repaxan wrote on 2020-07-02, 01:34:

Yes, I installed the 96.85 Vista drivers (despite that I'm using Win 7), which is the one listed as the latest Vista driver on Nvidia's download page. I'm not sure where you see the 160.02 one.

Thank you and all looks OK. I went in via Beta and Older Drivers and https://www.nvidia.co.uk/download/driverResul … .aspx/237/en-uk should download 160.02.

I have now released SIV 5.50 Beta-04 so please use >= to that.

mastergamma12 wrote on 2020-07-02, 01:43:

That's probably an Xp driver.

No, it's Vista-64, check the above link.

mastergamma12 wrote on 2020-07-02, 01:43:

I remember trying that driver on either Windows 8 or 10 with a Quadro FX 3000 and it was a major PITA.

UWP stuff crashed the system.

I am not surprised and in truth I am surprised the Vista driver runs on W7.

red-ray wrote on 2020-07-02, 07:51:

mastergamma12 wrote on 2020-07-02, 01:43:

That's probably an Xp driver.

No, it's Vista-64, check the above link.

Wow, wasn't even aware that existed.

Does that driver work on Geforce FX or is it Quadro only?

mastergamma12 wrote on 2020-07-03, 04:44:

Does that driver work on Geforce FX or is it Quadro only?

Looking at the [ SUPPOPRTED PRODUCTS ] tab then it's just Quadro. The latest GeForce (Rankine) Visia-64 driver looks to be https://www.nvidia.co.uk/Download/driverResul … .aspx/837/en-uk

If theory it should be possible to get the Quadro Driver to be used for a GeForce GPU by editing the .INF files, but as to if it would work I have no real idea.

Looking at https://www.techpowerup.com/gpu-specs/?archit … sort=generation then there are both GeForce + Quadro NV38 GPUs so it might also be possible to flash such as a GeForce PCX 5950 with the Quadro FX 1300 firmware and then use the Quadro driver.

Which Rankine GPUs do you have?

red-ray wrote on 2020-07-03, 06:45:

Looking at the [ SUPPOPRTED PRODUCTS ] tab then it's just Quadro. The latest GeForce (Rankine) Visia-64 driver looks to be https […]
Show full quote

mastergamma12 wrote on 2020-07-03, 04:44:

Does that driver work on Geforce FX or is it Quadro only?

Looking at the [ SUPPOPRTED PRODUCTS ] tab then it's just Quadro. The latest GeForce (Rankine) Visia-64 driver looks to be https://www.nvidia.co.uk/Download/driverResul … .aspx/837/en-uk

If theory it should be possible to get the Quadro Driver to be used for a GeForce GPU by editing the .INF files, but as to if it would work I have no real idea.

Looking at https://www.techpowerup.com/gpu-specs/?archit … sort=generation then there are both GeForce + Quadro NV38 GPUs so it might also be possible to flash such as a GeForce PCX 5950 with the Quadro FX 1300 firmware and then use the Quadro driver.

Which Rankine GPUs do you have?

Only a Quadro FX 3000 atm.

I used to have a bunch of FX 5200s and FX 5500s but that was a while ago.

red-ray wrote on 2020-07-02, 07:51:

Thank you and all looks OK. I went in via Beta and Older Drivers and https://www.nvidia.co.uk/download/driverResul … .aspx/237/en-uk should download 160.02.

It didn't want to install, said something about not being compatible with the hardware, further the PCI device ID (00FE) wasn't listed in the INF file, closest ones being 00F8 and 00F9. Seems like just a mislabeled driver to me.

Haswell system, in the main screen it shows the FIVR's Vcc as Vcore.

repaxan wrote on 2020-07-08, 20:08:

Haswell system, in the main screen it shows the FIVR's Vcc as Vcore.

Thank you for the report, bay I can't see an issue. SIV is displaying VIN0 VCCIN which is 1.76 volts. I guess you mean VCore (Max), do you? This is the maximum of the VIDs of the four CPU cores.

repaxan wrote on 2020-06-28, 19:42:

It's 'new to me', straight from ebay. I tried to pop off the heatsink and see what the die says, but it was held on so tightly that it felt like it might be thermal adhesive, so I stopped.

