Download inSpectre

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jan 17, 2018.

  1. tensai28

    tensai28 Maha Guru

    Messages:
    1,416
    Likes Received:
    365
    GPU:
    2080ti MSI X TRIO
    I just disabled it all since I'm on a 9th gen cpu. Partial protection is good enough and it wouldn't even let me enable one of them.
     
  2. nick0323

    nick0323 Master Guru

    Messages:
    985
    Likes Received:
    43
    GPU:
    Palit GTX970 Jetstream OC
    It's been slow ever since the April update from 2018. I turned off Windows Hello and that helped a bit as that tends to use quite a bit of memory shortly after logging in. Even with the October update it wasn't any better. Feels like a clean install of Windows now with the mitigations disabled.

    Incidently the InSpectre utility says the performance is bad on my 2500k but I couldn't feel/see a difference on my PC when I turned off the mitigations. Probably all in the numbers on a benchmark or all in my head.
     
    Noisiv likes this.
  3. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    If you're looking for a performance boost using this program it is pointless, I have purchased everything for 3D Mark.

    Restarted the systems did a 3D Mark benchmark which covers a CPU specific benchmark.

    GPU Score - 7859
    CPU Score - 8659
    Overall Score - 7859

    I then disabled the Meltdown and Spectre protection and restarted the computer, then opened up inspectre again to confirm protection had been disabled and then ran the same benchmark again.

    GPU Score - 7723
    CPU Score - 8423
    Overall Score - 7820

    Disabling Meltdown and Spectre via "InSpectre" does not increase performance, instead it lowers performance.
    My System:

    OS - Windows 10 Professional
    CPU - i7 8700k
    GPU - EVGA GTX 1080 SC
    RAM - Corsair Vengeance 32GB DDR4 3200MHz
     
  4. Astyanax

    Astyanax Ancient Guru

    Messages:
    7,986
    Likes Received:
    2,668
    GPU:
    GTX 1080ti
    your testing is faulty.
     
    Deleted member 213629 likes this.

  5. BuildeR2

    BuildeR2 Ancient Guru

    Messages:
    2,888
    Likes Received:
    146
    GPU:
    MSI 2080 Ti GX Trio
    Regarding the posts by @Astyanax and @Naraus this is what I found on my 4770k + 1080Ti machine.

    Before disabling anything using InSpectre my 3DMark scores were higher as well, but games had noticeable hitches and stutters. After disabling both options via InSpectre my synthetic scores did drop a tiny bit, but most games were markedly improved.

    I've kept all the stuff disabled and haven't looked back. I was sick of the stuttering and hitching across many games and engines. Again, this is just my experience. I hope at some point that as more and more performance sapping exploits are discovered, this tool will be updated so we can disable even more of them.
     
    Astyanax likes this.
  6. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    3D Mark is not faulty, and is being keep't up to date via Steam.
    "Ohh but your anti-virus could have something".
    I use Norton, it goes silent during fullscreen applications and only does background tasks and updates after the PC has been idle for over half an hour.
    "Ohh but you probably alt-tabbed during the test and something"
    3D Mark automatically cancels any test if you "Alt-Tab".
    OH BUT BUT- I did clean restarts, resources used were at a clean slate before every test.

    You know what websites are saying about the i7 8700K? that the Spectre and Meltdown patches have a "24.8 percent" performance impact for "My CPU Model", and just so you know 3D Mark does a "CPU" specific benchmark that averages out at 29 FPS on my CPU, that benchmark has no fps cap on it.
     
  7. Astyanax

    Astyanax Ancient Guru

    Messages:
    7,986
    Likes Received:
    2,668
    GPU:
    GTX 1080ti
    Reliance on one test to demonstrate something doesn't occur is FAULTY TESTING.
     
    Alessio1989 likes this.
  8. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    Also tested my Laptop - ASUS ROG Zephyrus S GX701

    OS - Windows 10 Professional
    CPU - i7-9750H
    GPU - RTX 2070

    Same result, lower performance.
     
  9. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    And the "24.8 percent" performance impact hmm?

    So what you're saying is that if I downgraded my GTX 1080 to a GTX 980 I would likely see "very close" GPU Scores, despite the massive gap in performance on official Benchmarking sites?
    Oh let me guess? you have done back to back Benchmarks in the past? hmm the performance got better after each test didn't it? That because everything from the last test is now already pre-loaded, what a surprise.
    BTW the patches are still disabled on my system and I just did another test, only takes a few minutes, hmm yeah... still not seeing that +20% jump on my CPU Score.

    FYI, 3D Mark doesn't do "one test" it runs "multiple tests".
     
  10. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    A logical answer to why my CPU score doesn't gain the 20% boost after the patch and see a score of 9500 - 10,000 is simple.
    Intel have had almost an entire year to fine tune the Spectre and Meltdown patches to the point it no longer has a large performance impact.
     

  11. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    Oh look how nice I found a 3D Mark benchmark of the same CPU back in 2017 before Spectre and Meltdown had been discovered:

    Intel Core i7 8700k, MSI GTX 1070 Gaming X

    His CPU score is "7606"
    Much lower then both my tests.
    FYI when 3D Mark does the "CPU Test" MSI Afterburners OSD reports a GPU usage of sub 15%, so no the 1080 won't increase my "CPU Score". Then again I am using a "Corsair Hydro Series H150i PRO" with 3 fans on both sides of the Rad.
     
