Isn't it true that for gaming one does not need more than 4 cores, you better off investing in a good videocard + I5 or Ryzen5? Streamers and video editing would benefit from 4+ cores, but the average player Threadripper is just overkill?
For the average gamer, a Core i5 is overkill and 95% of what they do, they wouldn't notice a bit if they dropped to a Core i3 (or AMD). But that hasn't stopped people from buying (or recommending) Core i7's.
I have to disagree with that. Depends on what type of games are you talking about. I play a lot of strategy games and many of those are multi threaded and getting more so. I got a letter from Stardock just a month ago saying they are now using software that will take advantage of as many cores as you have. Their contention was that the new Ryzen chips were a good investment if you like playing their games.
Would you define yourself as "average", or assume that the typical player plays a lot of strategy games by Stardock?
I don't know if AMD will have a straight CPU that's competitive, but you still need a GPU to do anything gaming-wise with that Pentium. Add in a $100 GPU, and your up around $165. AMD hasn't introduced their Zen lineup that is intended to compete in the budget arena yet.
The Ryzen 5's aren't competitive there, but they aren't really meant to be. I would expect a Ryzen-based APU to be competitive though - something with, say, a couple of Zen cores and a few Vega GPU units, could be extremely competitive with Intel's inexpensive Pentium, if you look at total system cost.
Seeing how underwhelming Rx 560 is, the chance any APU will be any faster than $100 GTX 1050 is unlikely.
And like I said, I cannot fathom what CPU performance they can provide for $65 - there does seem to be any room left.
@Ridelynn were you thinking of the i3 7350k as the gaming desktop chip for single core performance? Or something else?
Actually I was thinking of the Pentium G4560, because it's less than half the cost of an i3 7350. Moving that thinking to an i3, especially the 7350k, makes it a bit different, because Intel charges a lot of money (relative) for the Core brand in that tier, but the Pentium is at a nice price point (vs performance).
The i3 7350k doesn't make a lot of sense, because your right up against a Core i5 price tag there. The i3 7300 or 7100 would be the more appropriate benchmark - they considerably less than the 7350k, the clock doesn't go down that fast versus the price (but all are still roughly twice the price of the G4560) . The major difference between the i3 and Pentium (apart from Kaby vs Haswell) being clock speed and IGP.
But yes - because of single-core performance, the 2c/4t Intel chips are adequate for gaming purposes today, when paired with a GPU. I'm not claiming they win at any benchmark contests or futureproof against anything, but that "typical" games available today, under "typical" resolutions (1080p and lower), the low end Intel chips are fast enough to get you 60FPS in most games under at least Medium settings. For ~most~ people, they would be good enough.
The same thing can be said for an FX-x3xx though (not that it has superior single-threaded performance, just that it's adequate).
And I have no idea how a Ryzen/Vega APU will change that calculus, other than to say it has some promise, is supposed to be announced "soon", and details are not available yet.
the word AMD and redefining high end desktop performance go together like oil and water.
Italian dressing, Milk, and Mayo are all pretty tasty.
Now, I don't necessarily think AMD will redefine anything - but they are at least making Intel dance to their tune, which is impressive considering exactly how little market share AMD holds.
My opinion is that Intel isn't so much worried about AMD Ryzen per say. Intel knows they missed the boat on Mobile/SOC chips, and their cash cows (PC and Server) are what is floating the ship right now. Those cash cows aren't growth markets, so Intel has to find some way to grow in order to keep stock prices up (because that's how Wall St. works, if you aren't growing your dying) - that's why we've seen Intel push into (and out of) a lot of weird markets lately (physics via Havok, AV via McAfee, graphics via Larrabee, Storage via Optane, etc)
The new high-growth platforms (at least in so far as chip manufacturers are concerned) appear to be moving to Deep Learning/AI specific chips aimed at two markets: large cloud/datacenter use (Google, Azure, AWS, etc), and oddly enough, automotive (Tesla, Ford, etc) for the self-driving push. Their biggest competition there right now is nVidia, which is aggressively pursuing it. But AMD has all the right pieces to also be an aggressive player there, if they had some capital to push into R&D and marketing for that purpose.
So Intel is trying to preserve their market share in their cash cow markets right now (the lesser reason), and simultaneously deny allowing a potential rival to emerge and challenge them in an emerging market (the greater reason).
Pumping out some tech they've had on the burner for a while (KL-X, X299, more cores, etc) isn't exactly difficult for Intel, they'd already done all the R&D required, all they needed was a reason to produce the product. AMD gave them a reason, and Intel is jumping on it, not because they think all of these new SKUs with crazy core counts will sell in huge numbers (although that wouldn't disappoint them), or that they think it will really push computing forward, or that they think AMD is going to come in with 16-core chips and decimate their existing Core iX sales, but because it creates great press releases.
Wall St. reads that PR, and Intel's stock price will react accordingly. Intel doesn't have to sell a single chip, but if their stock price goes up by $1, that's the same as putting $4.7B in the bank. Once a company goes public, you really don't care directly about sales/revenue/profit so much as you care about how those things affect your stock price, because the entire worth of your company shifts to revolve entirely around your stock price.
And if you believe any of the AMD/Intel merger/cross-licensing/etc rumors (which Intel did deny, but that doesn't necessarily mean there isn't some truth to it), then that adds another layer to this interesting story.
