New 3DMark Sky Diver Benchmark Available

[btarunr]

Hot on the heels of all the hardware announcements at Computex, our new 3DMark Sky Diver benchmark test is ready to download and use today. Every single 3DMark user - more than a million and counting - will get Sky Diver as a free update. For new users, Sky Diver is unlocked and ready to use in all editions of 3DMark. And for a limited time, you can buy 3DMark Advanced Edition, which includes more tests, custom settings and other features, from Steam for only $9.99 (60% off).

Sky Diver is a new DirectX 11 benchmark test for gaming laptops and mid-range PCs. It's ideal for testing mainstream graphics cards, mobile GPUs, integrated graphics and other systems that cannot achieve double-digit frame rates in the more demanding Fire Strike test.



DOWNLOAD: 3DMark v1.3.708 with Sky Diver

Together, Sky Diver and Fire Strike let you test the full range of DirectX 11 graphics hardware. Fire Strike is equivalent to testing a system with a modern DirectX 11 game on ultra-high settings. Sky Diver is more like running a game on normal settings. As a general guide:

• If a system scores less than 2800 in Fire Strike you should run Sky Diver.
• If a system scores more than 12000 in Sky Diver, you should run Fire Strike.

Please note that Sky Diver and Fire Strike scores are not directly comparable.

3DMark Sky Diver benchmark tests in brief
Sky Diver includes a Demo, two Graphics tests, a Physics test and a Combined test. The Graphics tests measure GPU performance, the Physics test measures CPU performance, and the Combined test stresses both GPU and CPU. The Demo does not affect the score.

Graphics test 1 focuses on tessellation and uses a forward lighting method. Graphics test 2 focuses on pixel processing and uses compute shader-based deferred tiled lighting. The Physics test introduces a new approach that extends the performance range for which the test is relevant. The test runs through four levels of work starting with the lightest and continuing to the heaviest unless the frame rate drops below a minimum threshold. The Combined test contains both graphics workloads and physics simulations to stress the CPU and GPU. The test uses a compute shader-based deferred tiled lighting method.

View at TechPowerUp Main Site

 

[BazookaJoe]

Score : 27,193 (Just the Sky Diver Test)

Is this ok?

 

[MxPhenom 216]

i Cant even get it to run for me. Its just a black screen. Fire strike works fine.

 

[FM_Jarnis]

Sky Diver Game Test 1 or Combined Test is showing black screen with just the frame rate counter showing? I have a NVIDIA GPU.

NVIDIA issued the following statement on this issue.

"While testing the 3DMark Sky Diver benchmark, NVIDIA discovered an issue where some of their GPU configurations would render a black screen during benchmarking. NVIDIA has since solved the issue for these configurations and will be including the fix with their next driver instalment coming early next week. In the interim, if you experience this issue with your configuration please use the 335.23 WHQL driver. This driver is available at both NVIDIA.com and GeForce.com."

Desktop 335.23 WHQL:

• Win7/8.1 x32 - http://www.nvidia.com/download/driverResults.aspx/73779/en-us
• Win 7/8.1 x64 - http://www.nvidia.com/download/driverResults.aspx/73780/en-us

Notebook 335.23 WHQL:

• Win7/8.1 x32 - http://www.nvidia.com/download/driverResults.aspx/73783/en-us
• Win 7/8.1 x64 - http://www.nvidia.com/download/driverResults.aspx/73784/en-us

 

[d1nky]

will test this fully at the weekend, but will use 247 clocks in a bit

 

[oli_ramsay]

Nice looking demo, performs way better than fire strike

Got 22,220 with 7970@1170 core, 1725 mem and 3770 @ 4.4ghz

http://www.3dmark.com/sd/2132425

 

[TheHunter]

i Cant even get it to run for me. Its just a black screen. Fire strike works fine.

btw its only 1st part of the benchmark - gpu usage etc is maxed and fps meter running, 2nd part is fine. Also whole demo scene is ok too.


780gtx gpu boost 1228mhz, cpu 4.7ghz

http://www.3dmark.com/3dm/3241141?

 

[MxPhenom 216]

Scores (No CPU overclock on either):
Also this is using the bugged 337.88 drivers.

GPU at stock 1137mhz Skyn3t BIOS from OCN

GPU at 1293mhz core @ 1.24v in Afterburner ~1.19v load no LLC mod

 

[Lopez0101]

Guess I'll have to wait for a patch. Refuses to run an my 290.

 

[GreiverBlade]

Guess I'll have to wait for a patch. Refuses to run an my 290.

well i just did a run with my actual main rig ... take a deep breath ...
SKY DIVER
Add to compare
VALID RESULT
SCORE4534 with AMD Radeon R9 290(1x) and AMD Opteron 248
Graphics Score
14503
Physics Score
1190
Combined Score
2337
SHOW RESULT DETAILS
RUN DETAILS

Name
Description
AMD Radeon R9 290
Vendor
Asustek Computer, Inc.
# of cards
1
SLI / CrossFire
Off
Memory
4,096 MB
Core clock
947 MHz
Memory bus clock
1,250 MHz
Driver name
AMD Radeon R9 200 Series
Driver version
14.100.0.0