    Last edited: Nov 23, 2019
  12. -Tj-

    -Tj- Ancient Guru

    Messages:
    16,979
    Likes Received:
    1,843
    GPU:
    Zotac GTX980Ti OC
    I lost 400 points by cpu score on vs off..

    Initial retopoline fixed it, but now after few win updates scores went down again.

    Its only 1-2fps difference though. 40 vs 42fps.
     
  13. Astyanax

    Astyanax Ancient Guru

    Messages:
    7,986
    Likes Received:
    2,668
    GPU:
    GTX 1080ti
    intel haven't fine tuned anything.

    if you'd bothered doing your research you'd know why the mitigation on results can give higher static benchmark results while doing worse in real dynamic performance scenarios.
     
    Alessio1989 likes this.
  14. Alessio1989

    Alessio1989 Ancient Guru

    Messages:
    1,874
    Likes Received:
    513
    GPU:
    .
    It's your methodology which is broken. Disabling or change a couple of registry entry clearly is not enough, you are not using a scientific approach at all. You are trapped in a complex system noise scenario, which prove itself again that is not a Lego.
     
  15. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    Ah yes, "real dynamic performance scenarios".
    You mean like how 3D Mark doesn't just constantly spin a graphic like in furmark (which isn't that dynamic) and instead has the CPU deal with constantly evolving graphics and dynamic physics movements? ok... it's obvious you have not seen a 3D Mark benchmark run.
    Ah but there is the "random unpredictable movement of a real person playing a game", in which it is the RAM that handles and literally pre-loads to predict said dynamic movement.

    Or how about the fact that the "24.8 percent" performance hit that meltdown and spectre is mean't to have on the "i7 8700k" was proven with a "synthetic benchmark".
    Just did 10 more tests and still it doesn't get any higher than my initial test with the patches enabled.

    Also FYI between the space of 7 months post spectre / meltdown my BIOS received 3 updates all with the description "Update Intel Micro code for security vulnerabilities"
    https://www.msi.com/Motherboard/support/Z370-GAMING-PRO-CARBON-AC

    Here is a benchmark from Tom's Hardware on the i7 8700k using the same benchmark run that I did
    https://www.tomshardware.com/reviews/intel-coffee-lake-i7-8700k-cpu,5252-4.html

    Oh look at that a result of "7553" with no overclock, only their overclock to 4.9 ghz gets close to my stock performance.
    And when was this done? "October 05, 2017", hang on that's before spectre and meltdown even existed, shouldn't my CPU be getting scores in the sub 6000 range? and not 8500+?!
    I also went back and looked at the 20+ saved benchmark runs I did in the past 10 months and not once does it ever show my CPU getting sub 8000.

    On the subject of Intel not fine tuning anything, what have AMD being doing with their 3rd gen Ryzens in the past few months? pumping out a bucket load of bios updates to fine tune performance because of their initial false advertising?
    Are you under the assumption that base CPU performance cannot be improved with BIOS updates?
    Do you even bother to update the BIOS and the Intel Chipset Drivers?
     

  16. mbk1969

    mbk1969 Ancient Guru

    Messages:
    9,851
    Likes Received:
    6,803
    GPU:
    GF RTX 2070 Super
    You mean AMD chipset drivers? Intel ones are just inf-files without any actual drivers.
     
  17. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    "In computing, an INF file or Setup Information file is a plain-text file used by Microsoft Windows for the installation of software and drivers. INF files are most commonly used for installing device drivers for hardware components."
    https://en.wikipedia.org/wiki/INF_file

    The MSI link to my motherboard in a previous post also has Intel Chipset drivers which contains more than just INF files.
    And yes the AMD Chipset Drivers do include such files and other additional files like .dll and .img(??) because AMD Chipset Drivers include Graphic drivers while Intel provides iGPU drivers separately.
    Hence why AMD Chipset drivers are 350MB as apposed to Intel's 5MB Chipset Drivers.

    And for AMD Boards from MSI you're still installing Intel Network drivers.
     
  18. mbk1969

    mbk1969 Ancient Guru

    Messages:
    9,851
    Likes Received:
    6,803
    GPU:
    GF RTX 2070 Super
    @Naraus

    Here is the Win10-x64 folder from the Intel chipset driver package (downloaded from here https://downloadcenter.intel.com/download/28182/Chipset-INF-Utility?product=53):
    [​IMG]
    You can see only inf-files and cat-files (https://docs.microsoft.com/en-us/windows-hardware/drivers/install/catalog-files). So no actual binary driver files like sys-files.

    PS I replied to your post only with that thought - if Windows found all Intel devices (mostly system ones), then there is no benefits in updating Intel chipset drivers.
     
    Last edited: Nov 23, 2019
  19. Naraus

    Naraus Member

    Messages:
    14
    Likes Received:
    3
    GPU:
    EVGA GTX 1080 SC
    @mbk1969

    Yeah that is a nice point but I don't pick and choose, I update "everything" and perform maintenance on my system and physically clean it out every few months, and no surprises my hardware performs better than it did 2 years ago.
    A fact that is hard to swallow in those that believe "Corporations" (THE MAN) cause hardware to slow down on purpose to make you think about upgrading. Apple on the other hand..
     
  20. Holy crap..... I thought I drank coffee and got worked up .... only 9 messages in and making waves
    true... "planned obsolescence"
     
    Last edited by a moderator: Nov 24, 2019

Share This Page