Slashdot Mirror

Intel chipsets may be weak compared to current generation AMD and NVidia hardware, but compared to the cards of a decade ago they're pretty powerful.

Agreed. A previous Intel CPU (Ivy Bridge with HD 4000) runs Skyrim playably at 720p according to Anandtech. This puts it at least at parity with the PlayStation 3, which also runs Skyrim. In turn, because so many AAA PC games prior to 2013 were also released for PS3 and/or Xbox 360, they should have settings that scale down to PC hardware comparable in performance to those consoles.

They didn't, IT World got it wrong.

Intel responded to Anandtech's inquiry into the killing of the chip line and denies that it is dead and in fact is wondering where the bad information has come from:

http://www.anandtech.com/show/9639/the-death-of-intels-broadwell-is-greatly-exaggerated-socketed-broadwell-continues

Not true. The 5870 is still a capable GPU that is on par with a 750 ti. http://www.anandtech.com/bench...

It is widely known that DX12 will reduce draw call overhead, making weaker CPUs perform better relative to stronger CPUs. This is of course good for AMD, since they don't have high-end CPUs anymore though it's bit of a "scorched earth" result where gamers don't need expensive CPUs at all. But if you look at "Ashes (Heavy) DX11 to DX12 Scaling - Radeon R9 390X" and look at an extremely powerful CPU like the Core i7-6700 you're seeing 50-100% gains. If you're that severely bottle-necked by a 4+ GHz quad core then this is not a typical DX11 game.

We can compare the "typical" difference between a R9 390X and GTX980 in Anandtech's bench, though I have to substitute for a R290X "Uber" so the differences should actually be even smaller. Normally these cards are almost head to head, the question is not why DX12 is closing the gap but why there's such a huge DX11 gap to begin with. And the only reason I can come up with is because they're pushing way, way more draw calls than normal. Which may be DX12 enabling developers to do things they wanted to, but couldn't before or it could be to make someone look good/bad.

"developers can still use individual features if they’d like"

Third paragraph.

Feature sets are in addition to individual features. In any case, pushing the latest GPU hardware to the limit is the entire point of playing on a PC, instead of waiting several years for the next generation of console.

Reading AnandTech's review, they make a bold statement at the end:

"Sandy Bridge, Your Time Is Up."

That is an interesting thought, but is it really?

If you need USB 3, if you want some of the other newer chipset features, perhaps. But for performance?

In benchmarks, Skylake appears to be about 25% faster than Sandy Bridge. Sure, if you're doing video encoding all day or other CPU intensive applications, it is... (and if you ARE doing that stuff, why aren't you on Xeon?)

But for most desktop computer uses, you likely won't see any difference between the two. What is worse, is that most of the above gains came from Haswell, not Skylake.

http://www.anandtech.com/show/...

Look at the "Gains over Sandy Bridge" chart on that page. Look at the red lines, then the purple lines. The red lines are the Haswell gain over Sandy Bridge, then the purple lines are the Skylake gains over Sandy Bridge.

If we are talking MLC NAND, you are off by a factor of 10. http://www.anandtech.com/show/...

But since the XPoint cells are individually addressable bit by bit, the comparison is probably to SLC, in which case the relevant number of P/E cycles for XPoint would be 10E8 and you are off by a factor of more than 33,000.

Fury clearly beats 980 at 1440p and is closer to 980Ti-s than to 980 performance.
http://anandtech.com/show/9421...

What you call SSD now (both @ home & industrially - albeit, here writing up software in 1995 to do it for software ramdrives in 32-bit that did pretty ok - going a combination of that tech w/ hardware + software & massively LOW latencies in both in 2015 here):

Since 1990 w/ PCs it's the SINGLE best performance enhancement with the "best bang for the buck" there is as you notice it immediately:

Primary SSD = Intel 530 240gb Flash SSD (SATA 6) - OS & Program disk - latest 3.0 firmware & trim tools (vs. my WD Velociraptor -> http://www.anandtech.com/bench... )

* Used to use a CENATEK "rocketdrive" circa 2006-2010 before what's next below (PC-133 SDRAM 2gb unit on PCI 2.x bus).

Secondary "True SSD" = GigaByte IRAM 4gb DDR2-Ram based (SATA I) - for PageFile placement

Controller 4 Backup = Promise Ex-8350 128mb ECC ram caching controller (SATA 1/2) - for WD Velociraptor

RAM = 8gb Kingston DDR-3 (1gb for 64-bit NTFS Compressed Software RamDrive = webbrowser cache, hosts file, print spooler, %TEMP% ops, + %COMSPEC% location)

---

NTFS timestamps, all perf counters, & excess services off.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentations.

