Discussion in 'Fujitsu' started by HoyaNM47, Jan 1, 2014.
And this from an i5 Lenovo x240: LENOVO 20ALCTO1WW - Geekbench Browser
My impression from the way the Fujitsu reps talked is that the Q704 will only turbo up to full performance if plugged in to either the docking keyboard with the second battery or the performance dock with ac. I think the device design is intentionally set to never boost to full performance unless in one of those docks. Maybe I am wrong, but I was under the impression that even on AC power it's designed to never boost all the way up in tablet mode.
Now, if the device is still nor reaching 4600u performance in one of those docks, then it's probably the infamous Haswell heat regulation doing the throttling, which means insufficient cooling design and likely no software fix will make that right.
The 4300u may do much better if not vulnerable to the heat regulation.
Thanks so much for all these benchmarks. Fuji really needs to get it together with these devices. At this price point they should at least be in the midrange performance for each processor class.
I should say that all the tests were with the tablet in the dock.
I'm looking forward to getting the i5 to see what the results are.
Wow. Thanks for clarifying that. That's huge. In the dock on AC and performing that poorly is really worrying. The 4650u seems to have minor heat issues in the macbooks (causing them to underperform a bit) but nothing so bad as what you found. That really can't be just heat throttling, though with ruggedized / water resistant systems probably make heat management harder. Still, there's no way I would plunk down 2k (or even 1k) for a device performing that poorly,
Fingers crossed that the i5 produces much better results!
Japanese review of Q704 (QH77/M), i5-4200U, this page has quite a few performance charts, comparing Q704 docked (blue), Q704 tablet (red), Surface Pro 2 (green, same i5-4200U 1.6GHz / 2.6GHz max as in Q704), and Vaio11, i5-4210Y (purple)
Other Japanese review i5-4200U Q704
I don't understand Japanese, but with Google Chrome translate you can get a lot out of those reviews, and even without any translation, the benchmark program names are in English in the chart description, so even without translation program those can be quite interesting.
Thanks. If I am understanding what we see there correctly, then it seems those review indicate mediocre to poor battery life and performance significantly lower than Sp2 and Envy with the same CPU, even when docked. It looks not good at all.
I'd love to hold out hope that these issues will get sorted very quickly, but I do fear they have to do with heat issues limiting the turbo mode, which may mean it's a total design issue and not just firmware.
Fujitsu has a bit of a reputation of limiting their hardware with poor thermal management. The T9-- series have always gotten hot inside rather quickly.
This is a whole other level though.
HoyaNM47, have you tried running RealTemp in the background to see how hot the CPU is getting during the tests? I would be curious to see how much headroom we have, and if they have room to tweak performance settings in a BIOS update.
Here (Traductor de Google) explains more about that issue
Thanks @ HoyaNM47 for putting up with us
I was also about to pull the trigger on this one and now wondering.
Hopefully when you get the I5 we can have a second look and if things look as bad you can have a talk with Fujitsu to clarify things up...
Suddenly the Helix is back in the race as we know it has special cooling fans in the dock exactly for the purpose of allowing full power when docked. And these 20% are nice on the price.
The worrisome part is that I dug up the old post below on the Q702 that apparently suffered the same feat.
Last I found this that may suggest only the Desk Dock would offer the full performance.
"The Q704/H goes beyond the tablet-style and keyboard-style options of a two-in-one Ultrabook™. It is a three-in-one tablet(10), with a third option, cradle style, for maximum performance."
I think we will need some clarifications from Fujitsu...
A couple of posts up Hoya did mention that --> "I should say that all the tests were with the tablet in the dock."
... I know ... it is hard to accept isn't it? I want it to be not true also, because now I am left scrambling for other options.
Separate names with a comma.