The consequences of repasting an XPS 15 9560 and why most people shouldn't do it

There are quite a few threads on this forum for people who modify the cooling design of the XPS 15 9650, mainly via "repasting" (i.e. changing the thermal paste of the CPU and GPU to a better one than Dell uses). In this thread, I would like to make an argument for not doing this... or at least that it is a lot harder to do right than it first appears and stopping halfway might cause problems. I'm not an expert in this so more experienced people should feel free to correct me if I'm wrong.

Two phenomena combine to make this harder than it looks. The first is the nature of modern CPUs and GPUs: rather than operating at a fixed clock frequency when under load, they operate between a base frequency and a boosted one. The boost is intended to speed up tasks which do not take enough time to overheat the chip. If the maximum boost is sustained for a sufficiently long time, the temperature of the chip will increase which causes the chip to lower the clock speed. It will continue to lower it as higher temperatures are reached until it stabilizes at some point between the boosted frequency and the base one. Now, the crucial point here is that (as far as I can tell; again, people can correct me if I'm wrong) the input to the clock-lowering algorithm is the temperature of the chip itself, not that of everything else in the system.

The second phenomenon has to do with what repasting actually does. It is often lumped together with undervolting, but the two are fundamentally different in that undervolting literally lowers the amount of heat produced by the chip being undervolted at a given level of performance. The power used by the chip is roughly proportional to the square of the voltage so decreasing the voltage causes it to use less power and thus produce less heat. Thus, as long as you stay above the true minimum voltage, even the simple version of undervolting is safe and efficacious. Repasting does not lower the power usage or the heat output -- it merely moves the heat away from the chip better than the stock cooling solution.

So, what is the result of all this? The temperature of the chip itself is lower in a repasted system and thus it will operate at higher frequencies longer than it otherwise would... but the heat thereby produced has to go somewhere. In large machines with powerful cooling systems (such as the Clevo P650), the heat goes outside of the laptop and you wind up with a machine that produces more heat, but has a higher performance. Unfortunately, the cooling system of the XPS 15 isn't that powerful and the inside of the laptop also heats up.

This is bad news because in addition to the CPU and GPU, there are also a bunch of little chips that serve various purposes such as regulating the power delivered to different parts of the system. They do not benefit from your repasting, but they also produce heat and must be cooled by the overwhelmed cooling system. The more efficient your repaste, the greater the likelihood that one of these chips will overheat. Furthermore, they don't have algorithms that gradually decreases the clock speeds of the CPU and GPU; they simply wait until they reach temperatures close to boiling point (i.e. 95-99C) at which point an emergency override kicks in to prevent them from frying and they stop doing whatever it is that they're supposed to do. Since the ones that overheat first are typically those involved in delivering power, this causes the CPU and/or GPU to instantly downclock -- most likely not just falling from the boosted frequency, but below even the base one.

Now, if you were sufficiently dedicated to this, could you also provide cooling solutions to these tiny chips and thus end up with a system that runs hotter, but performs better? Yes, of course. You can read in the other threads about attaching thermal pads to them and which chips need more help than others... but how many people really want to do this? Unless the factory paste job of your specific machine is really bad, it's much easier to simply undervolt the CPU, accept Dell's intended performance level and enjoy your new laptop.

 

Agreed - those with very limited skills should think twice before repasting.

One key issue is that the Dell factory thermal paste job is frequently terrible. There is pleny of photographic evidence in this forum. I can't believe even Texan engineers accounted for that so plenty of users are getting crippled performance out of the box.

Good 9560 data does not exist but a lot of 9550 data indicates that:

-undervolt can provide ~10*C temp decrease
-repaste can provide ~5*C decrease

These steps can unlock a lot of performance for nearly free. Obviously other bottlenecks will appear. And when those bottlenecks are addressed, new ones will pop up.

* Note Kaby Lake has less potential upside to undervolting as it is essentially a factory overclocked SkyLake chip

* And YMMV as each chip varies. Heatsink warpage varies. Factory paste and user repaste vary. etc...

 

I thought you were going to say you broke your new XPS doing it.

Your post does have some weight but i'm afraid with this laptop it will still overheat and kick in the hardware throttle, the VRM is the main cause. The only difference is how long it takes to do it.