I place my custom hosts file on a software ramdisk by redirecting it in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via "DataBasePath" parameter - acts like a *NIX shadow password system)

I increased hosts' priority to its load/read too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008

APK

P.S.=> Lastly, changing a clock crystal/oscillator (?) in a 486SX/25mhz into a 33mhz one eventually had that system @ 133mhz via a Dx/4 Intel CPU & 8x the RAM of the original 1st system I had...

... apk

No theres three

WD/HGST
Samsung/Seagate
Toshiba

When WD bought out hitatchi's HDD buisness (which got renamed to HGST in the process) the regulators wouldn't allow them to keep the 3.5 inch drive part of the buisness as that would reduce the number of players to two. So that part of the buisness was sold to Toshiba (who already made 2.5 inch drives). http://www.anandtech.com/show/...

I would also note that having the same corporate overlord does not nessacerally imply having the same quality or lack thereof.

Samsung owns Seagate.

You got that backwards, Seagate bought samsung's HDD buisness.

In these form factors, it's no longer a question of peak CPU performance. These processors all thermal-throttle to the point where none of them are going to be performing at peak while in these form factors. The same i5 in the Surface will provide significantly more performance when in another form factor (like a NUC).

Of course Apple claimed they put a lot of effort into improving the sustained performance of the A8 over the A7, which already was better than other ARM chips. And it actually looks like they succeeded: http://www.anandtech.com/show/8666/the-apple-ipad-air-2-review/5

CPU = Intel Core I7 4790k (vs. my old CPU Core I7 920 -> http://www.anandtech.com/bench... )

Motherboard = ASUS B85-E

Video = EVGA/NVidia GeForce 970 GTX OC (+140mhz) 4gb GDDR5 RAM (vs. my old vidcard 470 GTX -> http://www.anandtech.com/bench... )

Primary SSD = Intel 530 240gb Flash SSD (SATA 6) - OS & Program disk - latest 3.0 firmware & trim tools (vs. my WD Velociraptor -> http://www.anandtech.com/bench... )

Secondary "True SSD" = GigaByte IRAM 4gb DDR2-Ram based (SATA I) - for PageFile placement

Storage HDD = Western Digital 7,200 rpm 8mb buffer 1tb HDD (SATA 6) - for downloads

Backup HDD = Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb HDD (SATA II) - for programming data

Controller 4 Backup = Promise Ex-8350 128mb ECC ram caching controller (SATA 1/2) - for WD Velociraptor

Burner = HP DVD+-RW Dvd 1265i (SATA 3)

RAM = 8gb Kingston DDR-3 (1gb for 64-bit NTFS Compressed Software RamDrive = webbrowser cache, hosts file, print spooler, %TEMP% ops, + %COMSPEC% location)

---

NTFS timestamps, all perf counters, & excess services off.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentations.

I place my custom hosts file on a software ramdisk by redirecting it in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via "DataBasePath" parameter - acts like a *NIX shadow password system)

I increased hosts' priority to its load/read too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008

* OS = Windows 7 64-bit fully patched + security "hardened" via CIS Tool (95% Score) + performance tuned.

APK

P.S.=> Used for development of APK Hosts File Engine 9.0++ SR-2 32/64-bit -> http://start64.com/index.php?o... & processing hosts using it...

... apk

CPU = Intel Core I7 4790k (vs. my old CPU Core I7 920 -> http://www.anandtech.com/bench... )

Motherboard = ASUS B85-E

Video = EVGA/NVidia GeForce 970 GTX OC (+140mhz) 4gb GDDR5 RAM (vs. my old vidcard 470 GTX -> http://www.anandtech.com/bench... )

Primary SSD = Intel 530 240gb Flash SSD (SATA 6) - OS & Program disk - latest 3.0 firmware & trim tools (vs. my WD Velociraptor -> http://www.anandtech.com/bench... )

Secondary "True SSD" = GigaByte IRAM 4gb DDR2-Ram based (SATA I) - for PageFile placement

Storage HDD = Western Digital 7,200 rpm 8mb buffer 1tb HDD (SATA 6) - for downloads

Backup HDD = Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb HDD (SATA II) - for programming data

Controller 4 Backup = Promise Ex-8350 128mb ECC ram caching controller (SATA 1/2) - for WD Velociraptor

Burner = HP DVD+-RW Dvd 1265i (SATA 3)

RAM = 8gb Kingston DDR-3 (1gb for 64-bit NTFS Compressed Software RamDrive = webbrowser cache, hosts file, print spooler, %TEMP% ops, + %COMSPEC% location)

---

NTFS timestamps, all perf counters, & excess services off.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentations.

