Question I'm getting intermittent b1 code on Asus motherboard after installing a used GPU ?

[sizzlinkola]

A friend gifted me their Nvidia RTX 3060Ti 8GB. About 25% of the time when I power on the computer from shutdown, the mobo reads "b1" and gets stuck on the Asus logo for 1-2 minutes. Then, it gets to the login screen and works fine.

Most of the time it boots quickly in less than a minute with no issues since I have a SSD, just like I did with my old Nvidia GPU. But I'm concerned that I'm getting this b1 code 25% of the time.

Any ideas?

Specs
MBD: Asus Z97 Deluxe
CPU: i7-4770
GPU: Nvidia RTX 3060 Ti 8 GB
RAM: 32GB (16GB x 2 )

 

[JohnBonhamsGhost]

and what have you found related to this "b1" code?

if the card wasn't functioning then you'd have no display whatsoever.

i'd guess it's more related to the very outdated system containing this card.

include the complete system specs with the make & model of ALL components.
not just this brief description you've offered.

 

[sizzlinkola]

and what have you found related to this "b1" code?

if the card wasn't functioning then you'd have no display whatsoever.

i'd guess it's more related to the very outdated system containing this card.

include the complete system specs with the make & model of ALL components.
not just this brief description you've offered.

Sure here's full specs.

Mobo: Asus Z97-Deluxe PCI-Express 3.0 (8.0 GT/s)
CPU: Intel i7-4470 @ 3.40 GHz, 3401 Mhz, 4 Cores, 8 Logical Processors
GPU: Nvidia GeForce RTX 3060 Ti 8GB
SSD: Samsung 850 EVO 500GB
RAM: Kingston 32GB (4 x 8GB) DDR3 933MHz
PSU:Thermaltake Smart M 750w

Note: I was incorrect in my first post about the the RAM config.

 

[Misgar]

You don't mention the PSU. My RTX 3060 (non Ti) pulls up to 170W. Can your PSU cope with the extra load?

 

[sizzlinkola]

You don't mention the PSU. My RTX 3060 (non Ti) pulls up to 170W. Can your PSU cope with the extra load?

I did in my reply above. It's Thermaltake Smart M 750W PSU. Unless you're looking for something more specific about my PSU?

It seems like it can cope? According to GPU-Z, it's at 19-20% power consumption while web browsing and such. While gaming, it's at 30%.

 

[Misgar]

Sorry, I didn't notice you'd already included the PSU spec. My apologies.

If it's the SP-750M PSU tested in this review, it looks like you've made a good choice and it will be more than capable of supplying the RTX 3060Ti.
https://www.techpowerup.com/review/thermaltake-smart-m-750w/

Here is a list of Asus BIOS error codes:
Code B1 = Runtime Set Virtual Address MAP End
https://www.hardwarecentric.com/asus-motherboard-error-codes-list/

This guy had code B1 and discovered his RAM was not fully compatible with his mobo (not a Z97).
https://rog-forum.asus.com/t5/z270/solved-maximus-ix-code-error-q-code-55-b1-cc/td-p/620652

The computer I'm using at this instant in time is as follows:
Gigabyte GA-Z97-D3H mobo and Intel i7-4770K.
Four 4GB (16GB total) Crucial BLS4G3D1609DS1S00 DIMMs at 1600MT/s (DDR3-1600).
Aida64 shows the DIMMs are running at 1.512V (not 1.65V XMP).
Memory Timings are 9-9-9-24 (CL-RCD-RP-RAS) from SPD.
I've applied a medium overclock multiplier of x41 (4.1GHz max) with a manual Vcore voltage set to 1.260V, but Vcore is currently around 0.900V at idle.
The system is stable.
I'm using the Integrated: Intel Haswell-DT GT2 - Integrated Graphics Controller in the i7-4770K, not a PCIe card like your RTX 3060Ti, because my main PCIe slots house an Aquantia AQtion AQC107 10Gbit Network Adapter and a Hewlett Packard SC44Ge Host Bus Adapter for my LTO4 SAS tape drive.

You say your RAM is Kingston DDR3 8GB RAM 933 MHz x 4 (32 GB total RAM).
Am I correct in thinking the 933MHz refers to the true memory clock frequency, with a DDR transfer rate of 2 x 933 =1866MT/s?

