something I've been curious about. What is the technology difference in the Cascade Lake-X and Coffee Lake, like the IPC, etc. I just got the Core i9 10940-X and I was able to overclock it to 5.0GHz easily, without even increasing the voltage manually. Core for core, does it match the 9900K? Obviously with multithreaded apps it will kick it's a$$. But what about core for core?
-
Leakers revise RTX 5090 and RTX 5080 power draw to 575W and 360W respectively — Up to 27% higher than last-generation RTX 40 GPUs
Soaptrail
Distinguished
- Jan 12, 2015
- 303
- 96
- 19,420
Also how is a Ryzen 3700X nine places higher than the 3800X in single threaded performance since it is based on test results not on price?
Last edited:
Soaptrail
Distinguished
- Jan 12, 2015
- 303
- 96
- 19,420
@PaulAlcorn How does this make sense that the Ryzen 3800X is nine places lower than a 3700X when you are not using price in your calculation?
Last edited:
Soaptrail
Distinguished
- Jan 12, 2015
- 303
- 96
- 19,420
How is a Ryzen 3700X nine places higher than the 3800X in single threaded performance since it is based on test results not on price?
cryoburner
Judicious
- Oct 8, 2011
- 5,773
- 1,394
- 34,840
Is there an echo in here? : P
Something does look wrong with those numbers though. The 3700X scores a 94.7% in the single-threaded chart, while the 3800X scores an 83.0%. Isn't the whole point of the 3800X that it's a slightly faster 3700X, not a significantly slower one? Other benchmarks I've seen have typically shown the 3800X performing a couple percent faster, not well over 10% slower. Some of the other numbers in that chart don't look quite right either, like the 2400G outperforming the 2600X at single-threaded performance. I would expect even the 1600X to be a little faster.
Were all these processors tested using the same methods, or were numbers just gathered from various reviews running different benchmarks on different test systems? I could see there being some anomalies between processors from different architectures and generations depending on what software is being tested, but not so much processors from the same generation, where higher-clocked models are performing below lower-clocked ones.
Something does look wrong with those numbers though. The 3700X scores a 94.7% in the single-threaded chart, while the 3800X scores an 83.0%. Isn't the whole point of the 3800X that it's a slightly faster 3700X, not a significantly slower one? Other benchmarks I've seen have typically shown the 3800X performing a couple percent faster, not well over 10% slower. Some of the other numbers in that chart don't look quite right either, like the 2400G outperforming the 2600X at single-threaded performance. I would expect even the 1600X to be a little faster.
Were all these processors tested using the same methods, or were numbers just gathered from various reviews running different benchmarks on different test systems? I could see there being some anomalies between processors from different architectures and generations depending on what software is being tested, but not so much processors from the same generation, where higher-clocked models are performing below lower-clocked ones.
- Feb 24, 2015
- 876
- 394
- 19,360
- Feb 24, 2015
- 876
- 394
- 19,360
Thanks for pointing that out. i will dig into the results to see what the disparity is here, I fear this may be a data entry error.
Soaptrail
Distinguished
- Jan 12, 2015
- 303
- 96
- 19,420
Amazing. Everyone who knows the tech agrees that Ryzen is the best choice for all except pro gamers and day traders. But somehow Toms found enough cherry-picked data to show the opposite - with Intel dominating the top of table after table of partial info.
Those less informed WILL be misled - no question. But ... Intel will be very pleased. Their "Meet Comp" pressure and support for misinformation is paying off. You've presented Intel, to the general reader, as the leader in almost every category! You should be ashamed.
Those less informed WILL be misled - no question. But ... Intel will be very pleased. Their "Meet Comp" pressure and support for misinformation is paying off. You've presented Intel, to the general reader, as the leader in almost every category! You should be ashamed.
Last edited:
The received wisdom is that milliseconds can matter in receiving info or placing such trades. So a few MHz faster would be worth any price or compromise in thermals etc. The same is supposedly true for pro gamers. Until now, Intel has had the fastest single-core chips - though not by much, and made on outdated silicon and running far above their stated power. Tom's leaves that part out.
Tom's also happens to omit the new Ryzen 3xxx XT chips, which bring single-core scores up even with Intel. At lower power and lower cost per core. All that is left out too.