I place my custom hosts file on a software ramdisk by redirecting it in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via "DataBasePath" parameter - acts like a *NIX shadow password system)

I increased hosts' priority to its load/read too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008

* OS = Windows 7 64-bit fully patched + security "hardened" via CIS Tool (95% Score) + performance tuned.

APK

P.S.=> Used for development of APK Hosts File Engine 9.0++ SR-2 32/64-bit -> http://start64.com/index.php?o... & processing hosts using it...

... apk

CPU = Intel Core I7 4790k (vs. my old CPU Core I7 920 -> http://www.anandtech.com/bench... )

Motherboard = ASUS B85-E

Video = EVGA/NVidia GeForce 970 GTX OC (+140mhz) 4gb GDDR5 RAM (vs. my old vidcard 470 GTX -> http://www.anandtech.com/bench... )

Primary SSD = Intel 530 240gb Flash SSD (SATA 6) - OS & Program disk - latest 3.0 firmware & trim tools (vs. my WD Velociraptor -> http://www.anandtech.com/bench... )

Secondary "True SSD" = GigaByte IRAM 4gb DDR2-Ram based (SATA I) - for PageFile placement

Storage HDD = Western Digital 7,200 rpm 8mb buffer 1tb HDD (SATA 6) - for downloads

Backup HDD = Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb HDD (SATA II) - for programming data

Controller 4 Backup = Promise Ex-8350 128mb ECC ram caching controller (SATA 1/2) - for WD Velociraptor

Burner = HP DVD+-RW Dvd 1265i (SATA 3)

RAM = 8gb Kingston DDR-3 (1gb for 64-bit NTFS Compressed Software RamDrive = webbrowser cache, hosts file, print spooler, %TEMP% ops, + %COMSPEC% location)

---

NTFS timestamps, all perf counters, & excess services off.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentations.

I place my custom hosts file on a software ramdisk by redirecting it in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via "DataBasePath" parameter - acts like a *NIX shadow password system)

I increased hosts' priority to its load/read too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008

* OS = Windows 7 64-bit fully patched + security "hardened" via CIS Tool (95% Score) + performance tuned.

APK

P.S.=> Used for development of APK Hosts File Engine 9.0++ SR-2 32/64-bit -> http://start64.com/index.php?o... & processing hosts using it...

... apk

The thing is, you're likely to want to in the future. With the most recent generation, Intel's Integrated graphics is actually better than AMD's best APU graphics.

http://www.anandtech.com/show/...

ASUS B85-E Motherboard

Intel Core I7 4790k CPU (vs. my last CPU Core I7 920 -> http://www.anandtech.com/bench... )

EVGA/NVidia GeForce 970 GTX video OC stock-oem (+140mhz) 4gb GDDR5 RAM (vs. my last vidcard 470 GTX -> http://www.anandtech.com/bench... )

Intel 530 240gb Flash SSD (SATA 6) - strictly OS & Program disk - latest 3.0 firmware & trim tools (vs. my
WD Velociraptor -> http://www.anandtech.com/bench... )

Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb (SATA II) - strictly for backup & programming data

Promise Ex-8350 128mb ECC ram caching raid sata 1/2 controller (SATA 1/2) - for backup WD Velociraptor

GigaByte IRAM 4gb DDR2-Ram based SSD (SATA I) - strictly for PageFile placement

Western Digital 7,200 rpm 8mb buffer 1tb (SATA 6) - strictly for downloads

HP DVD+-RW Dvd 1265i Burner (SATA 3)
8gb Kingston DDR-3 RAM (1gb for 64-bit NTFS
Compressed Software RamDrive = webbrowser cache, hosts file location, print spooler, %TEMP% ops, + %COMSPEC% location)

---

ALL disks = timestamp logical NTFS filesystem data = off + perf counters + excess services off & TCP/IP parameters = security & performance tuned.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentation on it.

I also place my custom hosts file onto it via redirecting where it's referenced by the OS in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via the "DataBasePath" parameter there which acts like a *NIX shadow password system)

I lastly increased hosts' priority assigned to its loads/reads too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008
"Name"="TCP/IP"

* OS = Windows 7 64-bit fully patched & performance tuned + security "hardened" via CIS Tool (95% SCORE).