 

The first line of throttling is indeed like this, and kicks in very quickly.
However, there are further lines of defence against the laptop frying itself to death, like the power limit throttling, which acts based on other temperature sensors around the laptop.
You are right that with an improved paste job, temperatures elsewhere tend to rise higher. But this should still not be dangerous.
The problem is that in order to produce a laptop as thin, the thermal design here is so borderline that power limit throttling tends to kick in after a few minutes of heavy load, and that power limit throttling isn't smooth. I don't think an intentional bad paste job is a solution, hehe. It is possible to undervolt, pad the VRMs, limit the CPU and iGPU in Advanced power settings etc. Question is then, does it make sense to have the i7 in such a laptop, if it overheats instantly. But some folks don't play games so much, and are happy if the i7 runs unthrottled when the GPU is idle...

Hehe, me too

 

I thought the same thing.

Agreed. Particularly for demanding games.

But for medium-duty tasks and games with lower graphics settings, the repaste and undervolt can open up better performance without throttling.

Of course, that better performance window is limited until someone figures how to keep the mosfets (well the whole vrm) below the factory set throttling threshold of ~78*C

 

oh this logic is bad... sorry but it is...

Thermal limits on alot of the components in the motherboard are rated for 120C in most cases. VRM, chokes, etc...they can handle more heat than your normal CPU or GPU because the internal components are made at a much much larger structure than the 14nm components of a chip... Your chip will fail in most cases before your VRM or motherboard components do if you're not running the VRM at a constant 110-120C which no one here at stock has seen temps even that close... I've seen 95C on mine but that's about the highest I've seen....

You might as well just massively underclock your GPU and use toothpaste as your thermal paste if you apply this logic...and be happy with a 1GHz CPU clock...

 

I figured I'd throw my two cents into this even though they are exactly that...

 

Exactly right in my experience. The repaste with undervolt has actually enabled me to run unthrottled but it requires a -400/-500 under clock on the GPU to be completely stable and unhindered. Next step is to tackle the VRMs

Sent from my SM-G935F using Tapatalk

 

You sure it's 78? I have some pretty conclusive evidence to suggest it's 98C when it gimps the GPU clock

Sent from my SM-G935F using Tapatalk

 

Mine goes up to 98C before it gimps the GPU (assuming the "Ambient" sensor in hwinfo is a VRM)

Sent from my SM-G935F using Tapatalk

 

It looks like almost everyone thought so. No, truth be told I haven't even opened this one yet and was thinking of just leaving it be.

As far as I can tell (based both on my own tests and on reviews), it seems to handle pure CPU tasks without issues. I ran Prime95 for half an hour yesterday and the CPU stabilized at 3.4GHz (which I believe is the maximum possible with 8 threads).

They might not physically melt until a higher temperature, but the firmware limits are below 100C and if they reach those, the clocks will be throttled.

There is always a tradeoff between performance and heat. If the stock paste job is really as bad as toothpaste, then yes, a repaste is in order. Mine doesn't seem to be that bad though.

 

I see the argument, but not the reasoning behind it. Yes opening and repasting your notebook is a so-so advanced technique, but apart from an on-going discussion I haven't seen a post/thread/tutorial
that does not advise extreme caution or knowledge before doing it. Though I see where you coming from, some people are careless.

This is in part wrong. For the XPS 15 9560, which is the main subject here for example, the Intel Dynamic Platform and Thermal Framework has multiple sensors across the Notebook like
CPU, GPU, VRM, Chipset, Ambient, Battery and RAM and maybe others. Having said that there is a separate throttle control on the CPU and GPU itself that reacts if they are in danger
of overheating (PROCHOT) as well as on most BIOS/UEFI (BD PROCHOT). To prevent IDPTF from kicking in you have to jump through quite some hoops to disable it permanently and endanger
your notebook.

I disagree that you always need the full package you will still benefit from an undervolt and repaste, the throttle will just kick in later or at higher constant clock speeds even without the pads.
The IDPTF will prevent overheating.

Please don't take this as condescending, but I am really unsure about this post. It's a combination of half-knowledge, sound logic derived from assumptions, and
well meant general information that is true for almost every notebook.

It's not a bad post, but for this wall of text I did not see it providing worthwhile information besides the general one.
Might be a skewed viewpoint from my side though.

 

Dell logic dictates testing the CPU and nothing else hence a crap stock paste will do 3.4 with prime95.