I have been trying to figure out what if doing on and given the Chipset ID is NV35 and the BIOS reports it as NV35 I am pretty sure it's an NV35 and have adjusted SIV in light of this.

Please let me know what [GPU Info] now reports for your card using the attached SIV64X V5.50 NV35-03 test SIV.

The test SIV also has the Menu->Help->Conversions->Prime Numbers panel I have been developing. In general it does OK on Intel + VIA/Centaur CPUs and rather badly on AMD, especially Ryzen. I suspect this is down to how the caching works on AMD. I would be interested to see how it does on your system and wonder if you have any thoughts as to why it's much slower on AMD.

red-ray wrote on 2020-07-08, 20:42:

repaxan wrote on 2020-07-08, 20:08:

Haswell system, in the main screen it shows the FIVR's Vcc as Vcore.

Thank you for the report, bay I can't see an issue. SIV is displaying VIN0 VCCIN which is 1.76 volts. I guess you mean VCore (Max), do you? This is the maximum of the VIDs of the four CPU cores.

I meant in the main screen it shows "Core 1.76 volts" (VCCIN value) when it should be saying "Core 0.90 volts" (Vcore value).

repaxan wrote on 2020-07-09, 21:05:

red-ray wrote on 2020-07-08, 20:42:

repaxan wrote on 2020-07-08, 20:08:

Haswell system, in the main screen it shows the FIVR's Vcc as Vcore.

Thank you for the report, bay I can't see an issue. SIV is displaying VIN0 VCCIN which is 1.76 volts. I guess you mean VCore (Max), do you? This is the maximum of the VIDs of the four CPU cores.

I meant in the main screen it shows "Core 1.76 volts" (VCCIN value) when it should be saying "Core 0.90 volts" (Vcore value).

No, the four per Core VIDs are reported per core just to the right of the % CPU load plots. The Core to the right of the L1 Code Cache information is the voltage supplied to the package which is VIN0 VCCIN. On some CPUs they are the same voltage, but on others such as Haswell + Skylake-X they are not.

red-ray wrote on 2020-07-09, 21:16:

repaxan wrote on 2020-07-09, 21:05:

red-ray wrote on 2020-07-08, 20:42:

Thank you for the report, bay I can't see an issue. SIV is displaying VIN0 VCCIN which is 1.76 volts. I guess you mean VCore (Max), do you? This is the maximum of the VIDs of the four CPU cores.

I meant in the main screen it shows "Core 1.76 volts" (VCCIN value) when it should be saying "Core 0.90 volts" (Vcore value).

No, the four per Core VIDs are reported per core just to the right of the % CPU load plots. The Core to the right of the L1 Code Cache information is the voltage supplied to the package which is VIN0 VCCIN. On some CPUs they are the same voltage, but on others such as Haswell + Skylake-X they are not.

I get that, my point is that it should not be labeled 'Core'.

repaxan wrote on 2020-07-10, 01:06:

red-ray wrote on 2020-07-09, 21:16:

repaxan wrote on 2020-07-09, 21:05:

I meant in the main screen it shows "Core 1.76 volts" (VCCIN value) when it should be saying "Core 0.90 volts" (Vcore value).

No, the four per Core VIDs are reported per core just to the right of the % CPU load plots. The Core to the right of the L1 Code Cache information is the voltage supplied to the package which is VIN0 VCCIN. On some CPUs they are the same voltage, but on others such as Haswell + Skylake-X they are not.

I get that, my point is that it should not be labeled 'Core'.

It should not be labelled VCore and Core is ambigous, but given the tooltip makes it clear what it is this is no big deal. The SIV64X V5.50 Beta-08 should say VCCIN when the sensor name contains VCCIN, but I am pondering if it should say Vccin, what do you think?