APK

ASUS B85-E Motherboard

Intel Core I7 4790k CPU (vs. my last CPU Core I7 920 -> http://www.anandtech.com/bench... )

EVGA/NVidia GeForce 970 GTX video OC stock-oem (+140mhz) 4gb GDDR5 RAM (vs. my last vidcard 470 GTX -> http://www.anandtech.com/bench... )

Intel 530 240gb Flash SSD (SATA 6) - strictly OS & Program disk - latest 3.0 firmware & trim tools (vs. my
WD Velociraptor -> http://www.anandtech.com/bench... )

Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb (SATA II) - strictly for backup & programming data

Promise Ex-8350 128mb ECC ram caching raid sata 1/2 controller (SATA 1/2) - for backup WD Velociraptor

GigaByte IRAM 4gb DDR2-Ram based SSD (SATA I) - strictly for PageFile placement

Western Digital 7,200 rpm 8mb buffer 1tb (SATA 6) - strictly for downloads

HP DVD+-RW Dvd 1265i Burner (SATA 3)
8gb Kingston DDR-3 RAM (1gb for 64-bit NTFS
Compressed Software RamDrive = webbrowser cache, hosts file location, print spooler, %TEMP% ops, + %COMSPEC% location)

---

ALL disks = timestamp logical NTFS filesystem data = off + perf counters + excess services off & TCP/IP parameters = security & performance tuned.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentation on it.

I also place my custom hosts file onto it via redirecting where it's referenced by the OS in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via the "DataBasePath" parameter there which acts like a *NIX shadow password system)

I lastly increased hosts' priority assigned to its loads/reads too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008
"Name"="TCP/IP"

* OS = Windows 7 64-bit fully patched & performance tuned + security "hardened" via CIS Tool (95% SCORE).

APK

ASUS B85-E Motherboard

Intel Core I7 4790k CPU (vs. my last CPU Core I7 920 -> http://www.anandtech.com/bench... )

EVGA/NVidia GeForce 970 GTX video OC stock-oem (+140mhz) 4gb GDDR5 RAM (vs. my last vidcard 470 GTX -> http://www.anandtech.com/bench... )

Intel 530 240gb Flash SSD (SATA 6) - strictly OS & Program disk - latest 3.0 firmware & trim tools (vs. my
WD Velociraptor -> http://www.anandtech.com/bench... )

Western Digital 10,000 rpm 8mb buffer Velociraptor 150gb (SATA II) - strictly for backup & programming data

Promise Ex-8350 128mb ECC ram caching raid sata 1/2 controller (SATA 1/2) - for backup WD Velociraptor

GigaByte IRAM 4gb DDR2-Ram based SSD (SATA I) - strictly for PageFile placement

Western Digital 7,200 rpm 8mb buffer 1tb (SATA 6) - strictly for downloads

HP DVD+-RW Dvd 1265i Burner (SATA 3)
8gb Kingston DDR-3 RAM (1gb for 64-bit NTFS
Compressed Software RamDrive = webbrowser cache, hosts file location, print spooler, %TEMP% ops, + %COMSPEC% location)

---

ALL disks = timestamp logical NTFS filesystem data = off + perf counters + excess services off & TCP/IP parameters = security & performance tuned.

Less work done on MAIN OS & Programs bootdisk = faster main drive doing less bs vs. REAL work + reduced fragmentation on it.

I also place my custom hosts file onto it via redirecting where it's referenced by the OS in the registry (for performance + security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Via the "DataBasePath" parameter there which acts like a *NIX shadow password system)

I lastly increased hosts' priority assigned to its loads/reads too:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\Tcpip\ServiceProvider]
"Class"=dword:00000008
"HostsPriority"=dword:00000005
"DnsPriority"=dword:00000006
"LocalPriority"=dword:00000007
"NetbtPriority"=dword:00000008
"Name"="TCP/IP"

* OS = Windows 7 64-bit fully patched & performance tuned + security "hardened" via CIS Tool (95% SCORE).

APK

Here is a link with info about AMD's planned next generation: http://www.anandtech.com/show/...

Oh man, it's worse than that. There are three options:
* 20gbps passive copper cable, USB-C connector, up to 2m long, supports Thunderbolt, USB3.1, and DisplayPort
* 40gbps active copper cable, USB-C connector, up to 2m long, supports Thunderbolt, USB3.1
* 40gbps active optical cable, USB-C connector?, up to 60m long, protocols not yet announced
Notice that you can't use DisplayPort on the 40gbps active cable. So in addition to having ports that look identical but support different functionality, you have cables that look identical but support different functionality.

