Geekbench 3.4.3

[jfpoole]

We're working on a new Geekbench 3 release that addresses issues that affect the overclocking community.  The first preview release, Geekbench 3.4.3 Beta 1, is now available for download:

https://geekbench.s3.amazonaws.com/Geekbench-3.4.3b1-WindowsSetup.exe

Geekbench 3.4.3 Beta 1 fixes the issues that prevented Geekbench 3 from working on recent AMD and Intel HEDT processors.

Geekbench 3.4.3 Beta 1 also fixes several issues that prevented Geekbench 3 from being used on AMD systems running Windows 10 on HWBOT.  

Give it a try and let me know what you think.
 

[Rauf]

Appreciate your efforts. But I think you should open up to the possibility to use geekbench with Benchmate. There is no future for Geekbench as a competitive benchmark without some sort of Benchmate integration. 

[Leeghoofd]

Let us test and screen first Tobias before drawing early conclusions

thx for the update mr Poole

[Rauf]

I was referring to this. I'd be sad to see geekbench gone from globals and would like to see a solution.

 

[Leeghoofd]

If they can fix some stuff on Matt's recommendations, maybe we don't need to go to this step

[l0ud_sil3nc3]

23 hours ago, jfpoole said:

We're working on a new Geekbench 3 release that addresses issues that affect the overclocking community.  The first preview release, Geekbench 3.4.3 Beta 1, is now available for download:

https://geekbench.s3.amazonaws.com/Geekbench-3.4.3b1-WindowsSetup.exe

Geekbench 3.4.3 Beta 1 fixes the issues that prevented Geekbench 3 from working on recent AMD and Intel HEDT processors.

Geekbench 3.4.3 Beta 1 also fixes several issues that prevented Geekbench 3 from being used on AMD systems running Windows 10 on HWBOT.  

Give it a try and let me know what you think.
 

Appreciate the update.

I know we (hwbot) are a niche group and really a minuscule demographic in the grand scheme of things, regardless it's cool to see this updated and will test it out.

Thanks again.

[_mat_]

I appreciate the update as well. A benchmark developer showing interest to work together with HWBOT and the overclocking community should always be appreciated. In my opinion it's an absolute necessity to cooperate to be able to keep the benchmark in the ranks long-term.

On 2/12/2020 at 6:29 AM, jfpoole said:

Geekbench 3.4.3 Beta 1 also fixes several issues that prevented Geekbench 3 from being used on AMD systems running Windows 10 on HWBOT.

May I ask what the specific changes are?

21 hours ago, Leeghoofd said:

If they can fix some stuff on Matt's recommendations, maybe we don't need to go to this step

It is impossible to do this in userland applications. Timers can not be verified correctly by using WIN32 API functions alone, direct access to hardware devices is needed (a kernel driver). The only workaround is to enforce HPET system-wide like I introduced with GPUPI years ago when I had no other solution. I doubt that this is the right way for Geekbench as a mainstream benchmark.

And even if a userland solution is found, 30 seconds with Cheat Engine are enough to change the timers to your liking. I've been there and if you think this through this is not the way to go with competitive benchmarks.

My recommendation is to integrate Geekbench into BenchMate again, which solves all problems at once. I'm up for it, because in the end it's the best for overclocking and the community to not remove points from Geekbench if not necessary.

[keeph8n]

On 2/12/2020 at 12:29 AM, jfpoole said:

Geekbench 3.4.3 Beta 1 fixes the issues that prevented Geekbench 3 from working on recent AMD and Intel HEDT processors.

 

I never saw an issue with GB3 running on 3970X or 10900X/10980XE? Is there a particular case that wouldn't allow GB3 to be run? 3990X perhaps?

[jfpoole]

 

6 hours ago, _mat_ said:

My recommendation is to integrate Geekbench into BenchMate again, which solves all problems at once. I'm up for it, because in the end it's the best for overclocking and the community to not remove points from Geekbench if not necessary.

