- Jul 24, 2006
- 22,856
- 277
- 57,490
-
Leakers revise RTX 5090 and RTX 5080 power draw to 575W and 360W respectively — Up to 27% higher than last-generation RTX 40 GPUs
- Status
[citation][nom]randomizer[/nom]Rather than posting so many comments, maybe you should keep it all in one. That way we keep the bullcrap to a minimum.[/citation]
Rather than posting so many comments, maybe you should keep it all in one. That way we keep the bullcrap to a minimum.
which clearly illustrates you have nothing to say, and that
You sir, are an idiot. really amazing. thank you for listening to yourself. You didn't really want to do that did you?
Rather than posting so many comments, maybe you should keep it all in one. That way we keep the bullcrap to a minimum.
which clearly illustrates you have nothing to say, and that
You sir, are an idiot. really amazing. thank you for listening to yourself. You didn't really want to do that did you?
- Jul 24, 2006
- 22,856
- 277
- 57,490
You know what would help this situation. Invite both Intel and AMD to bring their recommended configuration for benchmarking. Then all this complaining will disappear. I am absolutely sure that both Intel and AMD would love to participate directly with their CPU, recommended MOBO’s and RAM
zodiacfml
Distinguished
- Oct 2, 2008
- 1,228
- 26
- 19,310
i think, i missed a part. how come there's no athlon x2 in there?
amd only did well in the phenom in the memory bandwidth benchmarks. it is a true quad core chip only lacking in clock speed.
guys out there, do a little effort to find the prices. just focus on applications you would use most,its quite easy to narrow it down.
amd only did well in the phenom in the memory bandwidth benchmarks. it is a true quad core chip only lacking in clock speed.
guys out there, do a little effort to find the prices. just focus on applications you would use most,its quite easy to narrow it down.
[citation][nom]randomizer[/nom]*Yawn* Troll. Go play with Thunderman.[/citation]
You troll - or is it "just ok" when you do it?
(You sir, are an idiot.) Is this useful and productive? Thank you for demonstrating!! OK - all in one post. watch this, sir.
====
"In order to ensure that our testing procedure is suitable for coming generations of processors"
IS THIS why you use a year old mobo?
"which were put to the test using a fresh gauntlet of benchmarks"
WHATEVER was wrong with previous benchmarks? Real world conditions do not even involve benchmarks. I never use them. Do you? If you replace them are previous results nullified? Modified? Will you replace these as well by posting an /unbiased cpu list\?
"An X-Fi Xtreme Gamer sound card from Creative takes the pressure off the processor when it comes to audio calculations and ensures reproducible results."
This sounds fair; but in real world, this overlooks problems intel or AMD might have with their sound too. At any rate, not real world results - scores probably inflated.
How much "pressure" was "taken off" and how does that affect your results? You don't think this matters? Real world matters.
"Very few benchmarks are able to handle more than one processing core"
Why would you even bother saying this? And since nearly all cpu's are multicore, are all these benches useless therefore?
"Intel is now ahead in just about every category."
This also from your introduction. Do you detect any bias here? This statement is utterly FALSE!!!! Especially in real world conditions. How can you test without bias, if this is your approach?
"AMD is having trouble keeping up with Intel’s upper echelon of performance-oriented solutions and its current processor models offered only minor overclocking potential up until recently, when the SB750 southbridge improved scalability on Black Edition Phenoms."
And so you eliminate the REAL STORY. That's very biased!
New "see-through" bias. People will believe what they want. I believe truth and have some respect for professionaly presented legitimate testing. Bias is either deliberate or is introduced in ignorance - either way, it is seriously misleading and a serious disservice to manufacturers, sys builders, and especially users. Better to let the word of mouth do it all. When you contaminate the word of mouth, you create results such as my comments and the many other comments of disapproval. You are fooling yourself and hurting people.
You troll - or is it "just ok" when you do it?
