ASUS: The EN7600 GS Silent, and EN7800 GT TOP Silent
by Josh Venning on May 12, 2006 11:00 AM EST- Posted in
- GPUs
Overclocking, Heat & Power
We approached overclocking these cards with some caution because of what we mentioned before about heat issues. Since faster clock speeds equal higher temperatures on the GPU, more heat dissipation is needed, which can be a problem for cards with fan-less heat sinks. We found that the 7600 GS Silent did actually have trouble with overclocking, but interestingly, the Silent 7800 GT managed to achieve quite a high core and memory clock. Here are the clock speeds we achieved.Overclocking Results | ||
Core Clock | Memory Clock | |
ASUS EN7800 GT TOP Silent | ||
Default | 420 | 1240 |
Overclocked | 481 | 1290 |
% Increase | 14.5% | 4.0% |
ASUS EN7600 GS Silent | ||
Default | 400 | 800 |
Overclocked | 442 | 874 |
% Increase | 10.5% | 9.3% |
We tested these clock speeds by running the Splinter Cell Chaos Theory benchmark stress tests to verify stability. One thing to note is that these cards tended to get incredibly hot to the touch after several minutes of testing. Here are the performance results in Splinter Cell of the overclocked cards.
There's not much difference at all between the factory and user overclocked silent 7600 GS. This card was the worst of the two at overclocking and had a lot of trouble running stable (i.e. without graphical artifacts) during our stress tests. However, once we established the overclock of 442MHz/874MHz, the artifacts left and our tests ran cleanly. With the silent 7800 GT, the difference our overclocking makes is fairly significant. It did manage to boost performance by several fps, which may be a noticeable improvement during gameplay.
Power Consumption
We also tested the power consumption of these cards to see how much power they drew and if there was anything out of the ordinary. We acquired our power numbers by measuring the wattage of the total system at the wall outlet while idle and under load (stress tests). These are the results we recorded.
As expected, the EN7600 GS Silent drew much less power than the EN7800 GT TOP Silent. Neither of these cards are really power-hungry, but the fact that the EN7800 GT TOP Silent requires a 6 pin power connection reflects that its consumption is on par more with other cards of this type. The EN7600 GS Silent requires no external power connection, and its consumption is fairly low, which is good for those concerned with power draw.
Heat
We tested the temperatures of these cards as well to get a good idea of their heat levels generated during operation. We acquired the results here similarly to the power test, by taking the temperature of the GPU (in Celsius) while the system was idle and under stress testing. Here are the results.
As we mentioned above, both cards do get very hot to the touch after running any 3D application for a bit. Something interesting here is that the EN7800 GT TOP Silent runs cooler at idle than the smaller EN7600 GS Silent. At load however, the 7800 GT gets about 5 degrees C hotter than the 7600 GS. The EN7800GT cooler is obviously larger and better at dissipating heat, yet the card still gets hotter because it is a much more powerful part.
39 Comments
View All Comments
Seer - Friday, May 12, 2006 - link
This review is so messed up that it's not even funny. Okay, maybe it is funny to laugh at how pathetic it is. Let me start:1) The 7800GT Top Silent isn't available anymore. It was a limited edition run. GJ guys, there goes half of your article. (You even state this at the end that its not available. WTF dudes, what are you smoking?)
2) Extremely inaccurate test results. The O/Ced version of the 7600 GS is repeatedly listed as performing WORSE than the non O/Ced version. Niiiice. Also, performance for the 7600 GS INCREASES when going from a lower resolution to a higher one. Rofl.
3) Averaging core and memory % o/c's in a 50/50 weighted ratio gives the 7600GS an o/c os 9.9% avg, and the 7800GT TS a 9.25% avg. Yet you claim it overclocks better. Oh, right, you must have said that because the O/C version was performing worse. (I realize that the 7800GT TS is already factory o/c'ed. However, this means that you are guaranteed that stock clock. A real, consumer overclock is raising the clock to a speed that the chip [I}is not rated at. )
4) Hardly any mention of the EXTREME heat these cards put out. Into the air that should be going into your CPU, nonetheless. (Pointed out in an earlier comment). In fact, "the heatsink appears to be very affective at keeping the GPU cool" 95 C. Right, cool. In fact, if you understand the relationship between Heat and Temperature (two different things, people), you would understand that the 7800GT TS is in fact much worse for your case's ambient temps than the 7600GS, despite this articles claims to the contrary.
GJ guys, pat yourself on the back and go buy a few beers! You really deserve it! *rolls eyes*
Guuts - Friday, May 12, 2006 - link
Also, on page 5, the author writes: "What is even more nice..." More nice?! I hate to be the grammar police here, but I think what he's looking for here is "nicer".JarredWalton - Friday, May 12, 2006 - link
Fixed, thanks.Seer - Friday, May 12, 2006 - link
I forgot to mention why I think this shitty article came out: Asus paid them to do it. It's just like those shitty articles on Daily Tech about seemingly random product releases that are nothing other than shameless plugs.poohbear - Friday, May 12, 2006 - link
please tell your mom to stuff feeding u cocoa puffs for breakfast. they have TOO MUCH sugar for u.JarredWalton - Friday, May 12, 2006 - link
Actually, the 7600 GS Silent was released and they sent it to us for review (the same way we get nearly all products). Since it was a silent card and we hadn't looked at the 7800 GT yet, it was included as well as something of a reference point. Some people are interested in silence and don't care all that much about performance past a certain point, and that's where the 7600 GS fits in nicely.As for the other comments above, the results in the 1600x1200 graphs look like they were reversed, and in fact all of the results look odd. I'm checking with Josh to verify, but it could be the OC'ed 7600 is throttling down due to heat. I'm not sure about the stability of the 7600 OC'ed either - Josh mentions artifacts, so I'm not sure if the final results he posted are "clean" or not. A 9% OC shouldn't give you 35=65% more performance, or cut performance a lot.
Other than that anamoly, the 7800 GT overclocks "more" because averaging RAM and GPU OCs is not normally an accurate way of quantifying performance. Most modern games are hitting the GPU harder than the RAM bandwidth (unless you have really slow RAM), so 15% is more than 11%. Honestly, I wouldn't even think about OC'ing a silent card, though. Maybe if you want to add a low RPM fan to it?
Temperatures are GPU core AFAIK, so 95C is not the surface HSF temperature. Remember that these cards aren't actually generating more heat than any other card based off of the same graphics chips; the only difference is that they're not dissipating heat as quickly because they don't have fans. Unless you are using a graphics card that expels heat outside of the case, these cards aren't actually increasing the ambient case temperature.
Seer - Saturday, May 13, 2006 - link
Thanks for the response. I admit it was a bit of a troll. I was just disappointed with this article because I know AT can do so much better (IMO). Hopefully I elicited some motivation for improvement.LoneWolf15 - Monday, May 15, 2006 - link
Troll? More like a tool.If you're disappointed, there are professional ways to respond that don't make you look like a pathetic whiny git.
Seer - Tuesday, May 16, 2006 - link
Sorry for not blindly supporting any article put out by your favorite site D:LoneWolf15 - Tuesday, May 16, 2006 - link
I've disagreed many times here; if you read regularly, you'd have seen them, and you'll see them again. I just choose to do so in a manner that doesn't make me look like a 15-year old who's flunking 8th grade for the second time. Constructive criticism is one thing you apparently don't know how to do.