Pros: Don't pay attention to the other reviewer who couldn't get it installed, but DO pay attention to the specs which clearly state compatibility with the XEON line of processors, and clearly DO NOT state compatibility with the i7 / Core2 / other consumer grade CPUs. The fan included classifies as quiet even at full speed. With the included resistor cables, it's hard to hear it even once attached to the heatsink. Where normally that extra turbulence would cause extra noise, the rubber separators provide some key separation space and the interesting fin design seems to really help. 12cm solutions are too big for most cases and smaller solutions would require loud fans. The 92mm Noctua is exactly the right size for a server build that's going into a workstation chassis. (The Antec P193 for example cannot fit a 120mm tower cooler, but 92mm is perfect.) Since you will likely only use one resistor cable, the extra one is really useful for the other fans in your case!

Cons: The rubber strips that go between the fan and the metal fins of the heatsink can be a bit difficult to work with. You might not want to attach them to the fan itself, in case you want to repurpose the fan elsewhere. I'm sure you could pull them off if they get in the way. Really there aren't any cons to this heatsink. The higher-resistance cable can, for some fans (unfortunatly, for me this happened to a Noctua fan) eventually cause the fan to stop spinning. You will want to monitor fan speed if you use the higher-resistance "super-quiet" cable. (I forget which color it is, but I think it's the blue.) It might work great for fans that normally apply more power than the Noctuas. Mine stopped spinning after about a year, but I just replaced the blue cable with a black one and it was back in business.

Overall Review: On Newegg you have to watch carefully for reviewers who bought a square peg for their round hole, and vocally revile the manufacturer because it doesn't fit. Xeon motherboards differ from gamer / consumer motherboards in that they have the metal bracket built in, and it cannot be removed. This makes it very difficult to find a decent heatsink for a dual-CPU workstation build, as they tend to be built for 1U / 2U rackmount chassis, and make a noise like a vacuum cleaner. This heatsink is almost a one-of-a-kind solution, practically the only product on the market that allows for construction of a dual-CPU Xeon build that you would actually want to sit next to. For a company like Noctua who is so good at quiet coolers to consider this very small market and give us access to top quality cooling, earns them my respect. Until I had these in place, I had resorted to a mishmash of machine screws, nylon spacers and a bent socket-775 bracket to get quiet Xeon cooling.

Pros: Functions if you do not plug in the power connection.

Cons: In response to Mark T.'s review: I can verify this behavior, and I know why it happens after investigating the issue. It appears my review never showed up, likely because I purchased from Fry's. The electrical anomalies that cause the power LED to turn on etc. happen because the 4-pin power plug is electrically connected to the USB 2.0 port's 5v and ground pins. When the system is powered off, USB ports give standby power for wake-on-USB, and this product backfeeds that standby voltage into the entire system's 5v rail. Even more worrisome is that something somewhere acts as a diode when the 5v rail has power but the 12v one does not. So both the 5v and 12v rails get about 5v of power. This undervolts but slightly powers everything in the system, from the USB port header. My motherboard fluctuates, pulling between 700mA and 1000mA from the USB header, which is enough to confuse my DVD drive and cause its motors to occasionally start, stutter and stop. Needless to say, this is EXTREMELY BAD behavior, and recommendations to not use this product are well founded. Prolonged use is likely to cause damage to the hardware in your system.

Overall Review: I also contacted Raidmax and also received no reply.

Pros: Fast, while it worked.

Cons: Suddenly died for no apparent reason, just after 6 months. No hiccups or odd behavior, no blue screens or taking a long time to read a file. Worked fine yesterday, dead today. Investigation reveals the BIOS can tell something is present and tries for 60 seconds or so to read the drive, then gives up and the SATA port it was plugged into reads no device present. Plugged into an external SATA enclosure, the light simply stays red.

Overall Review: While the system was running, I spot checked temps around the SSD under various load conditions and it was always cool to the touch. I also had the pagefile remapped to spinning disks, to limit SSD writes. PSU in the system is a Corsair running at maybe 60% its spec, tops. Seems to be in good shape, as are all other components. Interestingly enough, since then the word "Enhanced" has appeared on the sticker, and the price has doubled from what's on my invoice. The manufacturer model number (MKNSSDCL60GB-DX) and the Newegg item number (N82E16820226151) have not changed, which is very odd. Industry standard is to add a revision number or at very least change the manufacturer part number if you make a change to the device, especially one that warrants doubling the price tag. This might mean the product has been updated, but the specs and the brand name of the controller have not changed, so I think it hasn't.

Pros: Unified wireless. On/Off switch on the base. Internal storage for extra Unified dongle means you can grab it from your desktop and and have a mouse to use on your laptop on the go. Side buttons and mouse wheel tilt actions are configurable. Other buttons are not. Scroll wheel has nice action. There's a slight detent when it turns. (I hate when it spins freely with no action that you can feel.) Buttons have nice action.

