Price/Performance Cores

[Fuzzykins]

That was the chip i always wanted when i was on my A64 3200+ system. $1031 release price though. I don't recall how many cores it had. Was it 2, or 4? I'm sure it was 2, but i genuinly can't find an answer on it (it has been 6 years since release).



http://www.xbitlabs.com/articles/cpu...0k-990x_3.html

269 for the 990x and 290 for the 2600k. Can't find an AATech score for the 990x. Dividing my 290 by your 274 gives me 1.058394160583942. So, a fair 990x score would be 254 (actual: 254.1586206896551). Price on the 990x at newegg is exactly $1000.

Erm, could you put that a little clearer?
I think I already have those in there. >.>

[Deleted]

Didn't look into new CPUs so I might be wrong but i7 920 tend to be cheap and rather powerful especially after OC.

[Cilraaz]

You're free to interpret the results of this whichever way you want. Personally, I'm more astonished that the 955 and 965 had more points per $ than the i5 and i7.

I'm not sure why that's surprising. Intel is performance. AMD is budget. It's been this way for a few years.

The i5 2500K costs 41.5% more and performs 29.3% better than the PII x4 965. Like any product, it's paying a premium to get more. If you factor in overclocking, though, it gets a lot closer. My 4.6GHz i5 2500K is 35.3% higher clocked than a stock 2500K (3.4GHz with Turbo Boost). If you assume linear performance, it would score around 358, which is 74.6% than the PII x4 965 for a 41.5% cost premium. Now, if you account for a moderate overclock on the i5 2500K, you have to account for a moderate overclock on the 965 also... say 4.0GHz? Assuming what we assumed with the i5 2500K, that's a 17.6% speed increase, or a score of 241. At that point, you're looking a differential of 48.5% performance for a 41.5% cost premium. That pretty much evens them up.

Then again, this whole theorycraft is bullshit because performance isn't too likely to be linear. So we'd need someone with a 4.0GHz PII x4 965 to run a benchmark against my system. Then we'd be assuming that there's nothing hindering performance on either system. Essentially, it could go in circles for a long time.

[Asera]

I'm not sure why that's surprising. Intel is performance. AMD is budget. It's been this way for a few years.

The i5 2500K costs 41.5% more and performs 29.3% better than the PII x4 965. Like any product, it's paying a premium to get more. If you factor in overclocking, though, it gets a lot closer. My 4.6GHz i5 2500K is 35.3% higher clocked than a stock 2500K (3.4GHz with Turbo Boost). If you assume linear performance, it would score around 358, which is 74.6% than the PII x4 965 for a 41.5% cost premium. Now, if you account for a moderate overclock on the i5 2500K, you have to account for a moderate overclock on the 965 also... say 4.0GHz? Assuming what we assumed with the i5 2500K, that's a 17.6% speed increase, or a score of 241. At that point, you're looking a differential of 48.5% performance for a 41.5% cost premium. That pretty much evens them up.

Then again, this whole theorycraft is bullshit because performance isn't too likely to be linear. So we'd need someone with a 4.0GHz PII x4 965 to run a benchmark against my system. Then we'd be assuming that there's nothing hindering performance on either system. Essentially, it could go in circles for a long time.

I can tell you right now, overclocking a K10.5 does not yeild the same amount of gain as overclocking an SB chip, or even a Nehalem chip. I'm sure you knew that though.

[Cilraaz]

I can tell you right now, overclocking a K10.5 does not yeild the same amount of gain as overclocking an SB chip, or even a Nehalem chip. I'm sure you knew that though.

Yeah, but for the sake of argument, I had to assume both would improve linearly. I have no clue of what kind of improvement curve each chip actually would show.