• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

A10-7850K Performance Review

They're not a review website but they're a system builder... so I guess they're receive official sample and they're not under NDA
 
Yeah I caught in the comments that AMD gave them explicit permission to publish that around launching pre-orders.
 
Maybe not...

Despite never communicating so to us, it has become clear that AMD intends for performance data to be withheld from the public until January 14th. In good faith we have proactively decided to take down this review until that time. Sorry everyone!
Click to expand...
 
Screen shots or other places where the review can still be read?
 
Wonder if someone sent Steve the benchmarks anonymously? You know. For science.
 
No one read it that can comment on what the review said, perhaps a few details remembered?
 
Image album: http://imgur.com/a/XVob4


"In fact, this performance bump means that the A10-7850K is actually able to match some low-end discrete graphics cards. In our testing, we found that it performs almost exactly between the AMD Radeon R7 240 and the NVIDIA Geforce GT 640.

Conclusion

Overall, the AMD A10-7850K is a great improvement over the previous generation APUs in AMD's A-series line. Especially in terms of gaming performance when using the integrated graphics, it is about 10% faster than the previous generation APUs. If you compare the gaming performance to an Intel CPU, the graphical capability is so much better that it really is not even in the same league.

In terms of CPU performance, the A10-7850K is overall an improvement over the A10-6800K. Especially in PCMark 08, the A10-7850K was about 10% faster than the A10-6800K. Our GeekBench benchmark indicated lower single-core performance, but as more and more programs use multi-core technology, that is really not even much of an issue."
 
Opening pre-orders without releasing quantitative data is hella shady anyway. Especially when the CPU performance is a null move from richland.
 
Also the Intel CPU was much more powerful at being a "pure CPU" using the testing suite in the review. So if you're going to toss in a dedicated graphics card, the Intel was impressive. The AMD APU as a whole blew the Intel chip out of the water as seen in the benchmarks PCjunkie linked. So if you're going to use the graphics part of the chips as intended, then the AMD is the clear cut choice. It's not even close. Like sticking a Lamborghini against a Yugo.

Which makes for quite the quandary. If you're building a SteamOS Box and want to do local gaming without a dedicated video card it's 100% in AMD's favor. If you want to stick in a dedicated video card then the best choice as an overall system platform is Intel. If you're streaming the game from another PC then who cares.

But what is missing from the test is Mantle. And there is a video of BF4 with 60 - 100 fps in another thread directly from the CES floor. So in the end we really need to wait for AMD to release the complete system before we pass judgement. HSA is supposed to do something for Office speed if I remember right. I doubt that Mantle was running on the test system as AMD hasn't released it yet.
 
@cageymaru

I honestly hope so. AMD really needs this not to suck. If we have to have everything perfect as in hardware, software and Mantle just to get good numbers then I don't think consider this very good news. To be honest those graphic numbers are way lower than I expected for nextgen steamroller with gcn inside. :(
 
whats interesting is it seems to stop making any real gains with memory faster than 1866.
 
pcjunkie said:
@cageymaru

I honestly hope so. AMD really needs this not to suck. If we have to have everything perfect as in hardware, software and Mantle just to get good numbers then I don't think consider this very good news. To be honest those graphic numbers are way lower than I expected for nextgen steamroller with gcn inside. :(
Click to expand...

They also need some serious design wins outside of steambox/etc.

An A8 in a brix/NUC type thing.

A kaveri T-series thinkpad for example with a matched dgpu.

or just a kaveri A10 in an X series thinkpad..
 
omega-x said:
They also need some serious design wins outside of steambox/etc.

An A8 in a brix/NUC type thing.

A kaveri T-series thinkpad for example with a matched dgpu.

or just a kaveri A10 in an X series thinkpad..
Click to expand...

Oh Lenovo, Apple, and others were showing off those devices in a video I linked in another thread. Here ya go!
 
Overclocking is what I want to see. OC the hell out of the GPU and see how it runs.
 
KazeoHin said:
Overclocking is what I want to see. OC the hell out of the GPU and see how it runs.
Click to expand...

I could OC my A10-6800K but dam that thing ran hot, even at stock. I tried wc and a good aftermarket heatsink/fan but the thing would throttle anything over 70c and it reached those temps pretty dam fast. Heck it seemed to throttle even though the temps weren't bad? I have to admit it was fun to play with and one of the most stable systems I ever used, even with those issues.
 
doesnt seem right that it would reach those temps with a good aftermarket WC/heatsink fan. its a 100 watt chip. I was able to keep my 4.9ghz FX8320 cool and my power meter was showing it pulling 335 watts at the wall under full load overclocked. I was using a thermaltake pro 2.0 all in one water cooler.
 
