Why are there no Sandybridge reviews using top end C2Q CPUs?

[cadaveca]

Yes, Sandybridge is faster and overclocks better, but in real world usage the difference won't really be noticeable.

Running dual GPUs on Sandybridge leads to like a 30% performance boost over 1156 with i7 870. Just ask triptex how much faster his games are now.

 

[newtekie1]

Running dual GPUs on Sandybridge leads to like a 30% performance boost over 1156 with i7 870. Just ask triptex how much faster his games are now.

See my post above about dual GPUs.

 

[cadaveca]

See my post above about dual GPUs.

Before the launch I was set on not buying into SB at all, but now that I have it, I really think it's important to highlight how, for specific scenarios, SB is THE way to go, even with cost considered.

I do not see SB getting alot of hype, but it should. Intel did a good job, and I'm definitely impressed.

 

[TRIPTEX_CAN]

I was set on bd until I saw the results of sb I might have made a mistake but the clocks and temps speak loudly. Sb is really as good as people say. Only those who haven't tried it still defend the c2q.

 

[newtekie1]

Before the launch I was set on not buying into SB at all, but now that I have it, I really think it's important to highlight how, for specific scenarios, SB is THE way to go, even with cost considered.

I do not see SB getting alot of hype, but it should. Intel did a good job, and I'm definitely impressed.

I absolutely agree, and if you are paying for dual-GPUs then you should be paying for propper supporting hardware as well.(Says the guy running SLi GTX460s on a Celeron...)

 

[LAN_deRf_HA]

Anandtech's cpu gaming comparisons are awful. Who is running a $300 cpu at 1680x1050 or lower resolutions and low quality settings?

That res is exactly what makes the anandtech benches better than most. As others have said those sorts of tests are typically done at lower resolutions. 1680x1050 is one of the most common high resolutions in use today. Most complaints I see against their reviews are from people who can't except just how big the gap is between certain cpus. I know AMD comes out looking downright useless for gaming, but many seem to forget phenom II only performed well compared to phenom 1, it still sucks compared to yorkfield and i5/i7. I mean even a stock Q6600 gives a X6 a challenge. Hopefully it'll be better with bulldozer, can finally switch teams without feeling like I have to lie to myself to justify it.

The whole issue is of course complicated by what sections of games are being compared. My friend had a bug that had his e5200 multi stuck at half speed, 1.6 GHz from 3.2 GHz. Just moving around in an empty crysis map he only lost like 5 frames with his 5850. Pretty remarkable given it's just a low cache dual core at such crap speeds, but during AI combat he lost a good 25fps.

 

[BababooeyHTJ]

No, that doesn't make any sense at all. Who is buying a $400-500 cpu+motherboard combo to game on a ~$120 monitor on low or medium settings? Those comparisons are about as useful as any other synthetic benchmark. I gloss right over those reviews.

Secondly as I said saying that the cpu don't have any effect on resolution and IQ settings is just ignorant.

Lastly, where is a four year old 2.4ghz quad core cpu giving a 3.2ghz (not including turbo boost) hex core cpu a run for it's money? Maybe on Anandtech's reviews. Phenom 2 is at worst on par with Yorkfield clock for clock and in some cases can outperform Nehalem.

 

[newtekie1]

IMO, both high resolutions and low resolutions are important. Low resolutions take the bottleneck off the GPU and place it on the CPU, so it highlights the real performance difference between CPUs. However, higher resolution place the bottleneck back on the GPU, and show what real world usage will be like.

Considering 1680x1050 is the second most common resolution on monitors today, it certainly is important that it is included in benchmarks, it gives people with these monitors an idea of what performance they can expect, and actually it is people with these "lower" resolutions that will see the most noticeable difference between CPUs. Oh, and just FYI, 1280x1024 is the 3rd most common monitor resolution, so that too is important in benchmarks as well, and again they will show a larger difference between CPUs than a higher resolution would.