If you're running at XMP DDR3-1866, try backing it down to DDR3-1600 or even stock DDR3-1333. As you're probably aware, memory timings often need to be relaxed when you fit four DIMMs instead of two, due to additional load on the IMCs.

With any luck, you won't get any B1 errors at 1333MT/s. If 1600MT/s is not fast enough for your needs and you find 1866MT/s is unstable, try relaxing the CL (CAS) timing by 1 or 2 clock cycles, e.g. if CL=11 at 1866, increase the value to 12 or 13. It might make 4 sticks of RAM more stable.

You might consider running MemTest86 at 1866 to see if the current timings are stable.

 

[sizzlinkola]

Sorry, I didn't notice you'd already included the PSU spec. My apologies.

If it's the SP-750M PSU tested in this review, it looks like you've made a good choice and it will be more than capable of supplying the RTX 3060Ti.
https://www.techpowerup.com/review/thermaltake-smart-m-750w/

Here is a list of Asus BIOS error codes:
Code B1 = Runtime Set Virtual Address MAP End
https://www.hardwarecentric.com/asus-motherboard-error-codes-list/

This guy had code B1 and discovered his RAM was not fully compatible with his mobo (not a Z97).
https://rog-forum.asus.com/t5/z270/solved-maximus-ix-code-error-q-code-55-b1-cc/td-p/620652

The computer I'm using at this instant in time is as follows:
Gigabyte GA-Z97-D3H mobo and Intel i7-4770K.
Four 4GB (16GB total) Crucial BLS4G3D1609DS1S00 DIMMs at 1600MT/s (DDR3-1600).
Aida64 shows the DIMMs are running at 1.512V (not 1.65V XMP).
Memory Timings are 9-9-9-24 (CL-RCD-RP-RAS) from SPD.
I've applied a medium overclock multiplier of x41 (4.1GHz max) with a manual Vcore voltage set to 1.260V, but Vcore is currently around 0.900V at idle.
The system is stable.
I'm using the Integrated: Intel Haswell-DT GT2 - Integrated Graphics Controller in the i7-4770K, not a PCIe card like your RTX 3060Ti, because my main PCIe slots house an Aquantia AQtion AQC107 10Gbit Network Adapter and a Hewlett Packard SC44Ge Host Bus Adapter for my LTO4 SAS tape drive.

You say your RAM is Kingston DDR3 8GB RAM 933 MHz x 4 (32 GB total RAM).
Am I correct in thinking the 933MHz refers to the true memory clock frequency, with a DDR transfer rate of 2 x 933 =1866MT/s?

If you're running at XMP DDR3-1866, try backing it down to DDR3-1600 or even stock DDR3-1333. As you're probably aware, memory timings often need to be relaxed when you fit four DIMMs instead of two, due to additional load on the IMCs.

With any luck, you won't get any B1 errors at 1333MT/s. If 1600MT/s is not fast enough for your needs and you find 1866MT/s is unstable, try relaxing the CL (CAS) timing by 1 or 2 clock cycles, e.g. if CL=11 at 1866, increase the value to 12 or 13. It might make 4 sticks of RAM more stable.

You might consider running MemTest86 at 1866 to see if the current timings are stable.

Thanks for your suggestions here. I was running it at DDR3-1600 originally so I backed down to DDR3-1333.

But I still got the b1 error code after shutting down 3 times. Didn't get the code the first two times, but got it on the third. Still got through fine though after 1-2 minutes.

Here's what CPU-Z says if that helps you gather more ideas:

https://imgur.com/a/XumtrFU

View: https://imgur.com/a/XumtrFU


Appreciate the help!

 

[Misgar]

Here's what CPU-Z says if that helps you gather more ideas:

That looks fine for 666MHz (1333MT/s).

I'd be inclined to give MemTest86 a chance to test both DIMMs. With memory of that age, it's not unknown for it to fail.

I was given more than 70 DDR3 DIMMs pulled from office machines and at least 6 of them failed MemTest86. Some failed instantly, bringing up dozens of red errors. Others took longer to reveal problems.

With your 2 x 16GB DDR3, a complete run of MemTest86 could take 3 to 4 hours. Be patient. Allow the tests to complete at least one full pass.

If there are any errors, remove one DIMM and test again. Repeat for second DIMM on its own. Any errors, consider replacement.

The RAM speed setting affects MemTest's results, e.g. stable at 1333 and 1600MT/s, unstable at 1866MT/s. You'll need to run separate tests at each memory clock speed you intend using, to make sure. Similarly, overclocking the CPU itself can affect stability.