Cons: Tracking. It's absolutely terrible at tracking surfaces. I have three other (much older) Logitech laser mice to compare to. The other mice work on every surface I throw at them except for a black wood surface that has very rough wood grain. The M510 only works on one surface I've found so far, and that is a brown faux-leather cover for a pad of paper. Surfaces the M510 cannot track on, but the V450, S520, and three other corded laser and optical mice track perfectly: * The fabric arm of the couch. * The pleather arm of another couch. * The very slightly glossy granite table, only -barely- reflective, with lots of fine matte stone patterns underneath a very thin very subtle gloss. It CAN track on: * The thick, uncomfortable binder cover. * Pant leg. * A piece of paper. (But the paper moves. And if I have to duct-tape paper to my table to use my mouse, something's wrong!) The biggest issue is the table, since the whole point of having the more expensive laser technology is to be able to use the mouse on the table without having to confine yourself to a mouse pad or other specific surface. This is not the driver issue reported by some others. I searched after buying this mouse, and lots of people are reporting the same kind of problems. The M510's laser optics look different from all the other laser mice, so it looks like simply a bad design that looked good on paper but doesn't work in real life. Also: Side buttons in games. If I assign them an in-game action, the settings screen reads the mouse event just fine and assigns it something that looks reasonable. But when I try to use the button in game, only one out of 10 tries actually causes the action to happen. For example, in Deus Ex: Human Revolution, if I assign one of the side buttons to "throw" and the other to "enter cover", I can press 'throw' seven times on average before he actually throws something. The same with 'cover'. Assigned to the middle mouse button or a keyboard key, the action happens immediately 100% of the time, on the first press. Also: Resolution. If I turn the sensitivity up enough that my usual 2-inch range of mouse motion covers more than half the screen area, the mouse starts to skip over some pixels two at a time. I hate mouse acceleration, but the only way to be able to reliably be able to place the mouse at an exact pixel (e.g. for Photoshop) is to have acceleration enabled. Also: Not a big fan of the texture of the rubber. It feels tacky, like it's already (after three weeks or so) starting to absorb oils.

Overall Review: The Unifying drivers seem to be able to configure other Logitech mice, even when they are not 'Unifying' mice. For example, the V450 shows up and for some reason has the middle mouse button configured to 'document flip' instead of middle mouse click. (Naturally, this can be set back, it just seemed interesting.)

Pros: Fan is reasonably quiet, though not the best. It's about as audible as your typical laptop fan on full blast. I'm sure it's fine on speed control inside a case though. (I was only able to test it full-tilt using a test power supply with no voltage control.)

Cons: Mounting hardware is nonfunctional. Didn't even make it to install. It has adapter brackets to convert to socket 1155, right? Well three of the four mounting holes in the heatsink block weren't bored correctly, and stripped the screws on the way in. As I was testing the last good screw on the last untested hole in the heatsink block, mild pressure trying to turn the screw also bored out the phillips head. The screw was still only half of the way in, and had to be removed with pliers. I can't attach the adapter brackets, so I can't install the heatsink.

Overall Review: I've put together plenty of stuff over the years. Positive I had the correct bits in the correct places, and I was careful that the screw was aligned correctly. They all started out fine, and then started stripping their threads 1/3 of the way in. One screw went in fine, and three identical screws in the other holes ripped their threads.

Pros: I have five of these in RAID. The only Pro is, I will soon be rid of them. Before anyone jumps the gun on the RAID thing, it was not the initial intended use and I do know what I'm doing. I put them in RAID because otherwise they'd sit in a box. My RAID controller is an Asus PIKE module with LSI chipset. Board architecture is Intel 5520 chipset with dual 5500 series Xeons. The LSI RAID controller has been configured to allow for very long timeouts. It's set to 150 seconds for device, array, actually for all timeout settings right now. When the problems started showing up I just cranked it hoping the drives would deal. The drives have been in service for about three months, mostly idling. They do occasional backups of a server and not much else. At least, until the backup was needed and I had to copy 2 TB's of the data.

Cons: These drives fail, and fail hard. They can't handle the file copy. Running idle, not doing anything, they stay stable. Copying data off the drives for more than a minute at a time, they park the heads and pause for anything from 0.25 to 8 seconds. During that time, the file copy halts. It's not an occasional but a constant thing. The system will copy for 20 seconds just fine, then go into a fury of drive head parking for two minutes. (NOT seek noise here.) In between the clicks there's a little bit of seek noise where the drives actually do their job, about 1/4 sec of seek between head parks. Basically the drives can't deal with long sessions of reads. The array has dropped six times today. (With 150 second timeout!) Thankfully the controller can re-populate the arrays without a reboot, but still I have six files buried in a 2TB production folder structure that I now have to write script to locate.

Overall Review: I'm going to belabor the details of my setup, because if I don't, for all anyone knows I have placed the drives on top of a wood-burning stove in a magnet factory. Rest assured the problems here are coming from Seagate's terrible drives. The drives are actively cooled with two dedicated 12cm fans that pull directly from relatively chilly San Fran coastline weather. The chassis is the Antec P193. The system is more than adequately powered, given a 1,000 watt supply for its measured outlet power draw of 220w under moderate load. (very hard to put this system under more than moderate load.) The OS is running on a 15,000rpm SAS drive on the same controller, so it is not a controller drivers issue. When the RAID arrays vanish, the OS stays running. What's interesting is as I've been writing, the copy moved to data at the beginning of the disks. Seems these drives get less stable the further in you go. My guess: the positive reviewers haven't filled them up yet.