(You sir, are an idiot.) Is this useful and productive? Thank you for demonstrating!! OK - all in one post. watch this, sir.
====
"In order to ensure that our testing procedure is suitable for coming generations of processors"
IS THIS why you use a year old mobo?
"which were put to the test using a fresh gauntlet of benchmarks"
WHATEVER was wrong with previous benchmarks? Real world conditions do not even involve benchmarks. I never use them. Do you? If you replace them are previous results nullified? Modified? Will you replace these as well by posting an /unbiased cpu list\?
"An X-Fi Xtreme Gamer sound card from Creative takes the pressure off the processor when it comes to audio calculations and ensures reproducible results."
This sounds fair; but in real world, this overlooks problems intel or AMD might have with their sound too. At any rate, not real world results - scores probably inflated.
How much "pressure" was "taken off" and how does that affect your results? You don't think this matters? Real world matters.
"Very few benchmarks are able to handle more than one processing core"
Why would you even bother saying this? And since nearly all cpu's are multicore, are all these benches useless therefore?
"Intel is now ahead in just about every category."
This also from your introduction. Do you detect any bias here? This statement is utterly FALSE!!!! Especially in real world conditions. How can you test without bias, if this is your approach?
"AMD is having trouble keeping up with Intel’s upper echelon of performance-oriented solutions and its current processor models offered only minor overclocking potential up until recently, when the SB750 southbridge improved scalability on Black Edition Phenoms."
And so you eliminate the REAL STORY. That's very biased!
New "see-through" bias. People will believe what they want. I believe truth and have some respect for professionaly presented legitimate testing. Bias is either deliberate or is introduced in ignorance - either way, it is seriously misleading and a serious disservice to manufacturers, sys builders, and especially users. Better to let the word of mouth do it all. When you contaminate the word of mouth, you create results such as my comments and the many other comments of disapproval. You are fooling yourself and hurting people.
[citation][nom]Area51[/nom]You know what would help this situation. Invite both Intel and AMD to bring their recommended configuration for benchmarking. Then all this complaining will disappear. I am absolutely sure that both Intel and AMD would love to participate directly with their CPU, recommended MOBO’s and RAM[/citation]
I appreciate your point Area51.
The problem is with too many sites. The people testing simply don't know or don't care or they do know and are lying. But some of the most trusted sites are lying; and it takes someone pointing things out before people start to see what is going on.
VERY FEW people will believe that intel just is not as good as it is made out to be. AMD might be behind in top top top crown peformance; but the rest of the line is amazingly equal or better. This turns into a pissing contest and little more. In real world conditions, the story is very different. And that's not only about price/perf ratio.
Simply most benchmarks hide the truth, or simply do not illustrate the truth. Most benchmarks are HAND PICKED for the reason the intel will look good. Other benchmarks are written for intel cpu's, compiled with intel compilers, and some have been known to have built in cpu "detours" for "other brands". Then in testing there are many scams for seeming honest, while fooling the reader. And then there are also mis-interpretations. Bias going into a test is visible here. This is one of the worst I have ever seen. Good lesson.
I appreciate your point Area51.
The problem is with too many sites. The people testing simply don't know or don't care or they do know and are lying. But some of the most trusted sites are lying; and it takes someone pointing things out before people start to see what is going on.
VERY FEW people will believe that intel just is not as good as it is made out to be. AMD might be behind in top top top crown peformance; but the rest of the line is amazingly equal or better. This turns into a pissing contest and little more. In real world conditions, the story is very different. And that's not only about price/perf ratio.
Simply most benchmarks hide the truth, or simply do not illustrate the truth. Most benchmarks are HAND PICKED for the reason the intel will look good. Other benchmarks are written for intel cpu's, compiled with intel compilers, and some have been known to have built in cpu "detours" for "other brands". Then in testing there are many scams for seeming honest, while fooling the reader. And then there are also mis-interpretations. Bias going into a test is visible here. This is one of the worst I have ever seen. Good lesson.