The Intel CPU is on par with the Godaveri over here http://www.anandtech.com/show/... .
Once you add a R7-240 the AMD chip is faster with dual graphics and sometimes faster by itself. Still cheaper buying AMD and the 240 card than one intel cpu.
Somewhat bogus benchmark because they didn't enable dual graphics for the AMD chips in the intel test. There is reason that you would want a AMD chip to crossfire later if you don't have a lot of money.

It'll be interesting to see how AMD does this time around. Apparently they're making use of HBM (see here for an in-depth description). If they make good use of this technology, which seems likely, they could produce a pretty dramatic increase both in performance and performance per watt. It seems unlikely that nVidia will be able to compete with older-generation GPU's (nVidia will be using similar memory technology next year, so if AMD pulls ahead, it probably won't be for all that long).

See subject: I'm happy w/ that & the 4790x I have now in place of my old 920 (was great for its day & still was outperforming many chips - not "everything" smokes it, far from it).

Kjella noted the 4790x I now use as "the best" along with the Phoronix website that another replier Hairyfeet noted that seconds his motion, stating the same:

"the Core i7 4790K is currently the best desktop offering Intel has until desktop Broadwell CPUs make it out months down the road" from http://www.phoronix.com/scan.p...

IT REALLY SHOWS IN MY CASE, for my needs (one of my main ones).

The NICEST part is that the 4790k was only the cost of roughly 1/3rd of the "TOP DOG" INTEL offers in the 5960x currently

AND

The 4790k puts up a great showing vs. even the 5960x (especially considering that has more cores and far more L2/L3 cache too) -> http://www.anandtech.com/bench...

* :)

(I try to get the MOST out of what I purchase, & usually do, as was the case with my older 920 - then, when I do buy as I stated once every 5-7 years, I see HUGE performance increases... just as I did yet again now, Core I7 920 vs. Core I7 4790k)

APK

P.S.=> Works for me... apk

"If you compare a 3770K (best of 2012) to a 4790K (best of today)" - by Kjella (173770) on Wednesday May 06, 2015 @07:43PM (#49634193) Homepage

Note the subject & bolded part of the quote: I just put one of those together today, & it absolutely SMOKES my last CPU (CoreI7 920) by a HUGE MARGIN when I run my APK Hosts File Engine 9.0++ http://start64.com/index.php?o...

* Over 4++ million records on deduplication, my old 920 (finally died @ mobo level after 7++ yrs.) took 57 minutes - this new Core I7 4790k only takes 31 minutes... made me happy!

(So did your comment!)

I typically buy every 5-7 yrs. & was planning on this 3 months or so from now, but when machines go, they go (what can you do, right?) - & I was forced to a wee bit early... It's nice to know I picked a good choice (I used Anandtech's benchmarks to give me a "rough indicator" of what I'd see 920 vs. 4790k http://www.anandtech.com/bench... , & it's around what I usually do, buying in 5-7 yr. timeframes (big increases usually)).

You're correct it's pretty good (especially in terms of "bang for the buck", which was the basis I picked up the former 920 I had on, vs. 990 etc. from that CPU family by INTEL): It even bears up WELL vs. the "biggest/baddest" from INTEL I know of on that site too, in the 5960X, &, for considerably LESS monetary outlay http://www.anandtech.com/bench...

APK

P.S.=> Nicest part was that I ordered a 4770k, & they told me they might have issues getting THAT one... so, they sent me the better one (which bears out in the benchmark page below too) for the SAME price (bonus!!!)... apk

"If you compare a 3770K (best of 2012) to a 4790K (best of today)" - by Kjella (173770) on Wednesday May 06, 2015 @07:43PM (#49634193) Homepage

Note the subject & bolded part of the quote: I just put one of those together today, & it absolutely SMOKES my last CPU (CoreI7 920) by a HUGE MARGIN when I run my APK Hosts File Engine 9.0++ http://start64.com/index.php?o...

* Over 4++ million records on deduplication, my old 920 (finally died @ mobo level after 7++ yrs.) took 57 minutes - this new Core I7 4790k only takes 31 minutes... made me happy!

(So did your comment!)

I typically buy every 5-7 yrs. & was planning on this 3 months or so from now, but when machines go, they go (what can you do, right?) - & I was forced to a wee bit early... It's nice to know I picked a good choice (I used Anandtech's benchmarks to give me a "rough indicator" of what I'd see 920 vs. 4790k http://www.anandtech.com/bench... , & it's around what I usually do, buying in 5-7 yr. timeframes (big increases usually)).

