Jump to content

Ashtefere

Registered User
  • Posts

    150
  • Joined

  • Last visited

  • Days Won

    4

Posts posted by Ashtefere

  1. Brian, according to that screenshot the fan on GPU2 is running at much lower speed than GPU1. Can you audibly check the fans when loading to see if GPU 1 is running noticably higher than GPU2?

    If that is the case, could be a dying fan. If the fan is OK, then it is a bios issue and needs an update.

    What I would do is try run a few wires from GPU2 fan to GPU1 fan header to test if the fan runs at the same speed, or just swap the fans around. I would say thats your issue.

    -Ash

  2. Uncompressible data is an unrealistic test for any drive. You arent going to be using 100% uncompressible data. So yeah, random is better for all benchmarks.

    So the tweaks didnt change anything huh? Thats a bit of a worry. The intelPPM trick will most definately help though, as it disables the auto cstate change.

    Looking forward to results with your m18x ;)

    -Ash

  3. An underclocked 6950 core can be unlocked to an underclocked 6970 core. And then overclocked to a 6970 level.

    Then people melt their laptops.

    But if I could, by the fucking gods you bet I would. I would do whatever ungodly cooling mods necessary to have two fully unlocked desktop 6970's in a laptop in crossfire. That would be just... that would just... it would... ack!

    -Ash

    • Thumbs Up 2
  4. OCZ's business practices are more than questionable.

    I would personally go for the Corsair Force 3 series at this point, over OCZ.

    Without a doubt, the corsairs should be the fastest drives out by about 5k-10k IOPS.

    I run a datacenter with many of the corsair force 2 drives, and havent had a single DOA (which is suprising for SSDS) whereas the OCZ equivalent drives had a 1 in 4 (yes, thats one in four) DOA rate. Stopped buying them once I figured that out and switched to the corsairs.

    Sandforce v1 drives have a few issues, for example if you fill it up completely with incompressible data you can never trim it to full performance again. Thats pretty fucking bad.

    The new Sandforce drives (vertex 3, force 3) are exceptional. They have excellent wear leveling (90 years+ at full thrashing all the time) and are the highest performance - so long as the manufacturer builds high quality drives with the chips (hence the OCZ issues...)

    The intel drives are higher build quality than any other drive, bar none. Its ridiculous. But you get slightly less performance and slightly less wear leveling reliability. We are still talking 60-80 years continuous use so its not something to worry about though.

    -Ash

    • Thumbs Up 1
  5. Australian alienware website just came up with pricing. Its fucking hilarious.

    Dell AU is even greedier than USA.

    Starting price is $3299 AUD.

    Yeah, fuck that.

    -Ash

    Hi Alex.

    I have been waiting for the M18x, and was using the Dell USA site to guage pricing before release.

    You see, because the australian dollar is quite a lot higher than the US dollar at the moment, I expected dell to have fairer pricing with the m18x when it was released in australia.

    Obviously, this is not the case.

    As you can see here: http://configure.us.dell.com/dellstore/config.aspx?oc=dkcwkr1&c=us&l=en&s=dhs&cs=19&model_id=alienware-m18x

    The USA pricing for the same laptop is $2269 USD (only $2101 AUD!!!)

    In australia it is $3299 AUD. This is nearly $1200 more for the exact same laptop, shipped from the exact same global location (malaysia was the last place I had my m17x shipped form).

    I know, personally, that the price to ship the laptop to australia is the same or less than it is to USA.

    I know that the AUD is higher than the USD.

    What I dont know, is what costs the extra $1200? Was the crate made of gold? Did an olympic swimmer drag it along the ocean by a golden string in their teeth?

    I can order this price through a USA drop shipper (as I did for my m14x) and only pay a very small price in customs duties. For the m18x this would be less than $200. I know dell must also pay this - so that leaves the extra $1000 unaccounted for.

    I am sorry, but I cannot buy any more products from Dell Australia. They have demonstrated unrivaled greed for no reason, and I am not going to feed that business model. Instead, I will be placing this order through Dell USA.

