• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Intel Arc A730M 3DMark TimeSpy Score Spied, in League of RTX 3070 Laptop GPU

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
47,670 (7.43/day)
Location
Dublin, Ireland
System Name RBMK-1000
Processor AMD Ryzen 7 5700G
Motherboard Gigabyte B550 AORUS Elite V2
Cooling DeepCool Gammax L240 V2
Memory 2x 16GB DDR4-3200
Video Card(s) Galax RTX 4070 Ti EX
Storage Samsung 990 1TB
Display(s) BenQ 1440p 60 Hz 27-inch
Case Corsair Carbide 100R
Audio Device(s) ASUS SupremeFX S1220A
Power Supply Cooler Master MWE Gold 650W
Mouse ASUS ROG Strix Impact
Keyboard Gamdias Hermes E2
Software Windows 11 Pro
Someone with access to a gaming notebook powered by Intel Arc "Alchemist" A730M discrete GPU posted its alleged 3DMark TimeSpy score, and it looks pretty interesting—10.138 points, which is somewhat higher than that of the GeForce RTX 3070 Laptop GPU or halfway between those of the desktop GeForce RTX 3060 and desktop RTX 3060 Ti.

Based on the Xe-HPG graphics architecture, the Arc A730M features 24 Xe Cores, or 384 execution units, which work out to 3,072 unified shaders. This is not even Intel's most powerful mobile GPU, with that title going to the A770M, which maxes out the ACM-G10 ASIC, with all 512 execution units (4,096 unified shaders) being enabled. It particularly raises hopes for a competitive high-end GPU for gaming notebooks, which can perform in the league of the RTX 3080 Laptop GPU, or the Radeon RX 6800M.



View at TechPowerUp Main Site | Source
 
I will hold all judgement and emotions until TPU's review is posted up.
 
Not quite a 3060Ti (desktop) though is it? https://www.3dmark.com/spy/28376992 i sit at 12341 Timespy Score on a 3060Ti Desktop GPU. Close but no cigar i feel Intel are going to be close but not close enough with these GPU's but we shall see.
 
Can't really say its good or bad since there's no real metric of comparison and not enough data really, especially for the mobile side.
 
My 2070 Super with i7 12700K scores about 800 points higher:
 
There are no game benchmarks, because there are no working drivers, but as you can see, Intel is trying to optimize their 3dmark driver as much as possible
 
My 2070 Super with i7 12700K scores about 800 points higher:
Actually, your Graphic score is very in line with the mobile GPU... your CPU score is like 70% higher...
 
Can't really say its good or bad since there's no real metric of comparison and not enough data really, especially for the mobile side.

If you compare it to their previous graphics performance, it's about 3000% improvement, so not terrible.

Im sure in games it's a stuttery mess from weak drivers.
 
Intel Arc (DG2-384) has gone from "Faster than a desktop 3070Ti" to "in league with a mobile 3070"
Just to remind people, that's 6144 cores at 1770MHz vs 5120 cores at 1290MHz, resulting in about 40% less performance.

I'm also wary of 3DMark scores. Intel Iris Xe laptop GPUs were excellent in 3DMark - outperforming Vega8 IGPs in the Renoir Ryzen 7 4800HS and Ryzen 9 4900HS and but sorely under-delivered in actual games and rendering applications, barely matching the low-power Vega5 and Vega6 in the TDP-restricted Ryzen 3 4300U and Ryzen 5 4500U respectively. I bought one for work and ignoring the games that needed driver fixes (not too many and only one I tested with it - Forza) it was overhyped and underwhelming. On the plus side it was an improvement on the previous generation of terrible Intel graphics.
 
How many eaten words in this thread i wonder. or i was wrongs?
 
Actually, your Graphic score is very in line with the mobile GPU... your CPU score is like 70% higher...
Not bad for a laptop, also it's not the highest tier yet, however my GPU isn't really new anymore and intel will have to fight soon with new gen GPU's.....
 
How many eaten words in this thread i wonder. or i was wrongs?

Somehow they will twist it, that you took their quote out of context or if you min/max/oc, it's better at niche thing somehow makes them right. :)
 
Read all your new comments. I believe we all are on the same page. Still no harmony between hardware to software in terms of games. Games are no doubt optimised for nvidia and amd. Newer titles might benefit from Arc gfx yet as before these gpus might suffer from driver support.
If the 3dmark can score good by proper driver tweaks I believe gaming benchmarks should come out real soon as well.
 
I'm guessing this was achieved using a driver that changes draw calls to no-ops.
 
By the time they release these cards nVidia and AMD would have released their next gen GPUs plus they can afford to reduce the prices for their previous GPUs.
 
3D Mark results is probably one of the worst and useless in my opinion. As usual, 3D Mark bench have this odd CPU physics test that is very out of place and significantly skew the final result. CPU is certainly important, but that is subjected to how GPU bound we are. So in this case, I am not surprised that the Alder Lake mobile processor actually contributed quite a fair bit to the score, unless the tester was using the same processor in the testing. In addition, all GPU drivers are heavily optimised to deliver their best in these short and predictable benchmarks. Just like Intel’s XE iGPU where they score really well in 3D Mark, but is not as impressive in actual games.
 
3D Mark results is probably one of the worst and useless in my opinion. As usual, 3D Mark bench have this odd CPU physics test that is very out of place and significantly skew the final result. CPU is certainly important, but that is subjected to how GPU bound we are. So in this case, I am not surprised that the Alder Lake mobile processor actually contributed quite a fair bit to the score, unless the tester was using the same processor in the testing. In addition, all GPU drivers are heavily optimised to deliver their best in these short and predictable benchmarks. Just like Intel’s XE iGPU where they score really well in 3D Mark, but is not as impressive in actual games.

The only GPU test that 3dmark released and that actually measured the GPU performance, independent if you had a sempron, pentium 4, a Core2duo, a Athlon 64 was 3dmark nature (3dmark 2001).

It was the perfect scene for overclocking contests. I guess they just tweak the driver to the point to create a false impression of it's performance. In games its a totally different image so far. Drivers or not, these are discarded compute chips afterall.

And what was Vega when raja released that thing? A discarded compute card.
 
Back
Top