Several possible errors on the AMD cpu table. Not only is the 5400+BE wrong, so is the 5000+BE. Mine is definitely a Brisbane. Its not a huge deal.
I would also like to see some of the older AMD CPUs, at least the ones at or above 2.4GHz seems fair, it doesn't even need to be all of them, but some comparison points would have been nice.
As other have said, prices would make the charts much more informative, but since they change so often, it would be futile. What would be nice would be a system of relative cost, but since the other hardware is always the same, you couldn't really include them since it wouldn't make sense for anyone to own a $50 cpu and a $200 motherboard and $200 memory and a $500 GPU (unless they were really trying to OC).
That does kind of prove a point about how the charts really become something that should really not be taken at face value. Its only useful as a general ordering (I never look at the exact position of anything on these charts, too many factors). The test system have nothing in common with real world system except for the very top end.
I do appreciate the work that goes into these charts, but I think that setting some budgets and building balanced systems to fit into those budgets is generally more useful. Maybe start at $500 and work up to $2000 in some increment (maybe $200 per step), list alternatives or maybe even build the alternatives. People should be able to fit any reasonably balanced configuration somewhere between the steps (or just build one of the recommended systems).
Keep up the work on the charts. I still use them as a general meter when looking for a new cpu or graphics card.
I would also like to see some of the older AMD CPUs, at least the ones at or above 2.4GHz seems fair, it doesn't even need to be all of them, but some comparison points would have been nice.
As other have said, prices would make the charts much more informative, but since they change so often, it would be futile. What would be nice would be a system of relative cost, but since the other hardware is always the same, you couldn't really include them since it wouldn't make sense for anyone to own a $50 cpu and a $200 motherboard and $200 memory and a $500 GPU (unless they were really trying to OC).
That does kind of prove a point about how the charts really become something that should really not be taken at face value. Its only useful as a general ordering (I never look at the exact position of anything on these charts, too many factors). The test system have nothing in common with real world system except for the very top end.
I do appreciate the work that goes into these charts, but I think that setting some budgets and building balanced systems to fit into those budgets is generally more useful. Maybe start at $500 and work up to $2000 in some increment (maybe $200 per step), list alternatives or maybe even build the alternatives. People should be able to fit any reasonably balanced configuration somewhere between the steps (or just build one of the recommended systems).
Keep up the work on the charts. I still use them as a general meter when looking for a new cpu or graphics card.
lunyone
Splendid
- Feb 22, 2006
- 6,517
- 0
- 26,960
I for one could care less about a $1500 CPU. 99% of builders won't even consider a CPU over $300, it isn't in the budget. I think I've only seen a couple of good reviews from several different sites that didn't seem too biased. I'm not an expert, but I can generally see where things aren't quite as equally as whatever review site depicts them. I would love to see reviews on a more equal or at best, as equally matched as possible. I'm pretty sure why review sites write the way they do, but if they at least made a neutral effort for both parties, they would at least earn respect from their readers.
I would like to see a review written with something along these lines:
First review topic: Sub $100 CPU w/Sub $100 GPU's (varying brands WITH UPDATED DRIVERS)
next test: Sub $100 CPU w/$100-$200 GPU's (varying brands WITH UPDATED DRIVERS)
next test: Sub $100 CPU w/$300-$xxx GPU's (varying brands WITH UPDATED DRIVERS)
Than up the CPU to $100-$150 with above GPU's
Than up the CPU to $150-$200 with above GPU's
and so on....You get the point.
This would shine some light on what kind of ACTUAL performance that you could get with a said CPU w/a said GPU. You pick the combo that you like the most (within your budget) and off you go to the store!!!
Now I'm sure most review sites won't do this, because it would show that you don't need a $1500 CPU to get similar performance out of a $200 CPU with said GPU. You get my point here. I think Tom's did something along these lines several months ago, but it was just a minimal attempt.
