Verdict: Fails to live up to its ambitious performance claims.
The Bigfoot Networks KillerNIC M1 review caused a stir in the CPC lab. While the £180 price tag is off-putting, it became one of the more endearing products we've tested this year. The lunacy of building an entire computer on a PCI card just to provide more consistent performance in online games is a strategy we admire.
However, as with most things in life, there's a cheaper alternative available. The HBB1 Broadband Booster from Hawking Technology purports to 'eliminate Internet lag in your network' by 'boosting your Internet apps as they exit your network'. It should therefore benefit online gaming, VoIP calls and P2P file sharing. This is quite a claim for a box costing less than £50.
However, while the HBB1 and Killer NIC share an objective, they go about achieving that aim in different ways. The Killer resides inside your PC, while the HBB1 has two RJ-45 LAN ports and essentially acts as a pass-through box between the broadband modem and the router. This is fine if you have a separate modem and router (as with many cable setups), but you're screwed if you use a router with a built-in modem (as with most ADSL arrangements). With an ADSL router, you'd have to place it on one of the router outputs, and either have the HBB1 only serve one PC or buy another router to place after it in your network.
We used the HBB1 between a modem and a router, and setup was a matter of plugging in the two LAN cables, followed by the power connector. The power, router and Internet LEDs lit up green and started flashing on the first attempt. 'Boost' emits a blue light - for some reason, blue means speed in the world of PCs.
The HBB1 is a hardware QoS (Quality of Service) engine. It analyses the data flowing into it, recognises that data of a certain type is from a certain application, and determines whether or not it should be prioritised. This is necessary for time-sensitive traffic due to the 'first come, first serve' nature of the TCP/IP protocol. Network traffic (which is sent in packets) enters the router and is then sent through in the same order in which it arrives. However, as some packets are larger than others, or the buffer is fuller in this second than it was in the previous second, the rate at which packets are sent through can vary from 10ms to one second. This is fine for file sharing, but rubbish for VoIP or gaming, as you'll miss parts of your phone call or get shot by some n00b without being able to do anything about it.
The HBB1 uses the Ubicom Stream Engine, which analyses each packet as it arrives before assigning it a priority. Next, it decides if there are any time-sensitive data streams - from a game or VoIP call, for example - in the queue and if there are, it fragments lower-priority packets into smaller chunks. This makes it easier to insert the game packets between packets of other data, and ensures that these other packets won't spend so long being pushed through the buffer in the broadband modem that the process will stall game packets further back in the queue. If the HBB1 works, a steady flow of game data should flow back and forth through the device, avoiding the lag introduced by the broadband buffer and the TCP/IP protocol.
Having read the bold claims on the box boasting 'up to [a] 400 per cent' increase in Internet application speed, we began testing. As with the KillerNIC, we used FRAPS in Lord of the Rings Online, CS: Source and Battlefield 2142.
Despite careful monitoring, there was no clear increase in either pings or the frame rate. With and without HBB1 installed in our test network, Lord of the Rings Online topped out at around 72fps, CSS at 60fps and Battlefield 2142 in the mid-50s. Only once was there a slight frame rate increase with the HBB1 installed, and that was a negligible rise of 3-5fps in the highly populated areas of Lord of the Rings Online. This could easily have been due to a drop in total server population, or it may have been server-side alteration effecting the change.
In other circumstances, the HBB1 seemed to be a little more at home. With the HBB1 connected, emailing 3MB worth of JPEG attachments via Outlook while surfing the Web was boosted in runs. On average, mail was sent around eight seconds quicker than it was without the HBB1 connected.
Conversations on Ventrilo were also less patchy in prolonged tests, and when combined with online gaming in CS:S.
Sadly, there's little you can do to tweak the HBB1 to improve performance. Using a web browser, you can view firmware version, upstream rate and other basic information. The five options that appear under the Configuration tab elicit little excitement. You can toggle 'Auto Rate Detection', change your upstream data rate, connection type and router address, or enable a static IP mode. These options are handled by the booster during setup, so there's very little point to them. Game data prioritisation is automatically switched on and can't be turned off, so clearly, the HBB1 doesn't work as a game data QoS engine.
Conclusion
The HBB1 works for some apps but not gaming, so there isn't a lot about which to be excited. It handles VoIP traffic quite well, so if you're desperate to cut some of the jitters from web-based phone calls then it might be worth a look. However, with a price tag of £50, we're not convinced the HBB1 is worth buying simply for more reliable VoIP calls. Given that the only other task it accelerated was a large email, the HBB1 is possibly the most boring product we've seen.