That is not pushing this laptop.

 

Well the OP is right in the part that the power limit throttling which kicks in later is, unlike primary CPU and GPU throttling, programmed as a rather stupid safety measure which kicks in at once and heavily, rather than limiting the performance to the thermally sustainable level smoothly. The folks that have programmed it have probably never heard of PID control and such. I think this is sth one could complain about to Dell, but I guess this is not likely to be improved quickly, if it is handled by Intel DPTF...
Advice to reduce FPS/quality and how to throttle the CPU manually to avoid this kicking in has been discussed several times, for instance using Advanced power settings or with consecutive profiles in Throttlestop (well that's also kinda step-wise).

It would've been nice to have a diagram/list of the known throttling measures in post one of the thermal issues thread. At least a table with the relevant sensors and threshold temperatures.

 

It appears I did not get my point across. Let me try again, this time with a hypothetical example. The issue is not about carelessness or about frying the laptop; careless people should not mess with electronics and you are right that IDPTF will prevent serious damage. My point was that even if you do everything right, the nature of the cooling system and the way the firmware reacts to the heat of various components might mean that you've made things worse for certain workloads.

Here's the example: suppose I'm playing a certain game which, with stock paste, starts off at 55 FPS (at maximum boost for both CPU and GPU), but, after a bit of time, the CPU and GPU heat up and the frame rate is reduce. Suppose that the entire system reaches thermal equilibrium when the frame rate is 45 FPS. That is, when the CPU and GPU are running at this level of performance, the two of them and all of the other chips on the motherboard remain at the same temperature. Obviously, all of these numbers are purely hypothetical and in reality, the temperatures of the various chips rarely stay fixed (because the clock frequencies change in discreet steps, they usually wobble slightly around the point where they would stabilize if the frequency binning was finer), but never mind -- this is close enough for example purposes.

Now, suppose I repaste this laptop with a better paste than Dell used. The result is that the CPU and GPU are now better cooled and thus their temperature stabilizes when the frame rate is at 50 FPS instead of 45 FPS... but this is producing more heat than before so some of the other chips on the motherboard (which do not benefit from the repaste) are not at equilibrium. Unlike the CPU and GPU, these chips do not have an intelligent downclocking scheme which would bring us down to at least 45 FPS or perhaps a bit higher (after all, we improved the cooling solution a bit). Instead, they do nothing as they heat up until they reach a certain threshold at which point they go into panic mode and drop the frame rate to, say 25 FPS. Once that happens, the system cools rapidly and once the offending chip reaches a different temperature threshold, it goes back to normal mode and the frame rate returns to 50 FPS (which is again too much so the cycle begins again).

The result is that instead of gaming at 45 FPS, the repasted system alternates between 50 FPS and 25 FPS every few minutes which I think most people would find to be a worse situation than they had before. Of course, this example is completely made up and whether a given workload would follow this kind of pattern depends on its nature, on the quality of the new and original paste and on the details of the cooling system and how the various chips react to heat. However, given that repasting allows the CPU and GPU to run at boosted frequencies longer, I think it is a reasonable assumption that it increases the chance of a workload running into the less intelligent form of throttling.

I was trying to word it so even people who don't know much about computers would understand (in retrospect, I could have skipped 80% of it given that most people here do not fit that description). Also, I tend to be verbose. The main point I was trying to make is that even if you do everything right, repasting an XPS 15 9560 may make the throttling worse depending on your workload.

 

Yes I understood what you meant in your first post, my disagreement does not stem from a lack of understanding. I really like the term "thermal equilibrium" though it really captures what you meant.


After re-reading and considering your latest answer I guess what rubs me wrong way is that on one side you present your thesis like solid, backed advice with some facts that are correlating, but not causing the problem and on the other side lack proof for your assumptions. Now as the opposition I am in a way no better since I also can not point to hard proof, though you admitted that you are for one no expert and for two have not opened your notebook, yet. What I do have are the hours experimenting with my 9550, (which is not the exact same model, but close) which directly contradict what you are writing here.
Sadly I have not documented these hours in text or picture. Only some reddit and forum posts remain of this. So all that is left for me is to say that I disagree with you on your point that repasting can make throttling worse, because I had a different experience.

 

I still agree that people with no handyman experience should NOT (EDIT!) be repasting.

But these thermal problems are complex.