Now I would be betting that a AMD x2 5400+ BE w/a 4870 would have similar results as a e5200 w/a 4870. Now they are both priced near each other and would probably perform pretty close at stock speeds, assuming your using similar resolutions. I think you get where I'm going with this too.
I would like to see a review written with something along these lines:
First review topic: Sub $100 CPU w/Sub $100 GPU's (varying brands WITH UPDATED DRIVERS)
next test: Sub $100 CPU w/$100-$200 GPU's (varying brands WITH UPDATED DRIVERS)
next test: Sub $100 CPU w/$300-$xxx GPU's (varying brands WITH UPDATED DRIVERS)
Than up the CPU to $100-$150 with above GPU's
Than up the CPU to $150-$200 with above GPU's
and so on....You get the point.
This would shine some light on what kind of ACTUAL performance that you could get with a said CPU w/a said GPU. You pick the combo that you like the most (within your budget) and off you go to the store!!!
Now I'm sure most review sites won't do this, because it would show that you don't need a $1500 CPU to get similar performance out of a $200 CPU with said GPU. You get my point here. I think Tom's did something along these lines several months ago, but it was just a minimal attempt.
Now I would be betting that a AMD x2 5400+ BE w/a 4870 would have similar results as a e5200 w/a 4870. Now they are both priced near each other and would probably perform pretty close at stock speeds, assuming your using similar resolutions. I think you get where I'm going with this too.
Thank you very much for the charts! It must have taken a very long time to test all those chips. Wow.
Regarding the Amd Vs. Intel, I was surprised to see how competitive AMD is. For some time I have read all about how Intel is doing, but the phenom 9950 competes very well (perhaps it even does better) than Intel Q6600, priced about the same. However, since amd chipsets and ddr2 ram are much cheaper, including very descent built-in mobo gfx (unless you are a gamer), the phenom 9950 is actually a much better choice. Unless you want to overclock, then maybe q6600 would be better.
Thanks again!
Regarding the Amd Vs. Intel, I was surprised to see how competitive AMD is. For some time I have read all about how Intel is doing, but the phenom 9950 competes very well (perhaps it even does better) than Intel Q6600, priced about the same. However, since amd chipsets and ddr2 ram are much cheaper, including very descent built-in mobo gfx (unless you are a gamer), the phenom 9950 is actually a much better choice. Unless you want to overclock, then maybe q6600 would be better.
Thanks again!
haider_up32
Distinguished
- Oct 2, 2008
- 2
- 0
- 18,510
What about AMD Athlon X2 6000+ (3.1GHz, 1Mb cache, 89W)?
These processors have been out for quite some time...
These processors have been out for quite some time...
chaohsiangchen
Distinguished
- Jul 28, 2008
- 479
- 0
- 18,780
[citation][nom]sighQ2[/nom]But the SouthBridge is a 690=OLD/ The test would be better with the NEWER M3A79-T Deluxe. [/citation]
Like I said, the project needs time to proceed. Tom's staff can't change motherboard half way through the test, otherwise we will never see an update.
And again, test has been done comparing 790FX/SB600 with 790GX/SB750, and 790FX/SB600 won 2 out of 3 memory test, while 790GX/SB750 won just a little bit on memory write test. Winrar score shows that 790FX/SB600 won clearly over 790GX/SB750.
http://www.sharkyextreme.com/hardware/motherboards/article.php/10703_3764251__3
790GX northbridge is basically 780G on steroid. Nothing specially about that northbridge. Southbridge doesn't improve CPU test score on stock speed. It only matters to overclock.
If Tom's want to do a Phenom/Q6600 rematch, they need to get SB750 boards. For CPU chart based on stock speed, that board is OK. You won't get magical 20% improvement by switching to newer boards, and the difference will be limited to within 5%.
