Jump to content

JCBiggs

Member
  • Posts

    853
  • Joined

  • Last visited

Reputation Activity

  1. Like
    JCBiggs reacted to RONOTHAN## in Mid Teir vs HPDT CPU's   
    Nope. Ryzen chips always boost until they hit 95C. If you're just running 2 threads they should be boosting up to somewhere around 5.5GHz-ish, with the 7900X being ~100MHz higher. 
     
    I mean, I guess, but at the same time there are still probably gonna be better options. The 5800X3D has such large and fast cache that the memory bandwidth of DDR5 would be borderline irrelevant, and in a lot of single threaded workloads (it will depend what you're doing for whether it helps or not, though most of the time it will). Raptor Lake is releasing in less than a month, and that should be about the same single thread performance of Ryzen 7000 but for significantly less money and (if Intel is to be believed) not a furnace this time around, unlike the Ryzen 7000 series chips which are next to impossible to keep cool. . 
  2. Like
    JCBiggs reacted to RONOTHAN## in Mid Teir vs HPDT CPU's   
    Yes.
     
    Not necessarily how it works, the higher up on a the product stack, the better the silicon quality and the higher it will boost on one thread. A 7700X will boost slightly higher than a 7600X, a 7900X will boost higher than a 7700X, etc. All core clock speeds will go down because the chips become utterly uncoolable trying to hit the same all core boosts than the lower core count chips achieve, but the single thread performance does still go up as you go up the stack. 
     
    Whether that actually matters is a different story, it's within 100-200MHz, completely unnoticeable. 
     
    It won't, the 7600X will be a lot faster, though realistically if you're already on AM4 you should just get a 5800X3D, for a gaming chip there isn't a reason to spend the $300 on a motherboard, $300 on a CPU, and $150 on a kit of RAM to get the same performance of just spending $400 on a new CPU and keeping everything else, the 5800X3D does match gaming performance of the AM5 chips. 
  3. Like
    JCBiggs reacted to bmx6454 in PC Bypasses Bios. Impossible to enter Bios   
    i would agree with trying without drives connected then.
  4. Informative
    JCBiggs got a reaction from bmx6454 in PC Bypasses Bios. Impossible to enter Bios   
    tried that.  didnt work.
  5. Like
    JCBiggs got a reaction from King_PIN in Linus's Home theater.   
    80 anything is super low imo.  My 2 sets of  klipsch towers are 97 and 98db. Obviously I'm going to pick sound quality first,  but sensitivity seriously pays off when it comes time to buy power.  That's not to say there isn't excellent speakers in the low 80s even... they just don't need to be driven with an avr. Even if it is a $6000 one.
  6. Funny
    JCBiggs got a reaction from Lightwreather in Linus's Home theater.   
    That escalated quickly
  7. Like
    JCBiggs reacted to Shike in Linus's Home theater.   
    First post, but love HT/audio so decided to post here first.  Rather than just the HT I'm going to list all rigs:
     
    My apartment livingroom:
     

     
    The audio for the home theater consists of:
     
    4x JBL Professional AC26 1x JBL Professional AC28 (Center Channel) Target Audio Stands (sand filled) front - Gator Frameworks in back with Skrim 2x DIY Peerless/Tymphany 18" subwoofers Marantz AV7701 Pre-Pro Emotiva UPA-700 Amplifier Crown XLi1500 for subwoofers MiniDSP 2x4 Software of Note: AudioMulch with various VSTs, Jriver, Tuneblade TV - Some 55" LG with IPS, fuck if I know off the top of my head. All audio comes from my PC using Voicemeeter Potato virtual loopback via Audiomulch.  First I used MSO (multi-sub optimizer) to equalize both subwoofers through the 2x4 independently using a calibrated UMIK-1.  The second requirement was global EQ, using REW focusing on MLP per Harman's AutoEQ study instead of multiseat correction I utilized PEQ VSTs within Audiomulch.  This results in roughly a 1dB/octove roll-off from low to highs.  I then also have various other VSTs including Waves for different RT upmixing, dynamic range compression for night listening, etc. to get it dialed how I like best (Dolby and DTS suck for this when it comes from non-bitstreamed sources).  Crossover is set to 80hz as traditionally standard - though I have the PC doing crossover and bass management so it can add a lowpass to prevent localization and a highpass to prevent any potential bottoming out of my subs. 
     
