Same exact hardware. The only difference is 32-bit WIndows 8.1 vs 64-bit Windows 8.1
The difference is OS on the same HW is 54% single core and 65% multicore.
IMO, any GB scores within 50% of each other are suspect.
Geekbench 3 Singlecore Score 1659 2567
Geekbench 3 Multicore Score 2775 4570
"[And I can assure you that Ash has used benchmarking to blatantly distort real world performance when it suits his needs.]"
Through "intention" or "ignorance", everyone is guilty of this offense. EVERYONE.
"it looks like core m will throttle:"
It may be a function of the chip or it may be a function of the SW that can set the multiplier to extend battery life (depending on initial BIOS settings and OS configurations). Core M is designed and configured for longer battery life so it will quickly throttle. I think the throttle behavior can be modified by changing the MSR.
I have seen my i7 4C/8T Sandy Bridge, Ivy Bridge and Haswell desktops throttle down from 3.4/3.5 GHz down to as low as 1.6GHz for periods of time even while under load (may be temperature protection). My i7-4790K CPU @ 4.00GHz Devil's Canyon never drops below 4.0GHz.
Single comparisons based on "best we have" benchmarks without knowing the conditions can lead to very wrong conclusions.
It is the best we have BUT "information" should be extracted from the "data" carefully.
"While I agree on both points, GB3 is the best native benchmark we have (even with it's many faults). BTW, some of the tests are modern (lua)."
Yep. I know your position. I also appreciate that you periodically point out GB shortcomings.
"Assuming battery and thermal load, they would have spent the full time turboed"
Have you personally monitored the turbo state of an Intel CPU during the tests? I have seen some interesting down throttles on other Intel CPUs under my workloads. While not disputing what you are saying, I remain skeptical.
"...but still it is better than nothing."
As long as people don't confuse it with being "something".
"Like Ashraf is an authority."
Do you think that someone needs to be an "authority" to post a Geekbench3 comparison? I think not.
Is the comparison "good" or "bad" for "who"??? You should look.
Meaningful benchmarks are HARD and Geekbench3 is one I have little respect for. It is a collection of old (some obsolete) open source micro-benchmarks that are then collected together. IMO, one can only glean very gross levels of information about the two systems being compared. MUCH CARE should go into drawing general conclusions. Geekbench is "inelegant".
Were they both plugged in or operating on battery? ....
The comparison shows a 0.988GHz Core M 2-core, hyperthread compared with a 1.50GHz Apple A8 3-core device. The Core M turbos to 2.0GHz but it is not clear how much time, if any, was spent turboed. Win81 vs iOS schedulers .... how many times was the test moved to another core for temperature control?
The single Core M core is faster on Geekbench3 at 0.988GHz that the A8 core at 1.50GHz. I am surprised that a single slower Core M core (running Windows 8.1) is faster.
Three A8 cores at 1.50GHz are faster than 2 Core M cores with hyperthreading at 0.988GHz.
You think your guess is better than your guess on Saturday for Monday's results?
Markets will tank tomorrowitakebackmyapology by itakebackmyapology • Oct 18, 2014 12:56 PM
I am 100% for certain
"Now, AD- post a link to an actual vote where Republicans voted to cut CDC funds. It didn't happen."
I don't remember giving any credit to either political party for the action. However, if you want your proof, then I think you can do a better job getting information you like than I can.
"Cherry picking numbers is the same as lying."
.... like choosing the "beginning of the Bush administration" instead of the last 10 years? 8-)
"Finally, please, stop initiating political topics. I don't want to respond, but feel obliged. "
I agree. But not because I am a Dem or Rep. I usually get called names from both sides. There are still some people between the political extremes.
"There have been no real budget cuts; just slower spending increases. "
There are probably some programs where what you say is true. The CDC is not one of them. It probably has not helped that the NRA has put a "hold" on the confirmation of the Surgeon General.
The 2003 CDC Program Budget was $6,755,534.000.
The 2013 CDC Program Budget was $6,291,000,000.
That seems like a $464 million budget decrease over the last 10 years.
The buying power of the US dollar has also dropped by 29% over the same 10 years.
"A vaccination will soon be developed by big Pharma to deal with the epidemic."
The vaccination will probably be priced so it will be widely available to the poor in west Africa.
"The one person who got in Africa did die because he went without care for too long."
I think it is less about African vs American and more about finances. Only one person did not have health insurance. Thomas Eric Duncan was in Texas and did not have health insurance. The other patients have health insurance. The doctors and hospitals will be reimbursed. Texas Health Presbyterian Hospital Dallas expected to not be reimbursed for the Duncan visit, they just did not expect it to cost $1 million for the Duncan treatment.
"simple program +, -, x, /, ^ and sqr()"
I would be interested in looking at the loop or an equivalent proxy loop to see the data dependencies if you have source .... or application you talking about if you can share it. 544x performance difference seems like a little extreme unless the RISC is an ASIC.
"Do you actually do anything with these Intel processors, or do ya just know the technicals?"
Yes and somewhat.
"What's the fastest Intel processor on the market that costs less than $25,000."
The answer would depend on the characteristics of workload you are running.
"I've run outta gas on this stupid RISC workstation and need something that can fly."
RISC machine must be really old or 32-bit. Makes sense that you are running out of gas.
"I noticed a HUGE difference between these Intel chips and the RISC stuff, like 544x faster than an Intel i7-2.46. "
See answer above. It depends on your workload. 544x faster sounds goofy but you already knew that ... being a genius and all. 8-)
"Intel Reports Record Quarterly Revenue of $14.6 Billion"
Earnings was not a disappointment. Even those with low valuations (Covello-types) said it was good BUT could not continue. It was the "Yes, but..." analysis. The "good this quarter, BUT" analysis it their only defense.
"Don't act surprised that this filthy stock got pounded."
Intel is YTD up 20% more than the market.
A periodic pounding is fine when I am 20% up over other investments.
"$28? why is $28 a target"
.... because $28 is 10% less than the Intel price today. If Intel were to fall to $28, the $28 target would become $25.
In reality, someone who is "long" thinks that $28 means $28.00. To the person spreading the negative target, $28 means $28.99. Once the stock price hits the $28 handle, they start their "I'm a genius!" dance.
The current 200-day moving average is $28.80 and rising. IF ..... Intel were to trade down to its 200-day moving average, it would "hit $28, through one set of eyes and .... not through the other."
8.. I thought $20.00 and 4% yield was a good price.
"Save yourself the embarrassment, dump it secretly before anybody finds out"
How can anyone be embarrassed while laughing at your humor. lol
"Covello has a 49% success rate recommending stocks with an average return of -1.5% per recommendation. He has recommended Intel 16 times with a success rate of 20%."
-1.5% average return per recommendation ??????????????????????
This is a stunning report card. A bull market makes even the densest rock a "genius". Covello is worse than a coin flip.
"... and it didn't need a reverse split, did it Alex?"
Nope, no reverse split. I just guaranteed that SQQQ would get to $40 one way or another. 8-)
Congrats on your good gains and timing. Your multiple similar threads .... looks like you are pleased with your good fortune too.
Always glad to everyone get ahead.
His and his apology guy posts are near useless for any purpose other than to scroll messages off the first page. It was a good guess that INTC would drop 5% and then it happened. Intel is usually down 5% the day before during the conference call so finishing up after hours was a big change.