 

[Melvis]

I know AMD comes out looking downright useless for gaming, but many seem to forget phenom II only performed well compared to phenom 1, it still sucks compared to yorkfield and i5/i7. I mean even a stock Q6600 gives a X6 a challenge. Hopefully it'll be better with bulldozer, can finally switch teams without feeling like I have to lie to myself to justify it.

This is what ive seen before and it has been told again and again. A Phenom I 9950 is = to a Q6600 so if you think a Q6600 has any chance competing to any X6 from AMD your sadly mistaken. I choose to go AMD for this reason and that's for gaming, it has been proven that an AMD CPU can hold its own just fine in gaming compared to any intel CPU, and more so at high res. Not this low res BS that everyone points to, who in there right minds play games at 1024*768??? using a quad core CPU. My CPU in any game i have run so far has never maxed out my processor once, still has room to breath.

 

[LAN_deRf_HA]

This is what ive seen before and it has been told again and again. A Phenom I 9950 is = to a Q6600 so if you think a Q6600 has any chance competing to any X6 from AMD your sadly mistaken. I choose to go AMD for this reason and that's for gaming, it has been proven that an AMD CPU can hold its own just fine in gaming compared to any intel CPU, and more so at high res. Not this low res BS that everyone points to, who in there right minds play games at 1024*768??? using a quad core CPU. My CPU in any game i have run so far has never maxed out my processor once, still has room to breath.

Did you miss the rest of my post? As discussed, 1680 is not low res. http://www.anandtech.com/bench/Product/53?vs=147&i=47.48.49.50

Not bad for a chip they don't even make anymore. I'd say it competes just fine. Even more so for it's replacement. http://www.anandtech.com/bench/Product/89?vs=147&i=47.48.49.50.59.60.61.62

Again, Phenom II was decent when compared to phenom I, which was just inexcusably awful. They needed to add 2 more cores to actually compete with Intel's worst chips.

 

[Melvis]

Did you miss the rest of my post? As discussed, 1680 is not low res. http://www.anandtech.com/bench/Product/53?vs=147&i=47.48.49.50

Not bad for a chip they don't even make anymore. I'd say it competes just fine. Even more so for it's replacement. http://www.anandtech.com/bench/Product/89?vs=147&i=47.48.49.50.59.60.61.62

Again, Phenom II was decent when compared to phenom I, which was just inexcusably awful. They needed to add 2 more cores to actually compete with Intel's worst chips.

If your talking just purely about gaming benchmark, then its a no brainier that a X6 isnt the best choice (that's why i got a quad core), games at this stage wont use the full power of 6 cores, but when it comes to everything else the 1055T will eat that Q6600 alive, and in the future games will need more and more cores so the X6 will only increase its lead in time. http://www.guru3d.com/article/phenom-ii-x6-1055t-1090t-review/10

And once again the 9950 is = to a Q6600 at stock clocks, so they do not need two more cores to catch up at all. Even the 965 runs better in games then my X6.

 

[LAN_deRf_HA]

What do you mean "if your talking purely about gaming", not only is this thread about gaming, you just said you bought AMD for gaming. Now you're saying the X6 isn't the best choice for gaming. Again, I'll bring up my point about how you basically have to lie to yourself to justify buying AMD at this point and time. If you wanted better application performance, you'd have gotten a 1156 for the same price and gotten radically better gaming performance as a bonus. I bought AMD when they were the best, they aren't now and haven't been for years. I'd encourage anyone supporting current AMD offerings above the $70 price point to reconsider their reasoning. I mean you just compared a Q6600 to a lower speed phenom 1. Phenom 1 had a severe clock for clock disadvantage, well outside of gaming. http://www.anandtech.com/bench/Product/53?vs=23

This isn't news. That was a well accepted fact at the time, as it should still be now. Reason would not lead someone to buy a phenom 1. You'd have to allow yourself to view things simply as you wished they were to justify that purchase.