Like I said, the project needs time to proceed. Tom's staff can't change motherboard half way through the test, otherwise we will never see an update.
And again, test has been done comparing 790FX/SB600 with 790GX/SB750, and 790FX/SB600 won 2 out of 3 memory test, while 790GX/SB750 won just a little bit on memory write test. Winrar score shows that 790FX/SB600 won clearly over 790GX/SB750.
http://www.sharkyextreme.com/hardware/motherboards/article.php/10703_3764251__3
790GX northbridge is basically 780G on steroid. Nothing specially about that northbridge. Southbridge doesn't improve CPU test score on stock speed. It only matters to overclock.
If Tom's want to do a Phenom/Q6600 rematch, they need to get SB750 boards. For CPU chart based on stock speed, that board is OK. You won't get magical 20% improvement by switching to newer boards, and the difference will be limited to within 5%.
Everyone knows you don't compare +$500 chips with subs: Apples to oranges and completely diff clientèle. People buying 9770s and the like are doing SERIOUS work (aside from rich tards)-work which DEMANDS an edge over the competition. Not just some guy sitting in his room gaming and claiming fps ownage over others. So dollar for dollar in the sub $300 range is actually more representative of REAL-WORLD users' applications: games, rar'ing, encoding files etc. So 100 vs $100, 200 vs $200, 300 vs $300 is the key chart ceiling.
For $100-300 AMD and Intel are sorta close, BUT Intel HAS the definitive roadmap that backs them: AMD's is shaky and speculative at best. Factor in most enthusiasts (anyone even reading these reviews) will probably OC also, which Intel completely dominates at. Thus I will stick with the E8400/Q6600/Nehalem series for now. Phenom is a nice upgrade if you've already got a newer AMD chipset, but for a NEW purchase; it's fools gold... anything less is just blatant denial.
For $100-300 AMD and Intel are sorta close, BUT Intel HAS the definitive roadmap that backs them: AMD's is shaky and speculative at best. Factor in most enthusiasts (anyone even reading these reviews) will probably OC also, which Intel completely dominates at. Thus I will stick with the E8400/Q6600/Nehalem series for now. Phenom is a nice upgrade if you've already got a newer AMD chipset, but for a NEW purchase; it's fools gold... anything less is just blatant denial.
I would like to see the "Price/Performance Index" again. That was always my main reason to use tomshardware for my benchmark comparison site.
Makes you wonder if the same people crying about this chart would be silent or smarmy if it were the AMD CPU's dominating. Even if TH would appease your every request it wouldn't make the charts much different if any different. Get over it. AMD has some good stuff coming and I imagine they will be extremely competitive beyond the price comparison. We all know what the reality right now is and the charts are just a visual of what we already know. I don't understand the discontent.
jitpublisher
Splendid
- May 16, 2006
- 8,747
- 2
- 31,960
Intel, Intel, Intel, Intel, Intel, Intel, Intel, AMD, Intel, Intel, Intel, Intel, Intel,Intel, Intel, AMD, AMD, Intel, Intel, Intel, Intel, Intel, Intel Intel, Intel, AMD, Intel, Intel, Intel, Intel, Intel, Intel, Intel, Intel, AMD, Intel, Intel, Intel, Intel, Intel, Intel, Intel, Intel.
Give us a break. AMD fanboi or Intel idiot, it does not take much to figure out the bias of this article. Completely worthless read.
Give us a break. AMD fanboi or Intel idiot, it does not take much to figure out the bias of this article. Completely worthless read.
biometriccoder
Distinguished
- Oct 3, 2008
- 2
- 0
- 18,510
I really love these charts when I'm looking to upgrade. However, this time it would appear that the list was a little short changed. In addition to missing the entire line of X2s, it is also missing several of intel's newer low cost quads. Namely, the Q9300, Q9400, and the Q8200. Since these all fit my low-cost niche, I would really like to see how they stack up against the big boys. Also, you got the E7200, but missed the E7300.
