is another option, if I am ready to bypass all the techs issues (I don't think I would have try all the slots on the motherboard myself hehe).
But...as a PvE player who value instancing and relatively high-end computers, I dunno...see, my computer is still much better than his fiancée, and I am about to change it...I don't update regularly, I change.
For example, the only places where I would have appreciate less lag would be lagville in CoV (Grandville) and some bases-missions with instanced grids, way too many grids prolly. He talk about lag in mayhems, and honestly, I didn't lag much at all in any mayhem ever. I wouldn't spent $300 for Lagville alone or 1 particuliar mission, since in CoV, I don't play in Lagville, I go to my mish, so I can endure the lag while I bump between mishs...even if it is really annoying.
If I would play WoW, well, I wouldn't do MC and if I lag in Ironforge, it suck but I would live with it...just like I was living with lag in the Bazaar.
Anyway, 1 more option to attack the evillagmonster is always good.
*Bite at the lagmonster*
I'm curious as to how it affects performance with regards to RF Online during Chip Wars? Some Chip Wars in my server have up to 500-700 people duking it out in the same map at the same time (which generally kills performance). If anyone has tested it on this particular game and has any feedback, please post. Thanks.
High end sys dont need it,low-mid end sys can use that money for better stuff.Killernic is kinda castrating themself imo
I agree there. I think that the network card is probably at the very bottom of the upgrade list on low to mid range machines. If I had 280 bucks to dump in to my system, I think I could get more of a performance increase from increasing my RAM or a new video card. I also have to say, as a network engineer, even with an on-board NIC on a cheap motherboard, it probably isn't having a noticeably adverse affect on your actual latency or throughput.
What I see this card doing is freeing up other resources for use in other places, It just doesn't make any sense to me to go this route, rather than just upgrade and increase those resources at the source.
When I read this review I hear 2 things. Reduce CPU needs and give game trafic priority.
One thing pops in mind when reading this story. If the network trafic is screwed because the CPU can't handle it, why not buy a second CPU (Dual core CPU). I am really curious if this NIC still has an advantage with a dual-core system.
Another thing is the trafic priority. This can also be done with a good router and/or software running on the PC managing the trafic. With a dual-core system the CPU load does not matter.
Same goes with the teamspeak software. Using a dual-core will solve the CPU problem.
Im in the "sure it can smooth things a bit on a lower spec machine but there are a bazzillion things that will give more bang for your buck" school of thought.
Internet latency and throughput is not going to be solved by a NIC and some of the slow down issues the author mentions are clearly server latency problems. Buy another stick or memory upgrade the graphics card or upgrade the motherboard and cpu and see noticeable performance increases across the board is my advice!
Cheers.
I'm still a spetic on the issue, but I think you figured there would still be those spetic's still around on the issue of this card. (Wasn't there some software, I think it was called "Gear" that did something similar to this NIC card that many FPS's and MMO have banned people from servers for?
I'm still using an onboard, non-nForce network card and was considering a Killer NIC but now I'll probably spend the cash on a new board and I won't have to worry about the extra slot, card not booting etc. It would mean upgrading all my system components though - so not as 'cheap' as the Killer NIC option but certainly more future-proof.
I'd like to see a review of the top end nForce vs. Killer NIC vs. the best other PCI network card out there.
Great review though
I don't know how much it costs them to manufacture the card, but I do know that the only other NICs on the market with built in network processing units cost upwards of $800 US, and they aren't really meant for consumers. Also, it's new technology, so the retail price is bound to drop as the cost to manufacture them goes down. Remember earlier this year when the new NVidia 7900 video cards cost over $500? Now you can pick up a new 7950 for less than $300. It's not just because the new 8800 cards are out, it's because the cost of manufacturing the chips for the 7900 series dropped dramatically.
Traffic prioritization can be done at your router (or via software on the system)
Offloading CPU only means you need a faster CPU which is probably cheaper to just get an upgraded CPU for your system.
All in all for power-tweakers this might be an option but I don't see it as something your average gamer is going to want to sink money on. Bleeding edge systems are going to see VERY little, if any, benefit from this sucker and, honestly, anything else you'll get a lot more bang for your buck by upgrading ram, processor, mobo, audio or video (or all of the above).
Good review, though. It's nice to see a well written and thought out review on this site. I would love to see more reviews done in this manner.
it's misleading to state this to people who don't know what hardware is and what it does. you say your self that you aren't a hardware site, and one of the first comments on this article even points out 'that this is the first positive review of the Killer NIC i've ever seen'.
then we have people chiming in with crap like:
This is a great article. It relates to the MMOG gameplay and is not tech-infested with charts. I had a fun time reading this article and I hope more like this will come!
god forbid we have charts. those might show how much something like this is a huge waste of money.
and:
I'm still a spetic on the issue, but I think you figured there would still be those spetic's still around on the issue of this card.
spetic?
bottom line:
don't waste your money on this crap. it'll give you as much of an increase in performance as a new exterior case (i.e. ZERO).
a faster connection will help lag, not a swanky new NIC with a big stupid K on the chipset.
i hope anyone actually reading this review does more research before they plunk down hard earned dollars for this thing.
