That is a bit more reasonable reply.Originally Posted by alpha754293
As for the onboard video:
You were a bit unclear. Regardless, the onboard video and CPU would be on a different bus/multiplierAlso, this might be a bit of a silly question - but does anybody know if anybody's going to be coming out with a LGA2011 board that has onboard video that can also OC?
The case is closed, but there is no side fan.
Smile
Now you're just baiting me, aren't you?That is a bit more reasonable reply.
Last edited by alpha754293; 03-27-2012 at 02:44 PM.
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
No, of course not. Its just that for once you replied without talking to me like I was a total idiot.Now you're just baiting me, aren't you?
Smile
Speaking about airflows - ideally, your case should be neutral pressure. (i.e. air is going in as fast as it is going out).
Surprisingly, there's only one chassis series from Supermicro that even have a bank of 120 mm fans. And two of them still supports Ultra320 SCSI.
http://www.supermicro.com/products/c...C942i-R760.cfm
Maybe I can mount the rad behind that?
But otherwise, it should have plenty of space for the Noctua D14.
Also, apparently, some of those 120x120x38 mm fans can push 120 CFM at 46.5 dB(A) at 3150 rpm. (I'm sure that the PWM will bring that down), but...that's quite a bit of air.
http://elcodis.com/parts/453100/KD1212PMBX-6_A_F.html
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
Yes, it seems you can mount the rad behind that. Of course, it will sound like a jet taking off...Speaking about airflows - ideally, your case should be neutral pressure. (i.e. air is going in as fast as it is going out).
Surprisingly, there's only one chassis series from Supermicro that even have a bank of 120 mm fans. And two of them still supports Ultra320 SCSI.
http://www.supermicro.com/products/c...C942i-R760.cfm
Maybe I can mount the rad behind that?
But otherwise, it should have plenty of space for the Noctua D14.
Also, apparently, some of those 120x120x38 mm fans can push 120 CFM at 46.5 dB(A) at 3150 rpm. (I'm sure that the PWM will bring that down), but...that's quite a bit of air.
http://elcodis.com/parts/453100/KD1212PMBX-6_A_F.html
Also, in your SSD thread it looks as if the RevoDrive is bottlenecked by the Gen 1 slot (~1GB/s) and I can't really explain the result for the Vertex 3...and the BSOD issues/dropping out could have been because of faulty NAND/incompatibility with your motherboard.
Have you thought at all about running normal SATA drives in something like RAID 5?
Last edited by BeepBeep2; 03-27-2012 at 05:37 PM.
Smile
46.5 dB(A) isn't that bad. I think that the highest I've hit is actually around 90 dB(A) (x6) (SunFire X4200). THAT was loud. Upside, CPUs had passive heatsinks and it was pushing 600 CFM.
The Vertex 3 was on a SATA 3 Gbps (or SAS 3 Gbps - tested it both ways).
It was such a shame when the RevoDrive kept dropped offline that OCZ issued an RMA for it.
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
Whoops, I was editing my post and didn't realize you had already replied.
Most of us here are sitting a foot or two away from our desktop in our homes so noise vs performance means a big deal and hence why I linked all the fan tests from Martin's Liquid Lab.
~45 dB(A) is about as loud as anyone here will possibly tolerate, I've got a 3800 RPM Delta that is rated at about 50 dB(A) for when I used to overclock/benchmark on air cooling
Now I'm down to 1850 RPM Nidec Servo Gentle Typhoons for normal use. They are much more tolerable to me than any other fan I own.
Now, if you want to go back to ATX Desktop Mid Tower cases I would recommend this one from Lian Li:
http://www.newegg.com/Product/Produc...82E16811112304
...use something like a Noctua NH-D14 or Thermalright Silver Arrow, and swap the fans out with 5 of these:
http://www.frozencpu.com/products/12...12B7AP-29.html
Smile
My workstation right now has two Dynatrons on it so that's about 50 dB(A) as well. If a system was going to be much louder than that, then they get moved into the spare room (a.k.a. the server room). Hence why the headless state.Whoops, I was editing my post and didn't realize you had already replied.
Most of us here are sitting a foot or two away from our desktop in our homes so noise vs performance means a big deal and hence why I linked all the fan tests from Martin's Liquid Lab.
~45 dB(A) is about as loud as anyone here will possibly tolerate, I've got a 3800 RPM Delta that is rated at about 50 dB(A) for when I used to overclock/benchmark on air cooling
Now I'm down to 1850 RPM Nidec Servo Gentle Typhoons for normal use. They are much more tolerable to me than any other fan I own.
Now, if you want to go back to ATX Desktop Mid Tower cases I would recommend this one from Lian Li:
http://www.newegg.com/Product/Produc...82E16811112304
...use something like a Noctua NH-D14 or Thermalright Silver Arrow, and swap the fans out with 5 of these:
http://www.frozencpu.com/products/12...12B7AP-29.html
It's not so bad if I have the side cover on it.
I couldn't sit/work next to the SunFire X4200 for more than 15 minutes at a time because it was so loud that it was actually giving me a headache. Literally. And that's even WITH headphones on.
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
Hitting 4.5 on a 3930 or 3960 will take roughly 1.3-1.38V, most likely as someone already mentioned no more than ~1.34. With appropriate cooling and decent airflow (high-end air, H80+ etc.), maintaining the system at safe temperatures will not be an issue.
However, 4.5 is considered a light to moderate overclock at best and I would be surprised if the CPU's life expectancy is affected much at this point. If configured properly, the system will be rock-solid stable no matter the workload. Your CPU might degrade after a year or two, thus needing a small voltage bump; but assuming your temperatures were within safe ranges and you didn't pump too high voltages through it, I wouldn't expect the CPU to degrade before the useful life span is over.
Buying the 3930K is definitely the right choice, the price difference doesn't justify the performance difference. As for the board, I haven't tried the Sabertooth but according to most opinions it appears to be very solid. Besides that, the P9X79 WS might be worth looking into as well but in terms of price, the Sabertooth is probably the best choice as others have already told you.
Thanks s4pphire.
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
I don't think that the RevoDrive was bottlenecked at all considering that it was nowhere close to the theorectical peak. I dunno. OCZ couldn't figure it out either. But that's kind of dumb if there's a NAND incompatibility with a motherboard. It's a PCIe card, and therefore; it should fit into any PCIe slot.
Are you talking about mechanical SATA drives or SSDs?
I'm currently running Intel 520 + OCZ Vertex 3 (+ 4 2.5" Fujitsu 73 GB 10krpm SAS 3 Gbps drives + a Hitachi 1 TB 7.2krpm SATA 3 Gbps).
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
The incompatibility problem would have more to do with the storage controller on the RevoDrive and how the motherboard BIOS was coded to handle storage over pcie...I'm not exactly an expert but I think that is how it goes.
I was talking about SSDs.
I would pick up four identical SSDs if I were you and run them in RAID 10 or RAID 5.
Smile
I don't know if that would really do anything for IOPs though, especially when I'm going highly randomized writes like swap.
RAID of any kind, (maybe perhaps with the exception of RAID 1 with two drives, and RAID 6 with three drives), works tremendously well with STRs but I don't think that it really make that much of a difference for random write performance. Besides, wouldn't that just add another layer of complexity to it? KISS.
flow man:
du/dt + u dot del u = - del P / rho + v vector_Laplacian u
{\partial\mathbf{u}\over\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} = -{\nabla P\over\rho} + \nu\nabla^2\mathbf{u}
Posting Permissions
Reply With Quote
Bookmarks