I don't know if AMD will have a straight CPU that's competitive, but you still need a GPU to do anything gaming-wise with that Pentium. Add in a $100 GPU, and your up around $165. AMD hasn't introduced their Zen lineup that is intended to compete in the budget arena yet.
The Ryzen 5's aren't competitive there, but they aren't really meant to be. I would expect a Ryzen-based APU to be competitive though - something with, say, a couple of Zen cores and a few Vega GPU units, could be extremely competitive with Intel's inexpensive Pentium, if you look at total system cost.
Seeing how underwhelming Rx 560 is, the chance any APU will be any faster than $100 GTX 1050 is unlikely.
And like I said, I cannot fathom what CPU performance they can provide for $65 - there does seem to be any room left.
Your post does not make much sense...
He makes a very reasonable post and you go off into who knows where and make a puzzling comment? He is talking about a combined CPU/GPU of which Intel has nothing to compete with such. Intel's CPU/GPU chips are the laughing stock of the industry. Sure you can add in a $100 to offset that, but if it is already on the AMD chip you are having to spend extra with the Intel version.
The laptop makers are all looking forward to a combined Ryzen/Vega chip as that will eliminate the need for a separate GPU card in all but the high end laptops.
I just ignore Gdemami for the most part and let him click LOL on everything. Thanks for the Epic badge buddy!
I suspect he doesn't even realize he's talking Polaris and not Vega, that for some inexplicable reason nVidia just released what appears to be a pretty useless 1030 that is in this niche as well, and that there is a very large and well defined market for modern AMD APUs that is poised to jump from Console to Desktop, and with Ryzen/Vega instead of Jaguar/Polaris in there could do so very well.
Comments
And like I said, I cannot fathom what CPU performance they can provide for $65 - there does seem to be any room left.
Your post does not make much sense...
The i3 7350k doesn't make a lot of sense, because your right up against a Core i5 price tag there. The i3 7300 or 7100 would be the more appropriate benchmark - they considerably less than the 7350k, the clock doesn't go down that fast versus the price (but all are still roughly twice the price of the G4560) . The major difference between the i3 and Pentium (apart from Kaby vs Haswell) being clock speed and IGP.
But yes - because of single-core performance, the 2c/4t Intel chips are adequate for gaming purposes today, when paired with a GPU. I'm not claiming they win at any benchmark contests or futureproof against anything, but that "typical" games available today, under "typical" resolutions (1080p and lower), the low end Intel chips are fast enough to get you 60FPS in most games under at least Medium settings. For ~most~ people, they would be good enough.
The same thing can be said for an FX-x3xx though (not that it has superior single-threaded performance, just that it's adequate).
And I have no idea how a Ryzen/Vega APU will change that calculus, other than to say it has some promise, is supposed to be announced "soon", and details are not available yet.
Now, I don't necessarily think AMD will redefine anything - but they are at least making Intel dance to their tune, which is impressive considering exactly how little market share AMD holds.
My opinion is that Intel isn't so much worried about AMD Ryzen per say. Intel knows they missed the boat on Mobile/SOC chips, and their cash cows (PC and Server) are what is floating the ship right now. Those cash cows aren't growth markets, so Intel has to find some way to grow in order to keep stock prices up (because that's how Wall St. works, if you aren't growing your dying) - that's why we've seen Intel push into (and out of) a lot of weird markets lately (physics via Havok, AV via McAfee, graphics via Larrabee, Storage via Optane, etc)
The new high-growth platforms (at least in so far as chip manufacturers are concerned) appear to be moving to Deep Learning/AI specific chips aimed at two markets: large cloud/datacenter use (Google, Azure, AWS, etc), and oddly enough, automotive (Tesla, Ford, etc) for the self-driving push. Their biggest competition there right now is nVidia, which is aggressively pursuing it. But AMD has all the right pieces to also be an aggressive player there, if they had some capital to push into R&D and marketing for that purpose.
So Intel is trying to preserve their market share in their cash cow markets right now (the lesser reason), and simultaneously deny allowing a potential rival to emerge and challenge them in an emerging market (the greater reason).
Pumping out some tech they've had on the burner for a while (KL-X, X299, more cores, etc) isn't exactly difficult for Intel, they'd already done all the R&D required, all they needed was a reason to produce the product. AMD gave them a reason, and Intel is jumping on it, not because they think all of these new SKUs with crazy core counts will sell in huge numbers (although that wouldn't disappoint them), or that they think it will really push computing forward, or that they think AMD is going to come in with 16-core chips and decimate their existing Core iX sales, but because it creates great press releases.
Wall St. reads that PR, and Intel's stock price will react accordingly. Intel doesn't have to sell a single chip, but if their stock price goes up by $1, that's the same as putting $4.7B in the bank. Once a company goes public, you really don't care directly about sales/revenue/profit so much as you care about how those things affect your stock price, because the entire worth of your company shifts to revolve entirely around your stock price.
And if you believe any of the AMD/Intel merger/cross-licensing/etc rumors (which Intel did deny, but that doesn't necessarily mean there isn't some truth to it), then that adds another layer to this interesting story.
The laptop makers are all looking forward to a combined Ryzen/Vega chip as that will eliminate the need for a separate GPU card in all but the high end laptops.
I suspect he doesn't even realize he's talking Polaris and not Vega, that for some inexplicable reason nVidia just released what appears to be a pretty useless 1030 that is in this niche as well, and that there is a very large and well defined market for modern AMD APUs that is poised to jump from Console to Desktop, and with Ryzen/Vega instead of Jaguar/Polaris in there could do so very well.