You're correct it's pretty good (especially in terms of "bang for the buck", which was the basis I picked up the former 920 I had on, vs. 990 etc. from that CPU family by INTEL): It even bears up WELL vs. the "biggest/baddest" from INTEL I know of on that site too, in the 5960X, &, for considerably LESS monetary outlay http://www.anandtech.com/bench...

APK

P.S.=> Nicest part was that I ordered a 4770k, & they told me they might have issues getting THAT one... so, they sent me the better one (which bears out in the benchmark page below too) for the SAME price (bonus!!!)... apk

I've got PC software I can't run on my PC without using an emulator because 64-bit windows won't run it. Also, there's no longer any such thing as "running natively" - today's x86 actually takes every instruction and translates it into microcode - in other words, the cpu is really just an x86 emulator.

If you look at the anandtech review, the original model #s were PPSTCK1A32WFC and PPSTCK1A8LFC.
Seems the PC police noticed and cut off the peepee.

Sorry, comparing a drive with 23% overprovisioning to a drive with 7% OP is no "proof".
Take a look at these
Pay close attention to the 256GB 850 pro at 12% and 25% OP.
Hell, even the "old" 256GB 840 pro at 25% OP can push ~30k sustained 4k random write IOPS.

840 EVO by chance? There's a confirmed bug (with firmware update and reconditioning process) that will slow an 840 EVP to a crawl. I've personally seen it happen with several laptops recently upgraded. Once I applied the update, performance resumed back to original spec. And, these were full from anywhere from 60% to 80%; didn't matter much. Link below for update

Samsung SSD 840 EVO Performance Restoration Software

In fact, that fix does not really FIX the problem, it just refreshes the cells (by reading/writing all the data), but then performance slowly deteriorates again. Apparently, Samsung will make a "real" fix available later this month.

http://anandtech.com/show/9158...

I would be surprised if a high-end smartphone in the world could out-compute a reasonably spec'ed desktop from the early 2000s (which was point at which computers began to rather consistently beat grandmasters.) The lack of CPU fan is the biggest limiting factor of all.

An iPhone 6 can do 77 GFLOPS. Deep Blue could only manage 11 GFLOPS. Now, Deep Blue had specialized VLSI chips that are hard to measure, and chess computations are going to be mostly integer, not floating-point, but the point stands that a modern phone has plenty of computing power for crushing puny meatbags at chess.

...is capable of sequential read and write speeds of 2,260 MB/sec and 1,600 MB/sec respectively. Comparable SATA-based M.2 SSDs typically can only push read/write speeds of 540 MB/sec and 500 MB/sec,

Non-SATA M.2 drives are already on the market. Comparing the newest drive to SATA-based M.2 drives does not help much, I'd rather see it compared to what it supersedes. In this case, I'm more interested in a comparison with a PCIe 3.0 4-lane M.2 SSD drive that doesn't support NVMe. The drive specification for the earlier non-NVMe SM951 is not that far off of that of the new drive. The earlier drive is rated at sequential read and write speeds of 2,150 MB/sec 1,500 MB/sec respectively. Again, not all that far off.

That being said...I'm curious to see the difference that NVMe makes in real-world benchmarks, and where the difference is...especially because I just built a new system with a non-NVMe SM951 SSD. :)

Turns out the initial fix didn't solve the issue completely. There's another fix coming soon: http://www.anandtech.com/show/...

whew. for a second there, i thought anandtech got scooped.

Does Samsung even have a competing product?

Yes. SM915.

Which seems to not exist according to Google. Perhaps you mean the SM951? This is a mSATA card that runs over PCIe, not a desktop computer card that runs over a PCIe slot. They are completely different things. Look at the size of the Intel card, it is a full daughter board with many large flash chips. The Samsung has like 3 chips on it. I would expect this would be a much lesser part that isn't designed for enterprise level loads like the Intel card is designed for.

Apple is using Samsung for mSATA, not PCIe

Guess which PCIe 3.0 x4 drive the PCIe SSDs in the 2015 mbp and mb are based on.

Which is why only quoting a portion of the sentence is silly. I said they are electrically similar, however, they are physically very different.

and theirs doesn't push near the speed of this product as this one pushes 2.4GByte/sec read and 1.2GB/s write.

Yeah, it's not like a 512GB SM915 manages to outperform a 1.2TB 750 for real world I/O traces.
Oh.