- For example, one important factor is that this is not a closed loop system. For example, hot radiators will transfer heat to the air more much more efficiently than warm radiators will. That is somewhat mitigated by factors such as shared heatsinks for the CPU & GPU and the screen blocking hot exhaust air.

- Also if Dell factory thermal paste application is substantially below spec, as posts in the 9550 and 9560 forums seem to indicate, one could assume the cooling system is underperforming engineering spec. In this case, I would argue that a decent repaste job that meets or exceeds spec could provide some additional "free" performance allowing the XPS to perform as it was engineered.

There is not much 9650 before and after data to support that point. But with the very similar 9550 there is some decent self-reported data in this forum that shows about ~5*C temp improvements available with a typical repaste (a summary of a lot of data is in google docs).

I know these comments don't directly address your point. And there are plenty of unknowns, chip variance, etc. so this requires trial and error. But I will say that a simple repaste and undervolt dropped thermals more than 15*C on my 6300HQ and allow me to run demanding software for hours without throttling anymore.

 

The i7 throttles heavily on stock as well. We saw major cinebench score reductions after repeated runs

 

It's really hard to prove anything here as the optimal course is almost certainly different based on the workload and the specific machine. Clearly, if the stock paste is so terrible that the CPU and/or GPU throttle below the base clock under the intended load, then you don't really care about what happens with the VRMs and are better off repasting (I noted this exception in my first post). And while I haven't opened my new 9560, I have tinkered fairly extensively with laptops before (although none of them were similar to the 9650). The reason I made this post is that it had never occurred to me before that repasting may be anything but an unmitigated good (as long as it was done properly, of course).

 

I think one of the biggest issues is no one has really done much testing of the VRM in stock form without repasting. I would welcome the OP or anyone to take some logs and see what the VRM is doing in stock form, my guess is it's still an issue, maybe it just takes an extra few minutes to get there.

 

I guess there already is experience regarding this. There is a lot of variability depending on the factory paste jobs though - for instance some stock units can run Prime95 without throttling, others can't. Mine was pretty throttled after just a few seconds, and indeed the power limit things didn't kick in because temperatures around the computer were relatively low, despite high CPU / GPU temperatures (Heaven). After repasting, running at faster clocks turned things around, CPA and GPU don't reach their thresholds, but power limit kicks in after a while.
However, the original state was surely not good, because the CPU and the GPU were cooking. In conditions where the cooling can't sustain stock max frequencies, I think it is a better idea to limit the frequencies intentionally so that power limit hard throttling doesn't kick in, and the CPU&GPU can run at lower temperatures.

 

Are there experiences with this? I haven't seen anyone monitoring VRM temps with stock settings yet. Did you play games or do something besides just prime95 and watch the VRM temp? Just because your CPU was throttling doesn't mean the VRM wasn't getting hot.

 

9550 i7-6700hq Prime95 (blend)

Stock / Repaste
Core clock MHz: 2000-3000 / 3090
Core temp max oC: 97 / 76
PCH temp max oC: 58 / 62!
Dell EC temps oC: 95 81 64 49 62 51 40 68 / 75 88! 64 49 67! 53! 59! 58
(! higher after repaste)

Temperatures after repaste were not higher in all tests, for instance everything was cooler after repaste running Heaven (which means that a bad paste isn't good in all circumstances). But I've also noticed when trying a few minor cooling mods after repaste, like opening and closing the small grille on the case edge near the VRM and lifting the rear while running Heaven+Prime95, that many of the Dell EC temperatures towards the steady state are simply strongly correlated with CPU&GPU clocks (when the fans are already maxed). So if you manage to let them run faster, some places inside get hotter, nothing weird about this. What we can do with VRM pads to sustain faster clocks
- even out the local temperature spikes which may trigger local temperature sensors earlier. - dissipate some more heat from the VRM area via air (default venting is poor) or chassis bottom.
Or just cap the CPU and iGPU power in Advanced power settings...

 

Do you mind listing out what the Dell EC temps are? I'm curious why some are so vastly different. Also maybe temps went up because fan speeds are down? I know my testing after repaste and undervolt the fans ran much slower.

 

I just can't bring myself to read any more of this thread, I'm far too ill to read the size of the posts.
to put this to bed for me, Toshiba asked us to use mx2 thermal paste on all board swaps and not pre applied paste pads.

