More than 10 years since it's last update Furmark is getting a v2 very soon!

[revenant]

https://www.tweaktown.com/news/9331...utm_source=newsletter&utm_campaign=newsletter

After 16 years Furmark is finally getting an update.. v2 is almost here... but if you just can't wait I grabbed this off the Geeks3D discord ...

FurMark 2.0.10 is available. Maybe the last version before public release ?

FurMark win64
https://www.geeks3d.com/dl/show/705/UEEYzxyKzVp6ayq

FurMark win32
https://www.geeks3d.com/dl/show/708

Linux version will be updated the first week of September...

Changes:
Version 2.0.10.0 - 2023.08.26
fixed Radeon RX 6850M XT name (XT was missing).
added support of AMD Radeon PRO W7900, PRO W7800,
PRO W7600 and PRO W7500.
added support of AMD Radeon RX 7900 GRE.
added support of NVIDIA GeForce RTX 4060 Ti 16GB
added support of Intel Arc A570M, Arc A530M, Arc Pro A60M and A30M.
(windows) updated with GPU Shark2 2.0.12
updated with GeeXLab 0.53.0 libs


Enjoy!

 

[Zarathustra[H]]

While once a very important tool to test for stability, I thought the overclocking community had dismissed Furmark as an unrealistic test, or "power virus" over a decade ago, with both AMD and Nvidia rendering it a pointless test, because the drivers detect it and detune the card while it is running...

Or am I misremembering?

I used a ton for overclocking stability testing years ago, but haven't touched it in many many years.

Why is it still around?

 

[revenant]

Oh.. I posted this being uninformed of that. I thought some people here might find this interesting. Maybe v2 will be better? idk.. I guess we just this roll off and organically die.

edit.. it could be useful to simulate max power draw and how thermals behave.

 

[riev90]

edit.. it could be useful to simulate max power draw and how thermals behave.

but still I wouldn't use Furmark to test my GPU stability just because of that, or what's usually called Power Virus.
the last time I used Furmark for testing purpose was before I sent my Powercolor RX 480 8gb to buyer (2018) and never used it again.

 

[Mad Maxx]

What I remember most about Furmark was the end credits to one of the mid-2000 versions. It was the first time I ever heard a song from Poets of the Fall. It was called Lift and it turned me into a huge fan of the group. I always liked that test video of space combat, too. Reminded me of the opening scene from the original Star Wars.

 

[Mazzspeed]

Where do you get the Linux version from? I didn't even know there was a Linux version.

 

[NattyKathy]

While once a very important tool to test for stability, I thought the overclocking community had dismissed Furmark as an unrealistic test, or "power virus" over a decade ago, with both AMD and Nvidia rendering it a pointless test, because the drivers detect it and detune the card while it is running...

Or am I misremembering?

I used a ton for overclocking stability testing years ago, but haven't touched it in many many years.

Why is it still around?

You are remembering right. On any card with enforced power limit it'll just throttle and result in a weird, unrealistic, low-clock low-voltage high-current state. And if there's no power limit, well, better hope you have really good VRM cooling or maybe kaboom. Not much utility in benchmarking or V/F stability testing with Furmark.

The "power virus" thing does make Furmark a useful tool tho for GPU modders, undervolters, Watercooling People, xoc'ers etc. Can be used to test loop heat capacity, worst-case VRM behavior, predict magnitude of transients, stuff like that.

I think things have kinda come around from Furmark being ridiculed to it finding a niche for the thing it was/is ridiculed for.

 

[Zarathustra[H]]

You are remembering right. On any card with enforced power limit it'll just throttle and result in a weird, unrealistic, low-clock low-voltage high-current state. And if there's no power limit, well, better hope you have really good VRM cooling or maybe kaboom. Not much utility in benchmarking or V/F stability testing with Furmark.

The "power virus" thing does make Furmark a useful tool tho for GPU modders, undervolters, Watercooling People, xoc'ers etc. Can be used to test loop heat capacity, worst-case VRM behavior, predict magnitude of transients, stuff like that.

I think things have kinda come around from Furmark being ridiculed to it finding a niche for the thing it was/is ridiculed for.

Yeah, I think there are two things that changed the general relevance of Furmark.

One as you say was the introduction of GPU power limits and the introduction of dynamic clocks, but the other was also a shift away from temperature being the universal limiting factor for overclocks, both on GPU and CPU.

There was once a time when thermals were like 95% of the problem, and if you could just keep the CPU/GPU cool enough you could probably get s higher clock out of it.

These days, with process size shrinking, thermals are less abd less of a limiting factor, which has shifted the limiting factor over to silicon quality, something which a thermal stress test like this doesn't really test for.

I see what you are saying about it being a niche test for those who have to stress test thermals, but I'd still argue that it is an unrealistic stress test, because Furmark produces an amount of heat that is wholly unrealistic in real use, and thus it is not a relevant stress test, even in thermally limited scenarios.

 

[sharknice]

I thought this said FutureMark at first. But yeah I remember trying to test my OC with this then gave up and just used the games I played.

 

[Zarathustra[H]]

I thought this said FutureMark at first. But yeah I remember trying to test my OC with this then gave up and just used the games I played.

I used to use 48 hours of Furmark to stability test my GPU overclocks, and 48 hours of Intel Burn Test to stability test my CPU.

If it passed these tests, it would be rock stable.

By modern standards it is overkill, but back then it was appropriate.