That's why Intel will be pleased with this article. Their product line - which is overall completely (sometimes laughably) inferior to AMD's - looks like the best! What more could Intel ask?
This has been going on ever since Intel process development hit the skids and they began scrambling for ways to appear constantly ahead, while actually being constantly further behind. Every month or so, Tom's delivers, in the form of a crazily biased "round-up" or "comparison" or "survey" piece like this one.
Wouldn't the important part here be the actual network. So a trader is going to want to be on a Fiber connection over DSL/Cable because of lower latency.
This hierarchy really showcases how outdated these types of ranking are. everyone involved in the tech enthusiast space knows how zen is dominating and a new generation is months away yet toms uses older lightly threaded games on clean installs of windows with zero software in the background using threads like antivirus rbg controls monitor controls streaming and other things that take cpu resources before you get to gaming and makes it appear as if intel is the king. they are not you will not realize these gains.
also these games.. im playing need for speed payback, its several years old and uses 16 threads at 64% on a 2700x. 16!!
have a bunch of crap running in the background, get new games, and have 25 tabs open then test again, id be interested in that hierarchy.
you could even call it a "real world" performance hierarchy.
dont be a subtle userbenchmark 2.0. id like to think toms is impartial but sometimes...
also these games.. im playing need for speed payback, its several years old and uses 16 threads at 64% on a 2700x. 16!!
have a bunch of crap running in the background, get new games, and have 25 tabs open then test again, id be interested in that hierarchy.
you could even call it a "real world" performance hierarchy.
dont be a subtle userbenchmark 2.0. id like to think toms is impartial but sometimes...
Blas
Distinguished
- May 18, 2007
- 730
- 8
- 20,165
Actually, they don't even use Fiber, because for what they use networks, there are faster ways: radio! I recently watched this documentary (in French, sorry!) that explains how the fast traders bought and maintain old radio antennae as close as possible to the datacenters where the actual trading is made (not Wall Street, that's just the old building where nothing interesting happens anymore) to gain those few ms they need to do their thing. Quite an interesting documentary to watch, hope you can find it in a language you can follow, in case French isn't fit.
In any case, for them every ms counts, so having a slightly faster computer is probably also worth it, considering how much they
Here is the price vs gaming performance scatter chart. (At July 2020)
I wanted to show multithreaded performance as bubble size, but because they used different labels on each table, it was made really difficult.
Note that the price is on logarithmic scale.
I wanted to show multithreaded performance as bubble size, but because they used different labels on each table, it was made really difficult.
Note that the price is on logarithmic scale.
jefferzbooboo
Distinguished
- Jun 21, 2012
- 38
- 17
- 18,545
Everyone in the tech enthusiast space knows that running additional programs in the background when doing benchmarks doesn't produce reliable results that can be reproduced, well except for you I guess.
johnwkuntz
Honorable
- Jan 11, 2018
- 40
- 2
- 10,535
Avro Arrow
Splendid
- Nov 12, 2009
- 4,280
- 1,440
- 26,090
This comment related to:
https://www.tomshardware.com/reviews/best-cpus,3986.html
The article's publication date is 08/30/2020 which is 1 day before this post.
Jeez, Chris I don't know how you manage to keep your hair (assuming that you do). Just ONE day passes and the prices of the i5-10600K and R5-3600X on this list are already wrong. That's one volatile market. I honestly don't know where your listed prices come from because I can't find any 10600K's for under $300 and I can't find any 3600X's for over $215 anywhere. Also keep in mind, the i5-10600K doesn't come with a cooler which drives its effective price up by at least $30 (Average Air Cooler Cost) making it even worse for the money. It should not be #1 with such a narrow performance delta and such a wide price delta.
This list:
i5-10600K - $299
R5-3500X - $249
Newegg:
i5-10600K - $330
R5-3600X - $215
Amazon:
i5-10600K - $317
R5-3600X - $209
Micro Centre:
i6-10600K - $300
R5-3600X - $200
I think that the i5-10600K should be knocked out of the #1 spot as "Best Gaming CPU for the Money" by the R5-3600X because even at 5GHz, the i5 is only 15% faster when both are OC and only 5% faster when both are at stock settings. However, the cost delta is 50% at Micro Centre, 52% at Amazon and 54% at Newegg which makes the i5-10600K a terrible value when compared to the R5-3600X. The cost delta that you have listed is only 20% which does make the i5 the better deal but that number is completely out to lunch so it renders your ranking completely invalid.