 

[Melvis]

What do you mean "if your talking purely about gaming", not only is this thread about gaming, you just said you bought AMD for gaming. Now you're saying the X6 isn't the best choice for gaming. Again, I'll bring up my point about how you basically have to lie to yourself to justify buying AMD at this point and time. If you wanted better application performance, you'd have gotten a 1156 for the same price and gotten radically better gaming performance as a bonus. I bought AMD when they were the best, they aren't now and haven't been for years. I'd encourage anyone supporting current AMD offerings above the $70 price point to reconsider their reasoning. I mean you just compared a Q6600 to a lower speed phenom 1. Phenom 1 had a severe clock for clock disadvantage, well outside of gaming. http://www.anandtech.com/bench/Product/53?vs=23

This isn't news. That was a well accepted fact at the time, as it should still be now. Reason would not lead someone to buy a phenom 1. You'd have to allow yourself to view things simply as you wished they were to justify that purchase.

Well if your comparing a X6 against anything else for gaming then that's your fault in the first place for bringing up the X6 compared to a X4 for gaming performance as a X4 seems to be the better choice for gaming as i explained above^. Lie to myself? i didn't have to do any such thing, i bought both my CPU's over 8months ago, and for the price performance they gave out it was unbeatable at that time. See now i don't get why your bringing up newer sockets when this is all about older sockets compared to the 2600K ONLY. You are the one saying and i quote> I mean even a stock Q6600 gives a X6 a challenge.

Just a reminder that the Phenom 1 that i posted was at the time the fastest AMD quad, at 2.6GHz there for the so called "severe" clock for clock disadvantage was only 200MHz. http://www.guru3d.com/article/amd-phenom-x4-9950-be-processor-tested/1

I would agree with you there it would not of been the greatest choice to get a Phenom 1, this is why i held of till Phenom II came along and it proved to be a very worthy contender giving any C2Q a very good run for its money, and in most cases better performance per $$. I must say that iam comparing the Prices for over here in AUS as intel is ALOT more expensive here then AMD is.

 

[BababooeyHTJ]

IMO, both high resolutions and low resolutions are important. Low resolutions take the bottleneck off the GPU and place it on the CPU, so it highlights the real performance difference between CPUs. However, higher resolution place the bottleneck back on the GPU, and show what real world usage will be like.

Considering 1680x1050 is the second most common resolution on monitors today, it certainly is important that it is included in benchmarks, it gives people with these monitors an idea of what performance they can expect, and actually it is people with these "lower" resolutions that will see the most noticeable difference between CPUs. Oh, and just FYI, 1280x1024 is the 3rd most common monitor resolution, so that too is important in benchmarks as well, and again they will show a larger difference between CPUs than a higher resolution would.

I agree with you but how many people in the market for a $300 cpu like 2600k will be using 1680x1050. What I don't like is Anand using Medium settings. I'll tell you right now that if anything most of the settings in FO3 are mostly cpu intensive on a modern video card.

 

[newtekie1]

I agree with you but how many people in the market for a $300 cpu like 2600k will be using 1680x1050.

A lot, a lot of people don't think that is a bad resolution. And a lot of people will be upgrading their computer, but keeping their LCD that they already have, and really why not do it this way? The monitor is still very good, and if you are someone like me, who spent a pretty penny on their 1680x1050 monitor, they will probably keep it if it still works. Heck, all my monitors are 1680x1050 or less with the exception of my main machine.

 

[Crap Daddy]

For me the best screen is the 22" with 1680x1050, not too large, not too small. Perfect. It also allows you to go for max settings with GPUs that are not quite the most poweful in the world. And a good CPU adds to a great gaming experience. So I think this particular res still has life in it and deseves to be benchmarked.

 

[BababooeyHTJ]

A lot, a lot of people don't think that is a bad resolution. And a lot of people will be upgrading their computer, but keeping their LCD that they already have, and really why not do it this way? The monitor is still very good, and if you are someone like me, who spent a pretty penny on their 1680x1050 monitor, they will probably keep it if it still works. Heck, all my monitors are 1680x1050 or less with the exception of my main machine.

And you leave all of the settings on medium while you are at it?

 

[newtekie1]

And you leave all of the settings on medium while you are at it?