Interesting, as the response time was faster for the Intel, not the Samsung, and the performance figures were less than half, I wonder if maybe AnandTech had their card misconfigured. The Samsungs on the test (of which one was slower and one was marginally faster) were both in AHCI mode, whereas the Intel was in NVMe mode, which according to the spec sheet isn't correct for the Intel (should also be in AHCI mode). Perhaps they had an old bios that doesn't support this card yet, or were in some other way messing up the test. SSDs don't magically perform at half speed, something has to be done to cause it to happen.

Does Samsung even have a competing product?

Yes. SM915.

Apple is using Samsung for mSATA, not PCIe

Guess which PCIe 3.0 x4 drive the PCIe SSDs in the 2015 mbp and mb are based on.

and theirs doesn't push near the speed of this product as this one pushes 2.4GByte/sec read and 1.2GB/s write.

Yeah, it's not like a 512GB SM915 manages to outperform a 1.2TB 750 for real world I/O traces.
Oh.

I see not one thing that says this is an x86. If it's not x86 it's still ARM and still windows RT even if they don't call it RT anymore. The result being you can only run software from the windows store, no legacy apps.

Its a 14nm Cherry Trail SoC. Don't confuse this with old crappy Atom. These are really fast. This is http://www.anandtech.com/show/...

Too bad the Titan X the article is about has the same 1/32 FP64 as the GTX 980, and NOT the 1/3 FP64 of the previous Titan cards. Source: http://www.anandtech.com/show/... GP is correct. This card is for suckers. People who want to play at 4k will be far better served by SLI/CF still, everyone else shouldn't be looking at either option, and instead getting the far-less-expensive-and-more-than-adequate-at-2560x1440 R9 290, R9 290x, GTX 970, or GTX980 single cards instead.

Less power consumption and better reception:
http://www.anandtech.com/show/...
On top of that they are a node ahead of Intel modems, which still use 28nm.

Qualcomm still seems to be at the forefront of cell modem design but the alternatives like Intel are probably becoming good enough at this point. They showed off some stuff at MWC: http://www.anandtech.com/show/...

Which is probably why OCZ was acquired by Toshiba.

There was a known issue with the Evo drives that samsung issued a fix for: http://www.anandtech.com/show/...

And USB 3 does not do everything I use Thunderbolt for on my Mac, including ferry USB3 over the same wire as video.

USB-C is in fact USB 3.1, and it very much does ferry USB and video over the same wire. VESA has standardized DisplayPort over USB-C. VESA's press release can be found here: http://www.vesa.org/news/vesa-... or AnandTech had a good article here: http://www.anandtech.com/show/...

and plug in my laptop with single cable and instantly my displays, USB3 devices, audio and networking all work ...

USB 3.1 has the same bandwidth as Thunderbolt 1 (10Gbps), there's no reason why a USB-C dock couldn't do all that, and be much cheaper than a Thunderbolt dock in the process.

USB-C also supports far more power delivery than Thunderbolt. Normal devices get up to 15W (Thunderbolt does ~10W), or devices can draw up to 100W if they implement v2 of the power delivery spec.

According to http://techreport.com/review/2... Intel TSX errata has been corrected

This can improve database performance (see http://www.anandtech.com/show/... )

If you leave biases aside, and want to compare the Macbook with something else, the Dell XPS 13 would be neck to neck. I would even venture to say that the Dell is far more computer for the money (i will qualify why). Here's a comparison.

http://www.theverge.com/2015/3...

The Dell XPS 13 is extremely well built (aluminum alloy, carbon fibre - I've held it, it is gorgeous) - might possibly be a hair lower than the new Macbook (which I haven't held) but really, I think you would be splitting hairs.

The XPS 13 however has two big things going for it - it has full blown Core i3/i5/i7 (2.1 - 2.4 GHz) as opposed to Macbook's Core M (measly 1.1 GHz) - which means much better performance (i would imagine 2x-3x better), much less thermal throttling, and better graphics (HD5500 vs 5300). The other big thing is its display. XPS 13 has a near zero bezel 3200 x 1800 pixel Sharp IGZO panel that is arguably the best and most cutting edge laptop panel one can get today.

You can read a review of this panel here: http://www.anandtech.com/show/...

And dimensions and weight. Macbook is 11" wide and 0.52" thick, and weighs 2 pounds. The XPS 13 is a bit heavier (2.6-2.8 pounds), but is only 12" wide and 0.33"-0.6" thick. The cool thing is that because of the near zero bezels, XPS 13 is a 13" screen while only being 12" wide (typical for a 11" laptop, not for a 13" laptop).

