CatMerc
Golden Member
- Jul 16, 2016
- 1,114
- 1,153
- 136
Didn't you mention earlier that there are no major revisions beyond A0? They're all A0x."New Horizon" demos ran on a system with B1 stepping CPU in it.
Didn't you mention earlier that there are no major revisions beyond A0? They're all A0x."New Horizon" demos ran on a system with B1 stepping CPU in it.
Didn't you mention earlier that there are no major revisions beyond A0? They're all A0x.
Didn't you mention earlier that there are no major revisions beyond A0? They're all A0x.
Have you guys considered the possibility of 8C/16T clocking to 4.3 GHz as rumored on all cores and staying within 95W TDP target?
No, since we know that 3.4Ghz already makes it cut dangerously close to the 95W power consumption in a light load.
The Stilt's predictions were all in line with the data available at the time. The data has changed, right from AMD themselves. The data keeps changing, and will become more solid at CES in two days. What's the problem? He's not to blame.
Also, "allegedly" i am devil incarnate.
Sometimes people are wrong. It happens.
As I said before, there are no Ax parts. Just A0x. Therefore you should advance in the alphabet
If someone says there is, he is either lying or being lied to.
A process that's dubbed low power should be a low power process. Its scaling should favor lower-power schemes for good efficiency. Otherwise it's not named well I'd say. If it indeed can get the rumored clocks maybe it should be called a medium power process or broad power or something.Doom2pro said:14LPP doesn't require low power and isn't limited to low clocks, as others have said, you can hit the efficiency sweet spot designed for 14LPP or go outside that range to get higher clocks at the cost of higher power.
Wrong. Cinebench doesn't use AVX or FMA instructions.
Must be an empowering feeling to you?
Enjoy it while it lasts.
A process that's dubbed low power should be a low power process. Its scaling should favor lower-power schemes for good efficiency. Otherwise it's not named well I'd say. If it indeed can get the rumored clocks maybe it should be called a medium power process or broad power or something.
Where did i say that it use thoses instructions..?..
I said that Cinebench was "updated" (from 11.5 to R15) to give Intel an advantage that couldnt be so big without those instructions, and this despite CB not using them, that is, it s telling how much it s biaised..
Heck ,even the two renderers used by Hardware.fr (within 3DS max) and wich are using those instructions dont display more than 15 and 18% advantage for the 4770K.
By looking AT's results for 4770K & 8370, going from R11.5 to R15 slowed them both down pretty equally.
58.055s (R11.5), 62.208s (R15) - FX-8370 = 7.153% slower
47.003s (R11.5), 50.568s (R15) - i7-4770K = 7.584% slower
In R15 FX-8370 performs better in relation to 4770K, than in R11.5. Don't you agree?
Are you sure that you took CB CPU scores..?.
Because they are not in seconds but in points, FX8350/4770K scores are 6.92/8.18 and 640/791 in CB 11.5 and R15 respectively.
I am, but obviously that won't change anything.
R11.5 score calculation = 400 / render time
R15 score calculation = 40000 / render time
R11.5 = 6.89, R15 = 643 - FX-8370 (AT)
R11.5 = 8.51, R15 = 791 - i7-4770K (AT)
Good to know. This page is the first time I've seen it mentioned.That was LPE
A process that's dubbed low power should be a low power process. Its scaling should favor lower-power schemes for good efficiency. Otherwise it's not named well I'd say. If it indeed can get the rumored clocks maybe it should be called a medium power process or broad power or something.
Perjaps...Or perhaps polaris was just a temporary console biproduct on 14lpe funded by Sony only used untill vega will come on 14lpp. Perhaps vega will later come in more than the two big sizes that we asume?
For 14nm, the wafer cycle time alone is around 90 days; then the wafers need to be shipped to the packaging plant where they are cut into dies, tested/binned and packaged. Then the CPUs are shipped to the distributors, which then ship to retailers. So that's probably closer to 4 months between the decision for volume production and first product being available to end customers. I wouldn't be surprised if AMDs "shipping in Q1" means shipping to distributors, with retail availability sometime in April.From tapeout to finished product it takes 3 months at least, so the final stepping must be in production now. A wafer does not take zero time to produce, quite the opposite.
4770K is supposed to run at 3.7GHz on when all four cores are stressed.
I could argue that there's no difference. In R11.5 FX-4300 is about 14% faster than i3-4130T. In R15 it is also 14% faster than i3-4130T. Conclusion, no bias towards Intel. But what we could get from these results is that cache size matters for R15. X4 845 has half the L2 per core compared to FX-4300. In R11.5 845 is about 6% faster than FX-4300, but in R15 it is exactly the opposite(6%). So half the L2 hurts X4 845 for about 12% in R15. Roughly the difference between your comparison between R15 and R11.5 results.Certainly but the 791 CB R15 score is at 3.53GHz, and comparaisons should be make at same clocks, at 3.5GHz for the 4770K the difference is 24% with CB R15 and 17.8% with CB R11.5..
So it s obvious that CB R15 was updated to give Intel an advantage, and your comparison is indeed wrong since you compare the FX at 4GHz in both benches to a 4770K at 3.5 for 11.5 and 3.7GHz for R15, and then state that the difference is the same with the two benches...
There s a 2.9GHz 2C/4T HW without boost in the charts below, you also have the FX 4C, Kaveri and XV, you can also see how CB R15 did compress the improvements in those AMD CPUs compared to CB 11.5 :
21% advantage for the XV Athlon 845 over the i3 4130T in CB 11.5 and 7.5% in CB R15, and we have people using the latter as "reference" to measure IPC, lol....
Where did i say that it use thoses instructions..?..
I said that Cinebench was "updated" (from 11.5 to R15) to give Intel an advantage that couldnt be so big without those instructions, and this despite CB not using them, that is, it s telling how much it s biaised..
That's a fallacy: all you can deduce from your initial claim is that the software characteristics are not the same.Certainly but the 791 CB R15 score is at 3.53GHz, and comparaisons should be make at same clocks, at 3.5GHz for the 4770K the difference is 24% with CB R15 and 17.8% with CB R11.5..
So it s obvious that CB R15 was updated to give Intel an advantage, and your comparison is indeed wrong since you compare the FX at 4GHz in both benches to a 4770K at 3.5 for 11.5 and 3.7GHz for R15, and then state that the difference is the same with the two benches...
It's just one rendering benchmark, and there are many other rendering engines that are more used and that don't show the same characteristics. So I wouldn't say it "reflects real world".No issue since Cinebench is a nice benchmark that reflects real world.
The update decreased Intels advantage slightly, but it was very minor. No issue since Cinebench is a niche benchmark that reflects real world. There is no favour for either Intel or AMD, it's a good
average value. You can be sure if AMD releases a competitive µarch in the future it will show up in Cinebench.