I wasn't disagreeing with your on that point, hence why I didn't address it or even include it in the quote. Why would you assume that just because I disagree with part of your post that I disagree with it all? That doesn't make any sense.

 

[yogurt_21]

No, that doesn't make any sense at all. Who is buying a $400-500 cpu+motherboard combo to game on a ~$120 monitor on low or medium settings? Those comparisons are about as useful as any other synthetic benchmark. I gloss right over those reviews.

Secondly as I said saying that the cpu don't have any effect on resolution and IQ settings is just ignorant.

Lastly, where is a four year old 2.4ghz quad core cpu giving a 3.2ghz (not including turbo boost) hex core cpu a run for it's money? Maybe on Anandtech's reviews. Phenom 2 is at worst on par with Yorkfield clock for clock and in some cases can outperform Nehalem.

*raises hand* now I'm certainly on all maxed out settings, just the lower resolution.

some of us want an actual monitor upgrade not just a small resolution bump while still ending up with a crappy tn panel.

the cheapest ips panel that meets my requirements is 500$ and that's cash I don't have right now, so I'm still on my 4.5 year old 1680x1050 monitor.

besides take a look at the resolutions in megapixels and there's really a small difference between 1680x1050 at 1.76 megapixel, and 1920x1080 2 megapixel.

resolution megapixels
640x480 0.31
800x600 0.48
1024x768 0.79
1440x900 1.29
1280x1024 1.31
1650x1080 1.76
1600x1200 1.92
1920x1080 2.07
1920x1200 2.30
2048x1152 2.36
2560x1440 3.68
2560x1600 4.1

you only really get a big jump if you skip up to the uber resolutions.

 

[BababooeyHTJ]

I probably shouldn't have even have bought up the 1680x1050 part, it's really not that low. It's mostly the low IQ settings that bothers me about anand. They also used to use a lower resolution in their cpu reviews (when I decided to stop paying attention to them) not too long ago IIRC.

 

[newtekie1]

I probably shouldn't have even have bought up the 1680x1050 part, it's really not that low. It's mostly the low IQ settings that bothers me about anand. They also used to use a lower resolution in their cpu reviews (when I decided to stop paying attention to them) not too long ago IIRC.

Again, the lower resolution, besides still being extremely popular and hence important, also shows the real difference in performance between the CPUs. That is why lower resolutions and lower settings are used in CPU reviews. If people are reading a CPU review, they want to see how much actual difference there are between CPUs. Quite frankly, if you only talk about large resolution and maxed out graphics settings, then my celerons would show little difference compared to my i7.

 

[LifeOnMars]

Again, the lower resolution, besides still being extremely popular and hence important, also shows the real difference in performance between the CPUs. That is why lower resolutions and lower settings are used in CPU reviews. If people are reading a CPU review, they want to see how much actual difference there are between CPUs. Quite frankly, if you only talk about large resolution and maxed out graphics settings, then my celerons would show little difference compared to my i7.

Absolute tosh :shadedshu Frame rates at any resolution are ruled by the lowest common denominator. If the CPU is slower than the GPU at higher resolution you are bottlenecked by the CPU, and vice-versa. Therefore a celeron compared to an i7 is still gonna show a big difference in alot of games, even at high res.

It's very true that the gap is reduced between CPU's at higher resolutions, normally though, this is an artificial representation because you are hitting GPU bottlenecks first.

 

[EastCoasthandle]

Here is some food for thought

A 5850 at 1Ghz was used to test a Q9550 vs i7 860 both at 4.00GHz. The results can be found here. If I recall correctly a 860 was faster then a 920. Here are some results using a QX9650 vs 920 at stock and overclocked. But this time using dual GPUs.

 

[LifeOnMars]

In that first test - GPU bottlenecked, plain and simples. Get a more powerful GPU paired with the CPU, or even a pair of GPU's and the better CPU would shine and come into its own. Crap test.

 

[EastCoasthandle]

I think it's clear that having a good CPU is beneficial when gaming. So, if anyone has the age old question, "Do I buy a CPU or a GPU?" If it's an SB CPU then get that 1st then get the GPU later.