Jump to content
Search In
  • More options...
Find results that contain...
Find results in...


  • Content Count

  • Joined

  • Last visited

Everything posted by Egad

  1. @fabiosapplou Problem is GeForce and Quadro drivers aren't the same. You can make it work but you'd have to do some tweaking and you probably lose 10 bit color in the process. In these scenarios lowest common feature tends to win. Typically people will pair a weaker, but still decent, Quadro with a 'gaming' card and run them both on GeForce drivers. For sanity's sake I'd suggest a M4000 SLI. Or if your specific software has good multi GPU support you can unlink them and let the software push out jobs to each GPU as it sees fits.
  2. @fabiosapplou Yes, setting aside 10 bit color that's correct. Here on some benchmarks on the Maxwell higher end cards. Depending on the specific task I might recommend the EVGA Titan X Hybrid, for the additional CUDA and RAM. Otherwise a 980Ti either with a hybrid cooler or reference blower would be ideal. The aftermarket fan coolers work well for gaming, but they dump the heat in the case. On extremely long renders this can get problematic. The reference blowers hit max temp faster but take much longer to heat up the entire case and potentially cause other components to throttle.
  3. People are more likely to have some Windows only applications that place fairly intensive demands on the system (games, professional software, etc). Even if you have good VM software, why deal with the hassle when a Windows key is under 100 dollars? I enjoy CentOS, but I still mostly run it in VirtualBox on my second monitor. I render using Windows only software and do my Ruby programming in a virtualized Linux. We keep talking about moving the administrative staff to Linux because 99% of what they do only requires a decent Webkit browser. However at the end of the day, one piece of ac
  4. Microsoft also is not terribly active in revoking older MSDN keys. They've always tolerated a certain amount of piracy on the grounds piracy still means market share. I say this not to encourage such activity, but rather because lots of people have been lulled into a false sense of security by the fact Microsoft is not terribly active in cracking down on MSDN key abuse. All it takes is one change in corporate policy, one engineer being told to hit the revoke switch, and lots of people would find their OS going poof.
  5. Pull the CMOS battery, give it a bath and it soak for anywhere from a couple hours to overnight, pull it out, remove any visible residue using a tissue and some alcohol. Set it somewhere safe and let all the alcohol evaporate, stick CMOS battery back in and hope it works. Although to be perfectly honest if I was paying you for a new PC, I'd expect my mobo never to have encountered energy drink. If you can save it and use it yourself, that's cool, but my two cents is professional ethics dictate the customer get a new one or a price reduction.
  6. Nope, FX-8350 doesn't really offer a fix to the poor single core performance of the FX series. Going from 6 cores to 8 cores is diminishing returns. Unless you have a specific game you really value your FPS in and you know it benefits from more cores, it's just not really an attractive move given your median game is four threads right now. You're better served to live with the bottleneck and wait for Zen/Kaby Lake and see which one you like or jump to Intel right now.
  7. The GTX 980Ti is GM200, whereas the M4000 is GM204. So the 980Ti actually has more CUDA (2816 vs 1664) cores and the ability to run faster thanks to GPU boost. It lacks the ECC and color depth though. In most rendering benchmarks, a 980Ti only lags slightly behind a M6000. So unless you need a Quadro specific feature, 980Ti. As a side note if you need to do lots of double precision work, the older Kepler architecture is actually better.
  8. I think the never preorder thing too broad of a generalization. I would agree never preorder from large studios that already have the cash to fund a game. However when you look at smaller studios, they don't have the luxury of having those resources. So to develop the game they have to get investors to back them. Given how hard the SEC makes it to offer out part of your company (which is why Kickstarter is structured the way it does and makes sure equity is never sold when you back something), a preorder is the next best thing. At that point you're making the deal that if we basically giv
  9. You could try something like offering 150 if he lets you come over and benchmark it, 100 if he doesn't. If he's legit and looking for a quick buck he gets slightly more money and still get a deal and some insurance. If he gets really hesitant over the the bench marking, it's possibly a scam.