    Thank you for letting me know, though Alex. Feel free to forward this along to your management so they can understand customer sentiment.

    I will also be letting other australian users of cheaper alternatives, as it is harmful to australian consumers to feed this kind of business model.

    -Jonathan

    Thats the email I sent the rep who notified me. Cant wait for the response!

    -Ash

    • Thumbs Up 1
  6. Worst case scenario the notebook SSD will be a few seconds slower than a fully powered desktop SSD. The benchmarks will show it to be a lot slower but real world usage you won't really see it.

    Yeah, I realise that but for the money I am paying, the premium I am paying, and the work I am doing I expect and need it to be 100%.

    I need to make package files and compile code many times an hour, sometimes 50 times or more, and over the course of a day or a week it adds up very quickly to be a massive unnecessary timesink.

    For gaming, no one is gonna miss it but for my work it is a dealbreaker unfortunately.

    -Ash

  7. Happened to me :(

    But it was my first ssd, 8gb. I thought I was so awesome! Early adopter and all.

    Such a scam!

    -Ash

    Edit: Huuy, if you have windows on the SSD, make sure the page file ISNT on the ssd.

    Writing to the drive causes the stutter, so pagefile, temp files, save game, etc will all cause the stutter.

    • Thumbs Up 1
  8. Re: Suttering

    This isnt microstutter - microstutter is very very different. Microstutter doesnt actually stutter, so its a bit of a misnomer. Microstutter is when 2 gpu's render their frames too close together, making the output look like almost half of the FPS currently being rendered. This is present on all multi-gpu systems, and can be somewhat rectified with vsync+triple buffer.

    What you are experiencing is most probably due to your SSDNOW having interrupt stutter, which happens with all jmicron SSDs. There isnt really a fix for this unless you buy expensive HDD caching software (i forget what it was called) or you toss it out. The whole jmicron generation was junk.

    Switching temporarily to a normal hard disk will show this to be the case.

    Sorry for your bad luck - whoever sold you the drive, try get a refund.

    -Ash

    • Thumbs Up 3
  9. Bulldozer isnt going to impress anyone because it isnt refined.

    Bulldozer is a fucking **amazing** jaw dropping design - it is the future of multi chip architecture (it would be perfect if it had intels ring bus tech though...) and the multi core system is amazing.

    Only problem is, its revision 1, and any performance hitches in it wont be attacked until further down the track.

    With each revision of bulldozer, it will increase in performance drastically. In about 3 revisions, AMD will be at least the same level of performance as intel, and then after that will come out on top again.

    For now, intel has reign. In 2+ years, that will end.

    -Ash

  10. Ivy is backwards compatible with Sandy bridge chipsets, thankfully, and seeing as CPU's dont have bioses it will work just fine - at worst it will display in the bios as "unknown CPU".

    Also, as for my "soon" comment on release, we are looking at a release around the time of bulldozer, so about June 11th for Ivy.

    Ivy is just entering mass production now, and intel usually mass produces for about 3-6 months before a release date in order to have enough stock to sell. Rumors are that production speed has been ramped up in order to steal bulldozer's thunder - and this is gonna be devasting for AMD.

    Ivy looks to be on average 30-50% faster clock for clock than bulldozer... :(

    -Ash

  11. Alright guys...

    Anyone considering a 2920XM should probably hold off on buying it!!!

    Intel announces 3D Tri-Gate transistors | bit-tech.net

    Check the link.

    Intel just dropped a fucking bomb on the microprocessor world. Ivy bridge is gonna drop soon, and will contain a much higher performance increase than originally hinted.

    Intel first said the ivy die shrink would be 20% more performance, but this is incorrect as they had an ace up their sleeve.

    Ivy bridge will now have a 37% performance increase over Sandy bridge, clock for clock!

    Thats nearly 40% boost... that is jaw dropping, and usually only tock implementations have that kind of increase in performance. In engineering terms, this is fucking insane!

    My recommendation is get the 2720qm for now, and ebay an ivy XM as soon as it drops!

    -Ash

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.