[H]ardOCP says:
I see a product such as this carving out a solid niche in the high-dollar boutique gaming boxes that we see sold from companies like Velocity Micro, Maingear, Falcon Northwest, and Dell's XPS class machines. When you are buying a $3000+ computer, I see many opting for the fancy NIC that they likely know nothing about to begin with.
http://enthusiast.hardocp.com/article.html?art=MTE1MywyLCxoZW50aHVzaWFzdA==
Anandtech says:
Our current opinion is, without FNapps, improved performance across a wider variety of titles, and a significantly lower price tag, this card is destined to be nothing more than an interesting footnote in the annals of hardware history.
http://www.anandtech.com/showdoc.aspx?i=2865&p=11
these are HARDWARE REVIEW SITES. not fluff guys that put it in their box and look around orgimmar for a sec.
First of all, i enjoyed reading the test. Very good writing, Jeremy.
As for it's usefullness, i suggest you first check your provider and his routings. My provider offers a fast ping routing for an additional monthly fee. This instantly changed my ping from ~200ms to 60-80ms.
I like the concept of the card with the ability to provide more specialized applications (like Ventrillo/Teamspeak) possibly even USB keyboards/controlles to free up more CPU resources. It might all be small improvements but it could be quite good if used to it's full potential.
Thanks for the interesting review.
It's interesting that you choose to quote from a preview article at [H]ardOCP, but ignore the fact that Killer NIC won an Editors Choice Silver award from them in the December 08, 2006 review.
http://enthusiast.hardocp.com/article.html?art=MTIzOSwxLCxoZW50aHVzaWFzdA==
Anandtech also notes that while they were unimpressed overall, the Killer NIC does do what it says: i.e. it lowers ping and increases framerates.
As for us not being a hardware site: it's true. I admit it. I review hardware from the Every-Joe perspective, so that the people who play MMORPGs can read a review and figure out if it is right for them and their game without trying to decypher a bunch of techno-babble. Does this mean I don't know what I am talking about and I am not qualified to techno-babble? NO. Truth is, my day job is at a pharmaceutical automation company supporting computers, robots, and bleeding edge technology. Before I worked at my current job, I worked for IBM supervising the inspection of chip carrier products for customers such as ATI, Nintendo, and Tivo. If I'm not qualified to review hardware, I don't know who would be.
You may want to actually read articles instead of cherry-picking comments that you think will support your point of view. Both sites you quote from confirm that the Killer NIC does indeed improve FPS. Have fun with your cute little blender-toast analogies over there in blissfull ignorance land.
I read the reviews when this crap first came out, and then just went back to the sites and punched Killer Nic into their search fields. The results I found weren't flattering, and I linked them. You'll notice that the results I linked weren't even BAD, per se... they just weren't good.
God forbid you confuse your readers with techno-babble. We wouldn't want to confuse their simple little minds. Insulting the intelligence your reader base is a surefire way to get new fans. Go ahead and recommend a $300 LAN card to them. I'm sure they'll appreciate it when they could have spent that money on a better video card or more RAM.
So many things contribute to lag, but to say the "killer nic" is for an 'everyday joe' that can't decipher a simple benchmark chart is stupid. A fool and his money are soon parted, I suppose.
Your title of "chip carrier product inspection supervisor" is certainly impressive. It's right up there with 'middle management at a software firm'... a.k.a. 'can't code himself, but took a business class, and knows how to organize shifts'. Congratulations on that. It must have gone smashingly well since you've now been promoted to 'everyday joe hardware reviewer for an MMO website'. I've worked in IT for 6 years as the IT Manager for a PDF conversion firm in the california bay area. We regularly slosh large amounts of data back and forth across the LAN on a daily basis. I'm not trying to throw that around like it means anything fantastic, I'm just saying that I don't "supervise the inspection of chip carrier products". I build boxes and put them out on the floor.
The Killer NIC has a place. It might be on a local counterstrike server in a gaming cafe, or a high end workstation at a video editing lab. To say an 'everyday joe' will get increased framerate in molten core is laughable, considering most motherboards these days come with onboard gigabit jacks, and the average DSL or cable line uses only a very small fraction of that pipe.
Good luck out there. You'll notice I registered just to post on this ridiculous article. Feel free to delete this post and ban this account, I won't be coming back.
I'm sorry, did someone kill your dog or something?
For those of you without serious attitude problems and superiority complexes, you will no doubt be intelligent enough to know that in order to supervise the inspection of any sort of micro-chip product, you must first know how to inspect it so that you can instruct your employees to watch for certain defects in certain chips, etc. Again, you cherry pick your commentary and choose to launch an attack on one aspect of my reply, conveniently ignoring the fact that I work in tech support.
You will also, no doubt, be aware that I would not insult the majority of our reader base by assuming that they could not decypher techno-babble. We merely assume that the average MMORPG.com reader would rather not, since they are most likely here to learn about MMORPGs and how/ with what to improve them. Were we stupid enough to go around insulting the reader base, we would not long survive in this market.
--"To say an 'everyday joe' will get increased framerate in molten core is laughable, considering most motherboards these days come with onboard gigabit jacks, and the average DSL or cable line uses only a very small fraction of that pipe."--
Have you read any of the reviews, by anybody? If you have, you clearly do not comprehend what is being said. YES, the Killer NIC will increase the FPS of most PCs or smooth out performance on most PCs. It has nothing to do with bandwidth. The question is not will it increase your FPS, it is do you want to spend that kind of money on it, and I fail to read anywhere where we or anyone else told people to go buy one. Worth is a personal opinion, and we invite all readers to make their own decision. We merely state that yes, it does do what it says.
Have a nice day.