Same amount of work done per iteration, more iterations per second... overclock, overrev If your motor starts performing less work per iteration the higher it goes, then that's a different problem
Of course I checked it Here's the bad boy... And the dual SSDs are going to be setup on a Raid0 configuration to host only applications and the OS. The 4.5TB of other hard drive space will handle big movie storage
Should verify but my recent ASUS based build has 64GB of RAM, which is totally nuts, so far I've never used more than 4% of it. With it, the i7 and the SSD system drive it's ready to log into in a few seconds. Overclocking these days is almost expected. Bios config abilities (and even autodetecting) are nuts flexible. But if it isn't being used as something like a gaming rig, there's really no point. One of my servers is dual purposed as a media server and a Hyper-V host with six guests running all the time. It's just a quad core and the box never even breaks a sweat.
Unless you're running server type applications like virtualization or SQL DBs, then you're wasting money on getting that much RAM. Your system won't be any faster. The speed of the RAM is the only thing that will matter once you get past 8 or so gigs of RAM.
Why would you even need to overclock these days? Or is the next project going to be adding an interstellar research facility to the brewery? Or possibly just trying to up the rankings for his Seti team? And over 4 TB storage just screams massive porn collection.
So let me get this straight...you're going to replace every piece of equipment you own except the one likely causing the stuttering and buffering...to fix the stuttering and buffering.... Yep...that's a Lever solution to the problem.
Actually... "Luckily for us, Sandy Bridge-E made some other changes that affect overclocking. While normal Sandy Bridge platforms use a bclk of 100MHz, the Sandy Bridge Extreme can use a bclk of 100MHz, 125MHz, 166MHz, or 250MHz. Just crunching the numbers, if that 4x the bclk ratio is in effect, then with the multiplier at 43, as it was for the single core turbo overclock with the 100MHz bclk, we are looking at a potential overclock of 5.375GHz with a bclk of 125MHz, 7.138GHz at a 166MHz bclk, and 10.75GHz at a 250MHz bclk. Obviously, pushing the bclk to 250MHz, or even 166MHz, is really out of the question. Even using the base clock as the multiplier at 3.6GHz, you're trying for a 5.9GHz overclock with the 166MHz bclk. But 125MHz bclk is right up our alley."
I wonder if I should mention that there's no need for a server to have a fairly good video card. After all who spends a lot of time gaming and watching video on the "server". I can only think of a couple of reason to drop a good video card in a machine and they all have to do with RemoteFX and virtual desktops.
Tesla's or K10/K20. GPUGP... I just got done with a contract where I had ten rack mounts with dual-tesla's (M2075's) at my disposal. Those things could do some amazing computing. -jim
Ahh...best I've got in the lab is a stack of Arris VIPre multi-format transcoders that are using Jet engines to do the job based on noise they make.
These were the toys: http://www.telestream.net/pdfs/datasheets/dat-Vantage-Lightspeed-Server.pdf And yeah, they were loud. -jim