    I tried using DIRAC as well but ended up getting similar results on my own so didn't really see the benefit outside of it being faster, especially since I prefer minimum latency configuration.
     
    The setup is perfectly capable of Dolby reference levels for home, though I never dare run it that high in regular use.  Nonetheless I get effortless room filling sound.
     
    Used lots of "audiophile" speakers but honestly the right PA speakers - if you go in with the proper mindset - can run circles around a lot of them when you understand the problems involved.  For example most people that complain about poor sounding subwoofers typically have ones with design defects or haven't worked to properly integrate them into the room and with the speakers - assuming the room itself isn't a massive issue alone requiring multi-sub coverage which is always beneficial.
     
    My PC setup is more subdued:

     
    4x NHT SuperZero XU 1x Dayton 12" HF (early make) - 300W Bash sealed Denon E200 AVR (doesn't need anything fancy really) V-Can Amp and NuForce DAC (not really used anymore) AKG K550 and Audio-Technica ATH-AD700 headphones  
    For those ready to scream that the speakers are next to each other, this was intended since I'm using Ambiophonics.  Similar to Polk SDA or Carver's Sonic Holography without the downsides this uses recursive ambiophonic crosstalk elimination.  The close placement reduces combing artifacts from ye olde' stereo triangle allowing for 160 degree soundstage per "ambiopole" - effectively I'm only missing 20 degrees to my very left and very right in this configuration.  And since it's run through Audiomulch there's also correction and other nice VSTs to use . . . say dynamic range expansion and compression which is handy in games like certain "gamer" headsets have for advantages . . . cough
     
     
    The bedroom and reading nook:

     
    NHT C3 Pair (bedroom) Dana 630i Pair (SB Acoustics drivers) (nook) All using sand filled VTI stands Audiosource Amp 100 Cambridge DacMagic (original) Cheap Asus netbook of some make O2 Amp and some old Grado SR-60 for night listening if needed Amazon Fire with gApps to act as control point Russound speaker selector for bedroom/nook speaker selection  
    PC is used as audio renderer/end-point for all audio save the TV (which is routed to a simple selector switch, other selection being PC/DAC).  PC is running EqualizerAPO and a VST (Waves MaxxBass) which uses harmonics to extend perceived bass response another octave out (cheating I know).  The PC also has TuneAero and Bubble Open Home Server installed.  Primary PC has Tuneblade installed and also a web driven interface.  I can effectively use the "tune" system for whole home synced audio or use BubbleUPN and open home for individual renderer use.  Use a Russound selector to bounce between the NHT and Dana's without surprise jumps in level.
     
  8. Like
    JCBiggs reacted to leadeater in Linus's Home theater.   
    Need to also mention speaker sensitivity is often done at 1kHz, lower frequency requires more power so you won't be doing 60Hz at 91dB using 1W if the speaker has a sensitivity rating of 91dB at 1W/1m. It's also a non-linear power increase as you go down in frequency.
     
    Singular sensitivity rating really is too basic anyway, frequency responsive graph is much superior if you can actually get one. Easily done for the induvial speaker/driver but not the product as a whole.
  9. Like
    JCBiggs reacted to leadeater in Linus's Home theater.   
    Should be fine since they set the speakers to small so the actual power required will be low. Personally I object to setting speakers to small and just use actually good enough stuff but eh not everyone can or is willing to do that.
  10. Funny
    JCBiggs got a reaction from IkeaGnome in Linus's Home theater.   
    That escalated quickly
  11. Like
    JCBiggs reacted to GDRRiley in Linus's Home theater.   
    you do realize hes got both a home theater setup and a normal TV living room right?
  12. Funny
    JCBiggs reacted to Roswell in Linus's Home theater.   
    He took his time disclosing it but...
     