Again, I am not saying Macbook is not good. It still seems to have Apple's obsessive attention to detail in terms of build quality and user centric design approach. But to say that it has no competion - that is no longer true. I do believe that the XPS 13 is a genuine alternative in just about every respec. The extra 0.8 may be an issue for some, but you also sacrifice a *lot* of computer for that. Then I would say, might as well get an iPad.

DirectX 12 isn't out yet and appears to be a Windows 10 only thing. Quick check shows windows 7 has DX11. I'm assuming you know what DirectX is ;-)

Source:
http://www.anandtech.com/show/...

eh?? I have an Asrock P4i65G Prescott P4 board next to me with an ECP parallel port on it. That's a 2006 vintage.

You are correct, it looks like that board was released in March of 2006, but it really doesn't help the case that parallel ports aren't outdated when it was a legacy support board even when it was released. LGA775 had been out for 18 months at the time and the AGP slot just speaks for itself.

(and blow me, it still works with the original processor I bought for it as well, a 2.66GHz P4)

You might want to grab a Kill-a-Watt or similar and test your power consumption. Prescotts weren't exactly known for being efficient in the first place and a lot has changed in the CPU world since then. Anandtech's CPU Benchmark Database has the Pentium 4 HT 660 which is a year newer, a full gigahertz faster, and has twice the L2 as yours. When you factor for the clock difference, a 10 watt Celeron is pretty much just as fast in single threaded loads and since everything has multiple cores these days multithreaded performance will be a whole different world.

With how cheaply you can get CPU power these days anything from the P4 era or older can be hard to justify keeping around both for power/heat reasons (particularly notable on Prescott chips) and performance. If you actually use that day-to-day I guarantee that a dual core chip would be practically a religious experience by now, not to mention if your electric costs are anything you care about it very well could save you enough to cover a lot of the upgrade price over the course of a few years.

I'm not saying you need really awesome gear, just that even cheap hardware these days is hugely better than that. Until a few months ago I was running on a Phenom X6 1045 that cost me $90 brand new when I bought it over two years prior. It still did just fine for me day-to-day and I felt no need to upgrade. Yeah it's six cores, but an end-of-life chip that was sub-$100 when bought is by no means a monster build.

We reached the point where for most day-to-day tasks you generally don't need any more performance some time ago, but that doesn't mean there's no reason to replace old hardware. I do contract IT work for a bunch of customers and at this point my line is Core 2 Duo. Anything older gets replaced when possible (made a lot easier by XP being EoLed with no security updates), anything newer gets upgraded to 4+GB of RAM and a strong recommendation of a SSD. Even a first-gen C2D is pretty hard to tell from a Core i7 in most desktop tasks if it has enough RAM and the SSD.

A comparision of a first generation desktop i3 (which is slightly newer than a first generation i7) from january 2010 to a current generation desktop i3 from may 2014 (there was a slight speed bump released in july but anandtech don't have that one in their list) can be seen at http://www.anandtech.com/bench... . We see that performance has less than doubled in over four years

We see a similar comparison when we compare a first generation desktop i5 from september 2009 to a current one from may 2014 http://www.anandtech.com/bench...

I'm not sure i'd consider it insane for a high end desktop part to have double the performance of a contemporary desktop part. I think it's more that we just aren't making the massive jumps in performance anymore that came from the move from 1->2->4 cores as the typical core count in the mid-mainstream or that came from retiring the crappy pentium 4 architecture.

A comparision of a first generation desktop i3 (which is slightly newer than a first generation i7) from january 2010 to a current generation desktop i3 from may 2014 (there was a slight speed bump released in july but anandtech don't have that one in their list) can be seen at http://www.anandtech.com/bench... . We see that performance has less than doubled in over four years

We see a similar comparison when we compare a first generation desktop i5 from september 2009 to a current one from may 2014 http://www.anandtech.com/bench...

I'm not sure i'd consider it insane for a high end desktop part to have double the performance of a contemporary desktop part. I think it's more that we just aren't making the massive jumps in performance anymore that came from the move from 1->2->4 cores as the typical core count in the mid-mainstream or that came from retiring the crappy pentium 4 architecture.

x86 did gain reliability features years ago, with the Nehalem-EX series and successors.

http://www.anandtech.com/show/...

Not sure if that's close enough for you. A year ago there were some additional RAS features (lower quality article : ) http://semiaccurate.com/2014/0...
Perhaps it doesn't go as far as the most paranoid mainframes but I wonder if such systems can be called a minicomputer.