Wonder how it would run with "dual graphics" mode of an added amd graphics card. Would the performance advantage be good enough to warrant an AMD rig over an Intel one?
 
buttons said:
doesnt seem right that it would reach those temps with a good aftermarket WC/heatsink fan. its a 100 watt chip. I was able to keep my 4.9ghz FX8320 cool and my power meter was showing it pulling 335 watts at the wall under full load overclocked. I was using a thermaltake pro 2.0 all in one water cooler.
Click to expand...

When I said hot I meant temperatures jumped right up to 55-60C+ from idle with pretty much any load. I tried a ThermalTake 2.0 (non-pro 2 fans) and a CM Hyper 212+ and they both had a hard time keeping up. You have to remember this thing is doing on average 4.2GHz plus using the integrated gpu. Check around as the stock heatsink/fan would severely throttle this chip so it was almost a must to put a better one on. Good little chip but it did run warm.
 
HvyMtl said:
Wonder how it would run with "dual graphics" mode of an added amd graphics card. Would the performance advantage be good enough to warrant an AMD rig over an Intel one?
Click to expand...

that's a question that a lot are asking.

Further more... what amd graphic card will Dual graphics will be optimised with the Kaveri.

According to apu13 presentation it will support asymetric crossfire on mantle enabled apps. Which will be awesome. Can you imagine what kind of performance you would get kaveri + R290x?

But how will be the performance under Direct3D and OpenGL under dual graphics?
 
HvyMtl said:
Screen shots or other places where the review can still be read?
Click to expand...
Bing has it cached.

OP: Thanks for the link. I checked the usual sites and none seemed to have samples yet.

Not a great review, but there's enough to gauge the performance. Puget tested the A10-7850 (3.7/4.0 + DDR3-2133) vs the A10-6800 (4.1/4.4 + DDR3-2133) and Core i5-4440 (3.1/3.3 + DDR3-1600... which is pretty close in performance to the i5-2500). You can see the boost in IPC vs the last generation, pretty much the 20% AMD claims, at least in the microbenchmarks. Due to the decrease in clock speed, it winds up being only ~10% faster than the A10-6800K.

What's surprising is the modest gain the GPU gets vs the previous generation. It looks like Iris Pro 5200 may still hold the crown in most games. Too bad it's really rare or Iris Pro might have made a difference. :p IOW, It's still pretty much HD4400/HD4600 for most top Intel CPUs this generation, and even the last generation of AMD APUs could easily beat those models.

edit: If you just want to see the benchmark results: http://imgur.com/a/ZiwCC
 
Last edited:
The graphic is only close to R7 240 which is very low end, not even close to a HD7750/R255 performance.
 
Very disappointing if true, and this test was with 2133. Not like the numbers will get any better.
 
FordFanboy said:
Very disappointing if true, and this test was with 2133. Not like the numbers will get any better.
Click to expand...

I ran their heaven preset on a 8750m (2gb ddr3, 384 GCN cores, a fair match to kaveri sans a few shader units).
44.3 avg fps.

34ish on the igpu of an A10-5757m (overclocked iGPU to 800mhz, 512mb alloted ram @ 1600mhz dual channel, 384 terascale4 shaders)

Which kaveri beats handily. I'd say *not bad at all* even if the mobile version ends up only matching that.
 
Last edited:
omega-x said:
I ran their heaven preset on a 8750m (2gb ddr3, 384 GCN cores, a fair match to kaveri sans a few shader units).
44.3 avg fps.
Click to expand...
44fps is what the A10-7850K got in Unigine using DDR3-1600 memory. http://i.imgur.com/czD5xlZ.jpg I can't imagine how many specs would have to be reduced to bring the A10-7850 into a mobile TDP, even a high one like 45W. You're going to be better off with a discrete GPU if you find the 8750m an acceptable baseline.
 
xorbe said:
Click specifications. Looks like the mobile chip loses 25% gpu MHz, 34% cpu base MHz, and 17% cpu turbo MHz, while cutting tdp 65%.
Click to expand...
The turbo speed hit is low because other thermal factors will reduce it quicker than desktop models due to laptop's worse cooling. While it sounds nice, it's kind of misleading to expect the highest speeds for any significant length of time while running a heavy load or even games. For quick bursts on 1c/2c, maybe it's accurate. When running multiple cores, it's going to be running much closer to base frequency, which is going to also affect real world gaming performance.

I'm not trying to be too hard on it, it should be pretty decent for an iGPU even cut down somewhat. Someone who plays modern games, or any other stressful gaming in other than low settings is still going to want a discrete GPU. Even low end ones will out perform the top iGPUs, desktop or mobile.
 
You must log in or register to reply here.
Back
Top