Imperious wrote on 2020-06-30, 12:53:

Too much hassle trying to install NT4, XP, or Win2000 on this computer but may try to boot xp off Hiren's boot cd and try SIV in that, but the questions been answered.

I wondered is SIV would work out the FSB, so got a Little Foot and tried it 😀. I set the FSB to 68.5 MHz as I felt 66.6 MHz was too easy.

As expected there is no MSR C0000087

red-ray wrote on 2020-07-10, 13:04:

It should not be labelled VCore and Core is ambigous, but given the tooltip makes it clear what it is this is no big deal. The SIV64X V5.50 Beta-08 should say VCCIN when the sensor name contains VCCIN, but I am pondering if it should say Vccin, what do you think?

Personally I feel like it would be more useful to show the Vcore and not the Vccin but that's just me.

red-ray wrote on 2020-07-09, 19:13:

I have been trying to figure out what if doing on and given the Chipset ID is NV35 and the BIOS reports it as NV35 I am pretty sure it's an NV35 and have adjusted SIV in light of this.

Please let me know what [GPU Info] now reports for your card using the attached SIV64X V5.50 NV35-03 test SIV.

Attached - I moved the card to an XP system where the driver works much better than it did on 7.

repaxan wrote on 2020-07-11, 20:31:

red-ray wrote on 2020-07-10, 13:04:

It should not be labelled VCore and Core is ambigous, but given the tooltip makes it clear what it is this is no big deal. The SIV64X V5.50 Beta-08 should say VCCIN when the sensor name contains VCCIN, but I am pondering if it should say Vccin, what do you think?

Personally I feel like it would be more useful to show the Vcore and not the Vccin but that's just me.

How is it possible to "show the Vcore" when there is one per core, so four in total? SIV already reports all four of them so to say the least I am confused

red-ray wrote on 2020-07-11, 23:14:

How is it possible to "show the Vcore" when there is one per core, so four in total? SIV already reports all four of them so to say the least I am confused

Just show max or average or something, this is already happening on a normal (non-fivr) system

repaxan wrote on 2020-07-12, 02:26:

Just show max or average or something, this is already happening on a normal (non-fivr) system

It's doing this as such as laptops don't report VCCIN and there is no choice, further your Westmere-WS does not report VCore, what should happen then?

It all comes down to personnel preference, I prefer VCCIN and I am unclear as to what you would like on such as your Westmere-WS system.

I have made some changes to the attached SIV64X 5.50 Core-09 test SIV, does it do what you would like and yes, you can change it on-the-fly 😀

Update: SIV64X 5.50 Beta-09 has Menu->Edit->Options->Show VCore (Max) 😀

Tried SIV on my H370 Stinger and like the Z370 Micro and Z390 FTW before it was fixed, SIO information isn't showing 12v. Though I can't find 12v on any other software.

mastergamma12 wrote on 2020-08-13, 22:51:

Tried SIV on my H370 Stinger and like the Z370 Micro and Z390 FTW before it was fixed, SIO information isn't showing 12v. Though I can't find 12v on any other software.

Thank you for the report and given the EVGA utility does not report +12 I suspect it's not reported at all. Give me a few hours and I will ask around then post a test SIV that should at least improve the other readings.

I spotted INUSE Wait 812 of 812 ms which tells me that some software is failing to release the SMBus hardware lock as the first 812 shows that SIV timed out waiting for it. If you exit all monitoring software and then just run SIV it should be INUSE Wait 0 of 812 ms, is it? Assuming so then please start the other monitoring software to see which one triggers the issue.

The INUSE Wait 0 of 812 ms is also displayed on Menu->Help->Lock Handle so when doing the test use this screen to check so we can also see if the offending utility uses the Global\Access_SMBUS.HTP.Method lock. Below you can see both SIV and HWiNFO do use the lock.

Update: I have attached the SIV64X V5.52 EVGA-00 test SIV that should fix all the temperature, fan and voltage names, does it please? BTW why is the RAM is running at 2.66GHz when the SPD says it's capable of 3.20GHz?