Sent from my SM-G920F using Tapatalk

 

So that we're comparing apples to apples, can you suggest tools for monitoring the VRMs and for stressing the system? I have the HWInfo log from playing a game last night and neither the CPU nor the GPU nor any of the "ambient" temperatures seem problematic: there is an occasional spike to 85C, but they mostly sit at or below 80C. I'm going to try making some graphs to be sure...

 

In the order of appearance in HWinfo64:
CPU Ambient Ambient Temp3 Ambient Ambient Ambient DIMM

 

I've only used HWinfo64 to monitor mine. I don't know what would be good to stress with, depends on what you're going for. Maybe a few loops of Heaven would be good?

 

I ran Unigine Heaven for roughly 20 minutes worth of loops (there were either 6 or 7 of them; I lost count). I didn't change any settings -- simply downloaded the free version and pressed Run. My CPU is slightly undervolted (-0.085 offset). Here are the highest HWInfo64 values (individual ones; they never reached this set together):

CPU: 85
Ambient: 85
Ambient: 74
Temp 3: 38
Ambient: 56
Ambient: 73
Ambient: 43
DIMM: 60

The GPU clock was mostly at 1670 MHz, but it occasionally briefly (a few seconds) went up to 1683 or similarly briefly fell to 1594 (or even 1569, but the latter never for more than 1 second). In either case, it always returned to 1670. Both fans were running at roughly 4900 RPM.

The one thing that gives me pause here is that the CPU does not appear to be fully utilized: HWInfo64 says there is no throttling, but it does not run at 3.4GHz the entire time and often dips below even the base clock speed for more than a minute at a time (its temperature when it does this is 70-72C). VRM throttling or simply no need for the higher clocks?

 

Well the first ambient temp reading is your VRM temp (AFAIK), it may be throttling your CPU, those temps are hot fyi... Was that with an undervolt or totally stock?

 

Agreed.

That first ambient temp is related to VRM, which above ~78*C generally throttles 9550 and 9560. That may be listed as PL1 throttling in some cases.

You can see that throttling flag easily in XTU...

 

I can also confirm that after adding thermal pads to my VRM, temps dropped from 105C to 82C. It probably would have went higher than 105C, but my 9560 always force shuts down around that range after a few minutes. CPU throttling exists when VRM is above 78~79C (800~900MHz). I unfortunately still throttle after the addition of pads, but my 9560 no longer freezes/automatically restarts when under load.

 

That was with an undervolt, but only a -0.085V one. I increased it to -0.110V and that dropped the temperatures by a few degrees. It's pretty bizarre though: looking at _sem_'s results, mine are hotter in some places and cooler in others and they're not where _sem_'s repaste changed things.

I am not seeing this. The CPU and GPU sometimes run at full power even when the first ambient is above 78C:
Core 0 - Core 1 - Core 2 - Core 3 - ... - CPU - Ambient - Ambient - Temp 3 - Ambient - Ambient - Ambient - DIMM - ... - GPU Clock
3490.6 3390.9 3490.6 3490.6 ... 71 81 70 38 52 72 41 58 ... 1683
3390.9 3390.9 3390.9 3390.9 ... 73 81 70 38 52 72 41 58 ... 1683
3390.9 3490.6 3490.6 3490.6 ... 74 81 70 38 54 72 41 58 ... 1683

Thanks, I will try it tomorrow. I was hoping to do it this time, but my XTU stopped responding during the Heaven run for some reason.

 

I guess shutting down is not normal. Maybe due to a really miserable paste job, overheating at min clocks?

 

In Heaven, the CPU is indeed not used so much, downclocks due to boredom.

 

So I can confirm that XTU sees "Power Limit Throttling" when I'm playing a game (it doesn't see Thermal Throttling, Current Limit Throttling or Motherboard VR Throttling). The Power Limit Throttling does not appear to be correlated with the first Ambient or any of the other temperatures. I am not sure what it means, but it has no impact on the gameplay (none that my eyes could tell, anyway).

I think that is it and it also happens in the game -- the CPU frequency is all over the place from 3.4 GHz all the way down to 1.6 GHz. I suspect the 7700HQ is way overkill for the 1050 so it will spend a lot of time underutilized unless the game has some CPU-specific features.

 

Good real world use is what matters...