I just wanted to point this out before someone else starts screaming about you being some kind of Intel shill because your numbers are so far off (and they are wildly in Intel's favour). I've been a member here long enough to know that you're fair and impartial but I just wanted to warn you that this may cause a storm of complaints. I can't comment on the other CPUs in the list because the 10600K and the 3600X are the only ones that I find interesting ATM.
https://www.tomshardware.com/reviews/best-cpus,3986.html
The article's publication date is 08/30/2020 which is 1 day before this post.
Jeez, Chris I don't know how you manage to keep your hair (assuming that you do). Just ONE day passes and the prices of the i5-10600K and R5-3600X on this list are already wrong. That's one volatile market. I honestly don't know where your listed prices come from because I can't find any 10600K's for under $300 and I can't find any 3600X's for over $215 anywhere. Also keep in mind, the i5-10600K doesn't come with a cooler which drives its effective price up by at least $30 (Average Air Cooler Cost) making it even worse for the money. It should not be #1 with such a narrow performance delta and such a wide price delta.
This list:
i5-10600K - $299
R5-3500X - $249
Newegg:
i5-10600K - $330
R5-3600X - $215
Amazon:
i5-10600K - $317
R5-3600X - $209
Micro Centre:
i6-10600K - $300
R5-3600X - $200
I think that the i5-10600K should be knocked out of the #1 spot as "Best Gaming CPU for the Money" by the R5-3600X because even at 5GHz, the i5 is only 15% faster when both are OC and only 5% faster when both are at stock settings. However, the cost delta is 50% at Micro Centre, 52% at Amazon and 54% at Newegg which makes the i5-10600K a terrible value when compared to the R5-3600X. The cost delta that you have listed is only 20% which does make the i5 the better deal but that number is completely out to lunch so it renders your ranking completely invalid.
I just wanted to point this out before someone else starts screaming about you being some kind of Intel shill because your numbers are so far off (and they are wildly in Intel's favour). I've been a member here long enough to know that you're fair and impartial but I just wanted to warn you that this may cause a storm of complaints. I can't comment on the other CPUs in the list because the 10600K and the 3600X are the only ones that I find interesting ATM.
Last edited:
Can someone more knowledgeable explain something to me...?
On these charts, and other comparisson articles, it is repeated that for gaming threads are not that important, and more i Clock speed.
Then when talking about older CPUs, it says yo ucan get a "substantial improvement" by upgrading your CPU to a new platform.
How exactly is that obtained? For example, I have an old 4790K, which has "just" 4 cores / 8 threads, but base freq. is 4GHz with 4.4GHz boost, stock.
How exactly would a newer processor with similar clock give me a "substantial boost"...?
On these charts, and other comparisson articles, it is repeated that for gaming threads are not that important, and more i Clock speed.
Then when talking about older CPUs, it says yo ucan get a "substantial improvement" by upgrading your CPU to a new platform.
How exactly is that obtained? For example, I have an old 4790K, which has "just" 4 cores / 8 threads, but base freq. is 4GHz with 4.4GHz boost, stock.
How exactly would a newer processor with similar clock give me a "substantial boost"...?
- May 29, 2013
- 38,221
- 3,624
- 156,790
Faster RAM, more efficient CPU, so more instructions per clock (IPC).
Look at this page -- https://www.cpubenchmark.net/compare/Intel-i7-4790K-vs-AMD-Ryzen-5-3600X/2275vs3494
The single core performance of the "slower" 3600x is higher than the 4790K
The total performance is biased by the higher number of cores.
- Oct 9, 2020
- 1
- 0
- 10
Is this a joke? It does NOT need an update for Zen 3 since those pricessors haven't been released yet.
How do you expect them to be benchmarked and added to the chart when they're not yet available?
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
-
-
Discussion PSU recommendations and power supply discussion thread - Tom's hardware- Started by Darkbreeze
- Replies: 2K
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
-
Facebook
Twitter
Instagram