  10. I personally would not describe it as decent, but that is subjective and also depends on how you value your time.
  11. So go search for R9 390 brands you were just supplied on Amazon. Are you really that helpless?
  12. I'd only buy it if it was in the system and I could run Aida64 on it for awhile. 125 seems like he potentially cooked it to the point it has a slight instability and now he's passing his failure on to someone else. Don't trust just a post test, show up and get him to download Aidq64 and show you in person.
  13. Go with the i5. It has better IPC and lazy developers are much more likely just throw everything into a couple intensive threads. People will be able to cite the edge cases where due to the open world nature of the specific game or because for some weird reason the developer spent a bunch of time adding in multithreading, but didn't make it gracefully step down to 4 cores and end up with it running like shit on the i5. So for a midrange, go with an i5 and focus more on the GPU since that helps the most. About the only time you'd do more cores in a budget build is when you were doing a
  14. If you have the rig listed in your profile, scrap your 4G 960 since the memory bus is too slow to make use of it, get a 970 or R9 390 and a SSD. If you have a Skylake i5 with 16 GB RAM, you should really have a better GPU than the 960.
  15. If you're used to lower end computers, I'd put the money have budgeted for the work on your laptop off to side and start saving. When you mess with laptops you are playing with the worst performance per dollar ratio out there.
  16. Pretty much what others have said: GTX 750Ti for ~120 dollars. , R9 380 for a~200. Given the caliber of the card you just killed, you could also consider an EVGA B-Stock GTX 750 for 90 dollars. I'd personally put down the extra 30 for the 750Ti, but if you need to save the money is an option. By the way don't try to overclock a B-Stock cards, they're often ones that EVGA had issues overclocking for their premium offerings. So plug it in and leave it be.
  17. Whichever is cheaper, unless the cheaper one has drastically higher latency. Or whichever one you think looks better through the case window.
  18. Given how cheap DDR3 is getting, 16 GB is kind of becoming one of those "Hey if you have the money, might as well do it while you're in there." Get one kit, have it all look the same, and remove any minor headaches that might come when you upgrade later. 16 is mostly useful because it saves you from having to close of other things when you do play the games that are RAM hungry.
  19. Perfectly fine for gaming and everyday use. Get 2133 MHz with the lowest latency you can find and enjoy. Really only worth pushing it for rendering workloads.
  20. PC for the mods that will come. Most Bethesda games are best with a combination of official patches and some fan made tweaks, and the PC will get the widest selection of mods (XBox ones will be more locked down).
  21. If you like the Platinum look you might also want to consider Klevv.
  22. Don't mess around with the integrated graphics. Get a cable for your monitor that can plug in to the second port on the Quadro. If your monitor cannot support that kind of input, then stick to one monitor.
  23. You can use both, but you will restrict yourself to lowest common denominator in many professional applications, which would be the iGPU. If the second monitor can accept display port, just get a display port cable off Monoprice or Amazon. Your iGPU is likely disabled via either OS or BIOS so that said professional applications only see the Quadro and perform accordingly.
  24. If you're buying VR, you want to go AMD. Oculus has openly stated they consider Maxwell's context switching a major problem which makes the 980Ti less attractive there. What would a single Fury or Fury X run you? If VR is a priority it might be wise to with one of those now, it will also deliver for ultrawide 1440p. Then when you buy the VR you can also determine if a second card is needed or not based on what you buy. Regarding the Fury though, if it is significantly more than a 390X, avoid it. It's kind of in an awkward spot, where if you can get it for not much more than a 390X it's s
  25. I still think people are overestimating how quickly DX12 becomes relevant. As of Sept, 53% of Windows installs are still 7 and Win10's growth has mostly come from people on 8.1 going to 10. Now obviously games/enthusiast types are more likely to upgrade and grandma is more like to still be on Windows 7, but that's still a big portion of the market booting to an OS that can't run DX12. Give DX12's unity with the XBox, we'll see some Xbox/Windows 10 exclusives that use that, likely at Microsoft's behest, but for people not taking marching orders from Redmond, they still have tremendous incent