Intel shows a lot of factors impact PL1, including power, current, voltage and thermal limits. There is not a lot of clarity there and Intel is always tweaking.

One backdoor trick I use on my 9550 to control throttling is to adjust PL1 and PL2 (and the max time) via XTU to something like 55w, 60w and 28 seconds. I did that only after I got my laptop thermals very low and stable.

- That can have some risks as skirts some thermal protection

- I don't know if it will work on Kaby Lake processors, particularly as the 7700HQ seems to
be killing the cooling system

- On newer Intel processors, Uncle Webb said ThrottleStop may not be able to access all the turbo registries. But...if you run ThrottleStop you need to match those settings in the TPL window or the registers keep getting rewritten

- You can't run XTU and ThrottleStop at the same time to prevent registry conflicts. The way I got around this was to make changes in XTU then shut down; those settings should stick and I never reopen XTU. Now I launch ThrottleStop every morning.

I have ThrottleStop running my CPU at max clocks all the time. That is basically by using SpeedShift with EPP=0. In that case, when you see CPU speeds drop off a cliff, you know there is throttling vs. "underutilization."

You can search the ThrottleStop thread to see why Uncle Webb believes SpeedShift increases snappiness and reduces power consumption.

 

Sounds like Throttling to me and unigine isn't a good real world test for latest games it's like 5 years old minimum. Use BF1 or GTA V to stress both or the ROG Realbench Stress Test

Sent from my SM-G935F using Tapatalk

 

Same issue I have and where this used to throttle CPU and GPU, after repaste and some work under the hood padding the chokes, now only CPU throttles but only down to 1.5 - 2GHz which makes no impact on in game FPS

Sent from my SM-G935F using Tapatalk

 

I agree there is something else going on here which a BIOS update would probably help with.also full control of the fans would be great Dell if you're listening!!

Sent from my SM-G935F using Tapatalk

 

Depends on the game in my experience. When I get PL1 throttling in overwatch it cuts my fps in half and I have trouble playing

 

You can use speedfan to control them, I never got as far as setting up profiles though.

 

@iunlock had the same issue as me and he has confirmed his CPU throttle doesn't impact in game performance in overwatch. Does your GPU also throttle? Double click on GPU clock speed in hwinfo and check the graph after you notice a difference in game

Sent from my SM-G935F using Tapatalk

 

Speed fan works to control XPS fans?

Sent from my SM-G935F using Tapatalk

 

It made them spin up and down on the 1st page so I will go with a yes. I never bothered going any further by making profiles or seeing if the bios kept overriding as I was checking for fan noise.

Sent from my SM-G920F using Tapatalk

 

I'm running medium/high details with 120 fps cap, I'm not sure if he was running ultra or not. The lower the detail level, the more the CPU speed will affect FPS. The GPU did not throttle.

 

I'm running it at Ultra maxed out.

Getting close to addressing the cpu throttling.

::iunlock::

 

Didn't work for me. Installed it but it doesn't detect the fans (or much else for that matter). Do I have to enable something in the BIOS?

Sent from my SM-G935F using Tapatalk

 

Not sure, maybe they changed something in the latest BIOS.

 

Just a little bump to this thread. It's not very scientific but here's what I found out with my 9560 (i5 7300HQ).

Stock:
About 2 minutes into ROG realbench my cpu throttled down to 1.4ghz

Stock paste with a -140mv undervolt:
ROG realbench again throttled the cpu down to 1.4ghz-1.6ghz after 6 minutes

Repaste with kryonaut and stock volts:
ROG realbench again throttled the cpu down to 1.4ghz-1.6ghz after 9 minutes

Repaste with -140mv undervolt:
Ran 2 back to back ROG realbench sessions @ 15 mins each without cpu throttling.

I am now trying a 1 hr session to see when it will throttle. CPU and GPU temps are much much better, although I'm still kicking around the idea of using my conductonaut on them to see where I end up. I will also play around with padding the VRM/chokes since I'm not sure when @iunlock is going to release his results, lol.

Not saying this negates your whole repaste is bad theory, but for me it shows that it was worth it.

*on edit*
after an hour of ROG I did get throttling of the CPU down to 2.4ghz after 34 minutes and it bounced between that and 2.8ghz the rest of the time. Not bad considering ROG destroys my 9550 with the i7 (with hyperthreading off it does much better).