Of course it might be the one of the IMC (Integrated Memory Controller) channels in the CPU, a bad connection in the CPU socket, a motherboard fault, or dust in the DIMM socket.

Check the Voltage applied to the DIMMs:
JEDEC stock is usually 1.50V (or 1.35V for DDR3L).
XMP is usually 1.65V, but I did have some dodgy OCZ which specified 1.70 to 1.90V.

I have an old Phenom 955 which requires a manual setting of 1.60V to work with 4 x 4GB Kingston DIMMs at 1333MT/s, where the BIOS normally sets 1.50V.

Do NOT confuse these voltages with "CPU Vcore". Nasty things will happen if you get things wrong!

 

[sizzlinkola]

With your 2 x 16GB DDR3, a complete run of MemTest86 could take 3 to 4 hours. Be patient. Allow the tests to complete at least one full pass.

If there are any errors, remove one DIMM and test again. Repeat for second DIMM on its own. Any errors, consider replacement.

To clarify, I have 4 x 8 GB. Does that change how you suggest I test?

 

[Misgar]

To clarify, I have 4 x 8 GB. Does that change how you suggest I test?

Almost the same as testing 2 x 16GB in MemTest86, but if the first test with all four DIMMs installed gives any errors, remove three DIMMs and test each DIMM singly, to find out which one(s) are faulty.

 

[sizzlinkola]

Almost the same as testing 2 x 16GB in MemTest86, but if the first test with all four DIMMs installed gives any errors, remove three DIMMs and test each DIMM singly, to find out which one(s) are faulty.

MemTest86 did not find any errors. Passed everything. Any other ideas?

 

[Misgar]

MemTest86 did not find any errors. Passed everything. Any other ideas?

Yes, but unless you have a large box full of spares, my ideas won't be of much use.

If it was my system, I'd start swapping components, one by one, until the fault goes away.

I'd start with another GPU card, because you say your problems started when you installed the RTX 3060. Then I'd change the PSU, then the i7-4770 for something similar, and finally the motherboard.

Changing an obsolete mobo is a pain. It's a major component and it can take some time to purchase a good, fully working replacement.

Along the way, if I have any doubts about a component, I swap it over into a similar system, e.g. if I think the CPU might be the culprit, I'd try it in another machine.

Second hand LGA1150 CPUs are cheap until you reach the 4770K (I paid around $25 for one on eBay). I'd pay less for the non-K i7-4770, but an i7-4790K can be expensive.

I'm sorry if you don't have an infinite supply of parts, but it's an excuse for not getting rid of things when upgrading, if they still work.

Spend (another) hour or two searching online for people with the same or similar faults. You might get lucky.

 

[sizzlinkola]

Yes, but unless you have a large box full of spares, my ideas won't be of much use.

If it was my system, I'd start swapping components, one by one, until the fault goes away.

I'd start with another GPU card, because you say your problems started when you installed the RTX 3060. Then I'd change the PSU, then the i7-4770 for something similar, and finally the motherboard.

Changing an obsolete mobo is a pain. It's a major component and it can take some time to purchase a good, fully working replacement.

Along the way, if I have any doubts about a component, I swap it over into a similar system, e.g. if I think the CPU might be the culprit, I'd try it in another machine.

Second hand LGA1150 CPUs are cheap until you reach the 4770K (I paid around $25 for one on eBay). I'd pay less for the non-K i7-4770, but an i7-4790K can be expensive.

I'm sorry if you don't have an infinite supply of parts, but it's an excuse for not getting rid of things when upgrading, if they still work.

Spend (another) hour or two searching online for people with the same or similar faults. You might get lucky.

Thanks for the help. May take it to a shop but not sure if it's worth the $$$ to fix this issue if it happens intermittently. And even when it does happen, I just need to wait another 1-2 minutes which is completely fine.

I'm more concerned if this causes any long-term damage. If it's just convenience, then I'm fine with it.

 

[Misgar]

I'm more concerned if this causes any long-term damage.

By long term damage, is the part you're most cencerned about the RTX 3060? The other components could be replaced/upgraded for less than the cost of the GPU.

You could probably buy a second hand mobo/CPU bundle of the same vintage on eBay for less than $50. I picked up an old working AMD FM2 mATX mobo with low-end A4 APU for $6. I upgraded to an A8 APU and used it in an experimental server build.