     
    In other words, this is fake, it's just a sponsored ad. Super unlikely that he's keeping them, it's just a glorified infomercial. I wouldn't waste energy worrying about whether or not he's using this stuff correctly.
  13. Like
    JCBiggs reacted to IkeaGnome in Linus's Home theater.   
    Not to brag, but TCL 65" tv and a sound bar with Bluetooth sub!
  14. Funny
    JCBiggs got a reaction from Slayerking92 in How to Remote locate Unraid Server driving VMs?   
    I need to move my CPU Tower.  Im thinking optical cables, 1 for each VM, but Is there a thunderbolt card that will work without Headers? I would be using a Mother/daughter card to expand a single pcie  slot to 3 slots. I have 12 lanes left. each card would use 4. I cant think of another way to do this. Maybe 6 thunderbolt docks/extenders?
  15. Agree
    JCBiggs got a reaction from Victor Golf in LTT Folding Team's Emergency Response to Covid-19   
    70c is not overheating. its perfectly normal. my quadros have been at 82c for 3 weeks now.   new GPUs are going to get just as hot. your goal should be to move enough air through the machine to end up with power limit throttling, not thermal throttling. 
  16. Agree
    JCBiggs got a reaction from Favebook in LTT Folding Team's Emergency Response to Covid-19   
    70c is not overheating. its perfectly normal. my quadros have been at 82c for 3 weeks now.   new GPUs are going to get just as hot. your goal should be to move enough air through the machine to end up with power limit throttling, not thermal throttling. 
  17. Agree
    JCBiggs got a reaction from TVwazhere in LTT Folding Team's Emergency Response to Covid-19   
    70c is not overheating. its perfectly normal. my quadros have been at 82c for 3 weeks now.   new GPUs are going to get just as hot. your goal should be to move enough air through the machine to end up with power limit throttling, not thermal throttling. 
  18. Like
    JCBiggs got a reaction from Victor Golf in LTT Folding Team's Emergency Response to Covid-19   
    ok ok.. ill join in..   i have 3 quadros running when workstation are idle, and about half my cpu's.     I think im putting out about 1.2 million per day.  Ill change my team to LTT shortly. 
  19. Funny
    JCBiggs reacted to GOTSpectrum in LTT Folding Team's Emergency Response to Covid-19   
    Folding@Home servers trying to get WUs out be like
     

  20. Like
    JCBiggs got a reaction from Daniel_SC in LTT Folding Team's Emergency Response to Covid-19   
    ok ok.. ill join in..   i have 3 quadros running when workstation are idle, and about half my cpu's.     I think im putting out about 1.2 million per day.  Ill change my team to LTT shortly. 
  21. Like
    JCBiggs got a reaction from GOTSpectrum in LTT Folding Team's Emergency Response to Covid-19   
    ok ok.. ill join in..   i have 3 quadros running when workstation are idle, and about half my cpu's.     I think im putting out about 1.2 million per day.  Ill change my team to LTT shortly. 
  22. Like
    JCBiggs got a reaction from TechyBen in Tesla pause German gigafactory construction after environmentalists protest cutting trees   
    this is why we should start teraforming deserts. it wouldnt even really be that hard now that the "global warming" is making the atmosphere more moist. 
  23. Like
    JCBiggs reacted to Master Disaster in Docker Updating, Nginx certs??   
    Can't answer the first question but for the second you need a DNS server running to translate domain names to IP addresses.
  24. Like
    JCBiggs got a reaction from leadeater in Leaked GPU Specs Suggest Xbox Series X Substantially More Powerful Than PS5   
    My point was more so based around the idea that the hardware manufacturers need to force higher quality code.   its amazing how fast a piece of highly optimized C runs on bare metal.   I know GPU's are basically as power efficient as you can get, so I dont know how much optimizing could be found there.
  25. Like
    JCBiggs reacted to Mathieu9836 in Ryzen 3800x Much Slower with XMP (very odd)   
    Memory timings set to auto will give you all sort of weird and awful timings, set them manually using Typhoon Burner and DRAM Calcultator.
×