The 6200U score is for the 64-bit version of Geekbench, while the 2500/5000 you quote is for 32-bit versions. 64-bit scores for 5200U are about 2800/5500.
No they are not. Plenty of 64-bit scores in the 2600/5200 range.
Highest Core i5 5200U Windows 64-bit score is 2766/5644, the first Skylake-U score is already faster than that.
Didn't know it was the fastest one... No idea why either. It was a clean Windows 7 installation, but no other "special treatment". But that's great, I guess. And perhaps puts the Skylake score in a little better light.
That's not quite true. You're sorting by multi-core score. If you sort by just single-core score you'll see the highest Windows single-core score is 2860.
Well, the 5300U is closer to the 6200U in base frequency (2.3GHz vs 2.4GHz), so I figured it was a decent match for an architectural comparison. No idea how well the turbo frequencies of those two match, though. Do we have that info for Skylake-U yet?Is there a reason we are comparing the 6200u to the 5300u? The 5300u has a 200mhz turbo clock advantage on the 5200u.
At any rate, comparing the 5200u to the 6200u, we can say that our first Skylake score has a 2% advantage single-core advantage over the best 5200u (windows) single-core score and a 9.7% advantage over the best multi-core score.
Not a bad way to start I guess. Another thing to keep in mind, the Skylake score was done with Windows 10. And I haven't found a Windows 10 score anywhere in the top marks for the 5200u.
K model is august 5th. Non K models is late september.
It is true. I was comparing single scores, I didn't look for the best ST and MT results individually.
Brunnis said:Well, the 5300U is closer to the 6200U in base frequency (2.3GHz vs 2.4GHz), so I figured it was a decent match for an architectural comparison. No idea how well the turbo frequencies of those two match, though. Do we have that info for Skylake-U yet?
The 6200U has a 9% advantage in base frequency over the 5200U. If some or all of that remains during turbo, it would account for a decent performance lead for the Skylake
Expect the 6700k to beat the 5820k by ~7% in everything but benchmarks that can utilize 6 cores. Basically, the 6700k is going to win in single-threaded measurements...the 4790k already wins in those.Skylake will be a good jump over Nehalem: OK.
Who cares if it's 70% IPC improvement or 100%: OK.
The real question is: since they will be priced roughly the same, will the 6700k be better than a 5820k in real life usage?
Even considering a 6700k OCed to 4.6-4.8 for daily usage and a 5820k OCed to 4-4.2 ghz. Will the higher clock and IPC, compensate for the lower memory bandwidth and the lack of 2 cores? I honestly don't think so.
I just think Skylake looks a great step forward if you look at performance per watt and iGPU performance but in terms of brute force will still be meh .
Anyway I'm really looking forward to some 6700k benchmarks because I will have to build a nice workstation in August / September and I guess the best bang for buck will be either the 6700k or the 5820k.
Skylake will be a good jump over Nehalem: OK.
Who cares if it's 70% IPC improvement or 100%: OK.
The real question is: since they will be priced roughly the same, will the 6700k be better than a 5820k in real life usage?
Even considering a 6700k OCed to 4.6-4.8 for daily usage and a 5820k OCed to 4-4.2 ghz. Will the higher clock and IPC, compensate for the lower memory bandwidth and the lack of 2 cores? I honestly don't think so.
I just think Skylake looks a great step forward if you look at performance per watt and iGPU performance but in terms of brute force will still be meh .
Anyway I'm really looking forward to some 6700k benchmarks because I will have to build a nice workstation in August / September and I guess the best bang for buck will be either the 6700k or the 5820k.
Here's the first picture of GIGABYTE's upcoming Gaming 3 motherboard. We're not allowed to name the chipset, that's as well-guarded a secret as Coca Cola's recipe, so you'll just have to use your imagination. This socket LGA1151 motherboard, is designed to strike a price-features sweetspot, and supports 2-way NVIDIA SLI and AMD CrossFireX. Built in the ATX form-factor, the board draws power from a combination of 24-pin ATX and 8-pin EPS connectors, and offers a 7-phase CPU VRM. The socket is wired to four DDR4 DIMM slots, supporting up to 64 GB of dual-channel DDR4 memory, and two PCI-Express 3.0 x16 slots (x8/x8, when both are populated). A third PCI-Express x16 slot (electrical gen 3.0 x4), is wired to the PCH, along with three other PCI-Express 3.0 x1 slots.
Storage connectivity includes two 32 Gb/s M.2 slots, three SATA-Express 16 Gb/s, and six SATA 6 Gb/s ports. Among the USB connectivity are two USB 3.1 ports (with at least one type-C port), at least eight USB 3.0 ports (from which four are given out via headers), and a number of USB 2.0/1.1 ports. The AMP-UP audio solution combines a 115 dBA SNR CODEC with ground-layer isolation, an OPAMP tuned for gaming and music, audio-grade capacitors, and EMI shielding. Networking is care of a Killer E2200 chip. Display outputs include D-Sub, DVI, and HDMI. GIGABYTE-exclusive Dual-UEFI BIOS is featured. Also offered is TurboOverclock+, a hardware-based overclock automation and stability testing system. This board will be a part of GIGABYTE's first wave of next-generation motherboards. It will include keys to some cool World of Tanks DLC.
No estimated price? For $130 or less, that looks like a pretty sweet board. SLI / CF, (PCI-E 3.0 x8/x8), dual 32Gbit/sec M.2 slots, etc. The only downside is Killer E2200 NIC, rather than Intel LAN.GIGABYTE's Next Gaming 3 Motherboard Pictured
More supercooled 6700K numbers?
1.9V?
More supercooled 6700K numbers?
1.9V?
That seems legit and it's very bad news... Going over 1.8V is lethal... And to reach only 5Ghz it might reveal that the chip will be harder to OC past some point.
That's a bit insane indeed, from such a voltage you could expect 6GHz, not 5... other shots in the link:
Why are there still USB2/1 ports on intel chipsets?
Are there Linux drivers yet for the E2200? If there are, then that board might be viable for me, otherwise, I'll stick with Intel LAN, thanks.
Yes please. I did a few days ago a full world recompile due to gcc-5.2.0. Here are my times:costel78, I use gentoo too and I have a i7-4790K. I can check the compiling times for the top offenders (chromium, firefox) so you can have an idea of the jump in performance (at least for haswell). Just give me a couple of hours to arrive home.
So... Maybe at 4.7 to above can be the wall to the low voltage.That's a bit insane indeed, from such a voltage you could expect 6GHz, not 5... other shots in the link:
HEDT has always been a generation behind.I wish Skylake-E was dropping next month instead of Skylake...sigh
hard to want to invest in x99 now, with it being 2 gen's behind when skylake drops. bleh