I disagree -- I think this is something we should fix in Geekbench itself, as then these fixes benefit all Geekbench users.

23 minutes ago, keeph8n said:

I never saw an issue with GB3 running on 3970X or 10900X/10980XE? Is there a particular case that wouldn't allow GB3 to be run? 3990X perhaps?

I don't know the full list of processors, but the 3990X (along with the W-3175X) was mentioned explicitly by @Splave.

[keeph8n]

Ah okay. Good to know as I have a 3990X in route

[_mat_]

33 minutes ago, jfpoole said:

I disagree -- I think this is something we should fix in Geekbench itself, as then these fixes benefit all Geekbench users.

How?

[keeph8n]

I would like to see GB and BM work together, as I think it would benefit HWBOT.

 

I do think GB should be fixed and correct for ALL users though

 

[yosarianilives]

Sounds like a lot can be fixed in user space app, which should be done then. And some requires a service like bm. So the real question is this, what's stopping geek from integrating/licensing benchmate services into their app? You have to install the app anyways so if it would need a service to be properly secure then how about it runs a service that makes it properly secure? 

[_mat_]

I got around to test the new version.

Ryzen 2700X, Windows 10

Bclock 104 MHz, 3744 MHz
Mem: 3198 MHz

Score: https://browser.geekbench.com/geekbench3/8844471

SC 3830
MC 30513
__________________________________________

Bclock 100 MHz, 3750 MHz
Mem: 3065 MHz

Score: https://browser.geekbench.com/geekbench3/8844479

SC 3961
MC 30575

My conclusions:

1. APIC Timer drift still skews the score. This is also true for all Intel CPUs below Skylake.
2. Neither CPU nor memory frequency is detected correctly.
3. <inserttoolhere> will get you a world record on air by hijacking API calls.

Edited February 13, 2020 by _mat_
Removed the name of the tool that fucks things up.

[jfpoole]

How did you run Geekbench? On our test systems that drastic a change to BCLK would’ve been flagged. 

[_mat_]

I opened it and clicked the benchmark button. Changed bclock multiple times, reran the bench each time.

[Splave]

so this is something that benchmate would catch?

Just thinking out loud.

• We still need to pay for key to unlock 64 bit geek3 to use it with benchmate 
• A valid link is still required from primatelabs website to be a valid score, so nothing is changing there
• We have extra protection via benchmate at no cost to primatelabs and they dont have to mess around paying people man hours with EOL geekbench3 anymore.

Seems it would be a win for all parties? Please correct me if im wrong.

 

I dont want to see geekbench removed. I think its an excellent benchmark. Very difficult on thread ripper and a nice challenge. Memory and fabric making a huge difference etc.

Edited February 13, 2020 by Splave

[Noxinite]

Not every bench needs to be integrated in to BM, especially not if the creator requests that it isn't.

The ideal solution would be AMD to fix the timer issue like Intel did, but that probably won't happen...

[_mat_]

16 minutes ago, Noxinite said:

The ideal solution would be AMD to fix the timer issue like Intel did, but that probably won't happen...

This is out of our control and will always be. We have to focus on the things we can control. Also there will always be the issues of the faulty CPUs/OS versions in between. More confusion to add to the whole disaster, great.

Anyways, BenchMate will go its way, with our without Geekbench. This can be much bigger than a single bench, no doubt about that.

Edited February 13, 2020 by _mat_
Read it wrong.

[Leeghoofd]

1 hour ago, Noxinite said:

Not every bench needs to be integrated in to BM, especially not if the creator requests that it isn't.

The ideal solution would be AMD to fix the timer issue like Intel did, but that probably won't happen...

Benchmate is a must to keep the database clean of bugged scores. Especially on topic here: Matt has analysed how Geekbench ticks and there's some internal things that need to be monitored and adressed. II is great news that the Geekbench ceveloppers still want to invest time in Geekbench3 and address some of the detected flaws done by Matt and the community. But we are not there yet...