[paste:font size="4"]AMD Opteron 248
Reported stock core clock
2,200 MHz
Maximum turbo core clock
2,200 MHz
Physical / logical processors
2 / 2
# of cores
1
Package
Manufacturing process
0 nm
TDP
85 W
[paste:font size="4"]GENERAL
Operating system
64-bit Windows 7 (6.1.7601)
Motherboard
Supermicro H8DCE
Memory
4,096 MB
Module 1
1,024 MB Kingston DDR @ 133 MHz
Module 2
1,024 MB Kingston DDR @ 133 MHz
Module 3
1,024 MB Kingston DDR @ 133 MHz
Module 4
1,024 MB Kingston DDR @ 133 MHz
Hard drive model
120 GB ST3120022A ATA Device

yep you can laugh a R9 290 in a 2003 dual opteron single core 4gb 266 DDR workstation

 

[RejZoR]

The latest 3DMark benchmarks are so disappointing i didn't even bother to run the test after Firestrike demo. First few tests were running at 4 digit framerates and the firestrike looked below average at best, yet it was running rather rubbish. Sorry, but i've seen games from few years ago that looked way more impressive in realtime as a proper game. Is churning billions of triangles for some stupid spikes on a magma soldier really all that impressive and important considering you never really have the chance to actually look at them with all the action going on? It's the whole reason i've been raging about it for ages ever since they introduced tessellation. Pointless wasting of compute power on stuff no one can even see instead of using it the other way around and making things run ridiculously fast on any hardware and still look just as good. But no, tessellate all objects to every possible limit and throw smooth framerate out the window for the sake of "technology". Pointless and idiotic. Has no one educated the damn developers how to use this thing? LOD was there years ago to increase performance on objects that are so far away that you can't tell the difference. It wasn't meant to pop in super duper detailed models over normal ones, it was to remove stuff that's wasting compute power.

I can easily say that Firestrike could run at 100+ fps with proper use of tessellation and pretty much no one could tell the difference in details quality.

Oh and Skydiver looked poor as well. I've seen better DoF in Crysis and less washed out rocks in so many other games. If you're setting a benchmark, at least try to drop my jaw with something impressive and not make my jaw drop because of yawning... like for example super impressive wavy water in 3DMark2000 at the end without a single used Pixel Shader. Or that wobbly quicksilver thingie or the Max Payne destroying the hallway. That was friggin impressive back in 2001. Or the insane beautiful nature with millions of grass strains and leaves on trees weaving around in wind. It was jaw dropping back then and still is whenever i see it. Way more than anything i've seen in 2013 or 2014...

Back to the drawing board Futuremark, you failed to impress me...

 

[Lopez0101]

I kind of agree. I don't think the recent 3DMarks are as impressive as they used to be. I can't remember which one it was, but the aerial dog figt with the bombers was awesome, along with the Sci-fi elevator and shootout thing, or even the floating ship with the dragon/leviathan thing.

 

[GhostRyder]

Wow can't wait to get home and try this, looks awesome.

 

[MxPhenom 216]

I kind of agree. I don't think the recent 3DMarks are as impressive as they used to be. I can't remember which one it was, but the aerial dog figt with the bombers was awesome, along with the Sci-fi elevator and shootout thing, or even the floating ship with the dragon/leviathan thing.

Why should they be impressive? They are a benchmark not a game.

Also for the 2 other benchmarks other than firestrike and this one, futuremark even says they are either meant to run on mid range systems or APUs or mobile devices. Firestrike is the only one they have released as of late for high end systems, then if you run it in extreme mode, its more for multi gpu systems.

 

[ZoneDymo]

Why should they be impressive? They are a benchmark not a game.

Also for the 2 other benchmarks other than firestrike and this one, futuremark even says they are either meant to run on mid range systems or APUs or mobile devices. Firestrike is the only one they have released as of late for high end systems, then if you run it in extreme mode, its more for multi gpu systems.

Well it does not HAVE to look impressive, but if its meant to task your system it should atleast look incredible right? I mean why else would it be so hard to run.
We/I want to see visuals we can't properly do yet in real time with current hardware.

 

[fullinfusion]

I always welcome new *fresh* benchmarking programs from 3DMark

Looks pretty good to me..

 

[RejZoR]

Why should they be impressive? They are a benchmark not a game.

Also for the 2 other benchmarks other than firestrike and this one, futuremark even says they are either meant to run on mid range systems or APUs or mobile devices. Firestrike is the only one they have released as of late for high end systems, then if you run it in extreme mode, its more for multi gpu systems.

Nonsense. If you want to push the envelope and showcase something breathtaking, something worth running at friggin 2fps on quad GFX configuration, you have to justify it by making it look brutally awesome. If it's just a synthetic benchmark, why would they even bother with all the visuals? Just make a progress bar and throw billions of triangles and textures and shader effects in the background and churn out the score at the end.

And when you see something run at 25fps on a high end graphic card and you think, this is totally meh, they are doing something wrong, because i know i was actually playing games that looked better and ran faster on hardware released 2 years ago. And current 3DMark is facing that problem. They haven't showed us anything that we haven't seen in an actual working game with perfectly playable framerate. TES:Skyrim is the first that springs to mind. It might not be exactly the same, but then again it wasn't released yesterday so it's sort of in sync with what i'm complaining about. And TES even has those flame atronach creatures that look very similar. Every older 3DMark was jaw dropping, because the sceneries were really an insane challenge for even the very latest hardware and got really fluid after 2 more egnerations of graphic cards. And they looked really impressive. This is neither. Despite the fact it was just released it was pretty much smooth on a HD7950 graphic card. Something it was released like ages ago and i'm not even using it in crossfire setup. It's not really pushing hardware much then, is it?

 

[TheHunter]

Ok I was testing mid high gpu/cpu OC and got this looked funny

http://www.3dmark.com/sd/2188981

 

[Prima.Vera]

26860 with no occ.
http://www.3dmark.com/3dm/3418214

 

[usmc362]