Are there any plans to go back and these newer procs? I believe that most, if not all of them were released in Q3.
Are there any plans to go back and these newer procs? I believe that most, if not all of them were released in Q3.
i m not sure what is the exact parameter used in linux compilation tests, while compiler are not really multi-threaded, the gcc make tool can be used to launch parallel jobs making better use of multicore processors using the -j option.
Let's see some benchmarks that matter like how long it takes to get to a usable desktop in various OS's. How about some multitasking benchmarks. Honestly, who cares how many seconds it takes to encode a video? (besides professional) Go grab a pop out of the fridge and make a sandwich. Seriously, I don't understand why there are so many useless benchmarks. My Pentium Mobile laptop running at 600Mhz (on battery power) seems to magically run Office, watch DVDs, surf the net, play the occasional lame game. Are you seriously telling me that the billions and billions spent by Intel and AMD are so you can get more frames per second on Crysis? OK, OK it's Toms Hardware I know, but maybe the whole industry needs a reality check. My desktop is a x2 3800, bought three years ago I think. I have a $25 graphics card. My 100GB old hard drive boots Vista in about 45 seconds. I can do everything except of course waste time on 3D games. I'm using DDR 1 RAM, not DDR2. I notice no difference in performance whenever RAM is compared on Toms, now they want me to move to DDR3??? Give me a compelling reason to send one more dollar to China so I can build the latest greatest machine which will : Play DVDs, run Office, surf the net, play music, organize family photos. The real crime isn't that people spend thousands on overpiced Macs, its that they spend hundreds of dollars on regular upgrades. Give me something compelling, like real AI that can use lots and lots of cores! Why do I have to tell internet explorer that I mean blah.com whey I typed blah.cmo? I look forward to future benchmarks having AI
i have a phenom x4 9750, and i would have bought a q6600 if i had known the difference at the time. i'm not biased towards one or the other, but the methodology in this test was flawed. no matter who beat who, tom's took a big hit in reputation by not showing neutrality and fairness. even to the casual observer, this test wasn't fair. the intel setup had unfair advantages (ddr3 RAM, for example) which makes the comparison charts nearly useless. i don't need to know if the qx can beat the phenom. i want to know how much more performance i can get by spending X. by not being fair, this chart is unfairly skewing customers towards intel (even though they do deserve the business.) if i had to choose between the q6600 and the 9750 again, these charts would push me towards the q6600, even though i know (now) that the 9750 holds it's own and even beats the q6600 in many situations (with even a 200mhz minimal overclock.)
hawk66
Distinguished
- Oct 4, 2008
- 8
- 0
- 18,510
For the average Joe, These tests are meaningless. I am still playing all of the new games on my AMD 2500+ and 6600GT. Did I fail to recognize spysweeper and Norton on the test setup. Why spend $500 on an Intel Quad for gaming when the AMD 6500 will do the same and keep the inside of the case cool. Cool cases matter.
- Status
TRENDING THREADS
-
Question New build wont post Gigabyte B650M Gaming Plus wifi , AMD Ryzen 5 7600X CPU, 32GB T-Force RGB DDR5. Need help!
- Started by Notmiserable
- Replies: 27
-
News TSMC's wafer pricing now $18,000 for a 3nm wafer, increased over 3X in 10 years: Analyst- Started by Admin
- Replies: 31
-
Discussion PSU recommendations and power supply discussion thread - Tom's hardware- Started by Darkbreeze
- Replies: 2K
-
Question Need Help Building a $1500 pc for Editing, Design, and Blender
- Started by Abdullah127
- Replies: 4
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
-
Question Which graphics card should I be looking at for my specific needs?- Started by LewisMann
- Replies: 1
-
[SOLVED] CPU speed slows down once unplugged and running only on battery.- Started by Dalleuh
- Replies: 10
Facebook
Twitter
Instagram