AMD will NEVER fix their issue, surely not on short temr, If they wanted it would already have been done ages ago. Other point is we can't rely on Windows 7 forever. Installing older OSses is a real pain. Win10 also brings huge performance benefits and new users too will probably not downgrade to have legit fun!!!

The road ahead is (fingers crossed) that the Geekbench development team and Matt can work out a gentlemens agreement on futher allowing the OC community to use both products. Allow us to dream that they maybe join forces to improve the current and future products of both parties involved.

 

But this communication has to happen without any interference of any other external parties or members. They are talking again, which is already a big bonus for HWBot and its users. Now a simple innocent quote like yours Noxinite can lead to a total escelation of the topic. Something which needs to be avoided at all costs. So plz leave the thread to the ones that can make it better. Just provide info on findings when using the new Geek Beta version.But there will  be no product bashing nor your personal vision on the matter is required here! This is for all users, not for a single individual!

 

Like Allen said, most of the benchers want to keep on using Geekbench 3/4/5, but also agree it has to be secured, thus avoiding  that we don't run into the same issues as before (especially on AMD)

 

Fingers crossed these gents can make it happen!!!!

 

 

 

[jfpoole]

1 hour ago, Splave said:

• We have extra protection via benchmate at no cost to primatelabs and they dont have to mess around paying people man hours with EOL geekbench3 anymore.

For what it's worth we still sell and support Geekbench 3.  For the vast majority of users we recommend Geekbench 5 as it has much more modern benchmark tests, but if people hit bugs with Geekbench 3 we'd like to hear about it.

[l0ud_sil3nc3]

3 hours ago, Leeghoofd said:

Benchmate is a must to keep the database clean of bugged scores. Especially on topic here: Matt has analysed how Geekbench ticks and there's some internal things that need to be monitored and adressed. II is great news that the Geekbench ceveloppers still want to invest time in Geekbench3 and address some of the detected flaws done by Matt and the community. But we are not there yet...

AMD will NEVER fix their issue, surely not on short temr, If they wanted it would already have been done ages ago. Other point is we can't rely on Windows 7 forever. Installing older OSses is a real pain. Win10 also brings huge performance benefits and new users too will probably not downgrade to have legit fun!!!

The road ahead is (fingers crossed) that the Geekbench development team and Matt can work out a gentlemens agreement on futher allowing the OC community to use both products. Allow us to dream that they maybe join forces to improve the current and future products of both parties involved.

 

But this communication has to happen without any interference of any other external parties or members. They are talking again, which is already a big bonus for HWBot and its users. Now a simple innocent quote like yours Noxinite can lead to a total escelation of the topic. Something which needs to be avoided at all costs. So plz leave the thread to the ones that can make it better. Just provide info on findings when using the new Geek Beta version.But there will  be no product bashing nor your personal vision on the matter is required here! This is for all users, not for a single individual!

 

Like Allen said, most of the benchers want to keep on using Geekbench 3/4/5, but also agree it has to be secured, thus avoiding  that we don't run into the same issues as before (especially on AMD)

 

Fingers crossed these gents can make it happen!!!!

 

 

 

The voice of reason coming through clutch

[jfpoole]

Geekbench 3.4.3 Beta 2 is now available for download:

https://geekbench.s3.amazonaws.com/Geekbench-3.4.3b2-WindowsSetup.exe

Beta 2 improves the in-app validation routines, and makes it far more obvious whether Geekbench considers a result valid (or invalid or questionable) due to timer issues.  Beta 2 will still let users view invalid results as well as upload them to the Geekbench Browser.  The Browser still needs to be updated to display the validation status of the results, but that should be straightforward once the client code is working properly.

Give it a try and let me know if you hit any issues.

[Splave]

Hello this version doesn't run at all on 64/128,tr 

Dll swap still works on old 3.4.2 so not the os. 

 

Thanks

[jfpoole]

@Splave, it may take a while for the benchmark to initialize on TR3 systems (it's proportional to the number of hardware threads).  How long did you let Geekbench sit at the "Benchmark Progress" dialog?